diff --git a/pytorch/output/altra_10_30_Oregon-2_1000.json b/pytorch/output/altra_10_30_Oregon-2_1000.json deleted file mode 100644 index 3ce3252..0000000 --- a/pytorch/output/altra_10_30_Oregon-2_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [37.36, 22.88, 22.36, 22.72, 22.52, 22.2, 21.96, 21.8, 21.48, 21.48], "matrix": "Oregon-2", "shape": [11806, 11806], "nnz": 65460, "% density": 0.0004696458003979807, "time_s": 1.5312557220458984, "power": [26.68, 27.84, 28.48, 29.92, 30.0], "power_after": [21.16, 21.32, 21.16, 21.16, 21.16, 20.88, 20.92, 20.76, 20.96, 21.2], "task clock (msec)": 64.81, "page faults": 3244, "cycles": 82069432, "instructions": 78292700, "branch mispredictions": 319703, "branches": 19996903, "ITLB accesses": 26988315, "ITLB misses": 5988, "DTLB misses": 14570, "DTLB accesses": 36879854, "L1I cache accesses": 30465174, "L1I cache misses": 293085, "L1D cache misses": 487330, "L1D cache accesses": 31932249, "LL cache misses": 545501, "LL cache accesses": 558084, "L2D TLB accesses": 204746, "L2D TLB misses": 25302, "L2D cache misses": 314594, "L2D cache accesses": 1828047, "instructions per cycle": 0.9539812582107307, "branch miss rate": 0.01598762568383714, "ITLB miss rate": 0.00022187379982781437, "DTLB miss rate": 0.0003950666399058955, "L2D TLB miss rate": 0.12357750578765886, "L1I cache miss rate": 0.009620329101025322, "L1D cache miss rate": 0.015261374167538278, "L2D cache miss rate": 0.17209294947011755, "LL cache miss rate": 0.9774532149282187} diff --git a/pytorch/output/altra_10_30_Oregon-2_1000.output b/pytorch/output/altra_10_30_Oregon-2_1000.output deleted file mode 100644 index 78b89f4..0000000 --- a/pytorch/output/altra_10_30_Oregon-2_1000.output +++ /dev/null @@ -1,158 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3394980 queued and waiting for resources -srun: job 3394980 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 583, 584, ..., 65459, 65460, 65460]), - col_indices=tensor([ 2, 23, 27, ..., 3324, 958, 841]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(11806, 11806), - nnz=65460, layout=torch.sparse_csr) -tensor([0.9231, 0.7723, 0.0509, ..., 0.0839, 0.6982, 0.3459]) -Matrix: Oregon-2 -Shape: torch.Size([11806, 11806]) -NNZ: 65460 -Density: 0.0004696458003979807 -Time: 1.5677142143249512 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/Oregon-2.mtx 1000': - - 64.81 msec task-clock:u # 0.013 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,244 page-faults:u # 50.056 K/sec - 82,069,432 cycles:u # 1.266 GHz (59.04%) - 78,292,700 instructions:u # 0.95 insn per cycle (76.75%) - branches:u - 341,509 branch-misses:u (90.97%) - 33,032,555 L1-dcache-loads:u # 509.704 M/sec - 478,674 L1-dcache-load-misses:u # 1.45% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 31,508,310 L1-icache-loads:u # 486.184 M/sec - 297,528 L1-icache-load-misses:u # 0.94% of all L1-icache accesses - 49,358,091 dTLB-loads:u # 761.613 M/sec (27.83%) - 88,514 dTLB-load-misses:u # 0.18% of all dTLB cache accesses (14.82%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 5.016393105 seconds time elapsed - - 16.759527000 seconds user - 31.429551000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 583, 584, ..., 65459, 65460, 65460]), - col_indices=tensor([ 2, 23, 27, ..., 3324, 958, 841]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(11806, 11806), - nnz=65460, layout=torch.sparse_csr) -tensor([0.8423, 0.9339, 0.8037, ..., 0.5953, 0.0649, 0.1559]) -Matrix: Oregon-2 -Shape: torch.Size([11806, 11806]) -NNZ: 65460 -Density: 0.0004696458003979807 -Time: 1.516484022140503 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/Oregon-2.mtx 1000': - - 319,703 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 19,996,903 BR_RETIRED:u - - 4.945699041 seconds time elapsed - - 16.431978000 seconds user - 29.752452000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 583, 584, ..., 65459, 65460, 65460]), - col_indices=tensor([ 2, 23, 27, ..., 3324, 958, 841]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(11806, 11806), - nnz=65460, layout=torch.sparse_csr) -tensor([0.8058, 0.2922, 0.1227, ..., 0.2176, 0.9496, 0.8838]) -Matrix: Oregon-2 -Shape: torch.Size([11806, 11806]) -NNZ: 65460 -Density: 0.0004696458003979807 -Time: 1.6458909511566162 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/Oregon-2.mtx 1000': - - 26,988,315 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 5,988 ITLB_WALK:u - 14,570 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 36,879,854 L1D_TLB:u - - 5.011871473 seconds time elapsed - - 16.529942000 seconds user - 30.438432000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 583, 584, ..., 65459, 65460, 65460]), - col_indices=tensor([ 2, 23, 27, ..., 3324, 958, 841]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(11806, 11806), - nnz=65460, layout=torch.sparse_csr) -tensor([0.7728, 0.1182, 0.3337, ..., 0.2555, 0.2523, 0.5746]) -Matrix: Oregon-2 -Shape: torch.Size([11806, 11806]) -NNZ: 65460 -Density: 0.0004696458003979807 -Time: 1.529954433441162 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/Oregon-2.mtx 1000': - - 30,465,174 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 293,085 L1I_CACHE_REFILL:u - 487,330 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 31,932,249 L1D_CACHE:u - - 4.954100105 seconds time elapsed - - 16.282966000 seconds user - 28.926724000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 583, 584, ..., 65459, 65460, 65460]), - col_indices=tensor([ 2, 23, 27, ..., 3324, 958, 841]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(11806, 11806), - nnz=65460, layout=torch.sparse_csr) -tensor([0.5613, 0.3211, 0.1739, ..., 0.5461, 0.1391, 0.8387]) -Matrix: Oregon-2 -Shape: torch.Size([11806, 11806]) -NNZ: 65460 -Density: 0.0004696458003979807 -Time: 1.5726752281188965 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/Oregon-2.mtx 1000': - - 545,501 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 558,084 LL_CACHE_RD:u - 204,746 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 25,302 L2D_TLB_REFILL:u - 314,594 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,828,047 L2D_CACHE:u - - 4.866549675 seconds time elapsed - - 16.609257000 seconds user - 31.381282000 seconds sys - - - diff --git a/pytorch/output/altra_10_30_as-caida_1000.json b/pytorch/output/altra_10_30_as-caida_1000.json deleted file mode 100644 index dbd6ca7..0000000 --- a/pytorch/output/altra_10_30_as-caida_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [21.6, 21.64, 21.88, 22.08, 22.2, 22.32, 22.36, 22.04, 22.0, 21.96], "matrix": "as-caida", "shape": [31379, 31379], "nnz": 106762, "% density": 0.00010842726485909405, "time_s": 2.6254467964172363, "power": [30.92, 29.2, 29.52, 29.72, 29.72, 31.72], "power_after": [21.04, 21.28, 21.04, 21.16, 21.16, 20.96, 21.04, 20.88, 20.56, 20.84], "task clock (msec)": 61.4, "page faults": 3507, "cycles": 78967021, "instructions": 94334531, "branch mispredictions": 325893, "branches": 19069753, "ITLB accesses": 27181279, "ITLB misses": 5995, "DTLB misses": 17412, "DTLB accesses": 37016930, "L1I cache accesses": 31535482, "L1I cache misses": 292676, "L1D cache misses": 471752, "L1D cache accesses": 33119145, "LL cache misses": 540894, "LL cache accesses": 554700, "L2D TLB accesses": 191772, "L2D TLB misses": 23711, "L2D cache misses": 306195, "L2D cache accesses": 1755986, "instructions per cycle": 1.1946066827061894, "branch miss rate": 0.017089523917797993, "ITLB miss rate": 0.00022055621444450792, "DTLB miss rate": 0.00047037936425305935, "L2D TLB miss rate": 0.12364161608576851, "L1I cache miss rate": 0.009280847522799873, "L1D cache miss rate": 0.01424408752097918, "L2D cache miss rate": 0.17437211913990203, "LL cache miss rate": 0.975110870740941} diff --git a/pytorch/output/altra_10_30_as-caida_1000.output b/pytorch/output/altra_10_30_as-caida_1000.output deleted file mode 100644 index 70077ee..0000000 --- a/pytorch/output/altra_10_30_as-caida_1000.output +++ /dev/null @@ -1,163 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3394983 queued and waiting for resources -srun: job 3394983 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 0, 0, ..., 106761, 106761, - 106762]), - col_indices=tensor([ 106, 329, 1040, ..., 155, 160, 12170]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379), - nnz=106762, layout=torch.sparse_csr) -tensor([0.4886, 0.3652, 0.5691, ..., 0.6466, 0.4355, 0.8397]) -Matrix: as-caida -Shape: torch.Size([31379, 31379]) -NNZ: 106762 -Density: 0.00010842726485909405 -Time: 2.6297245025634766 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/as-caida.mtx 1000': - - 61.40 msec task-clock:u # 0.010 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,507 page-faults:u # 57.117 K/sec - 78,967,021 cycles:u # 1.286 GHz (61.13%) - 94,334,531 instructions:u # 1.19 insn per cycle (95.16%) - branches:u - 365,239 branch-misses:u - 33,334,312 L1-dcache-loads:u # 542.906 M/sec - 457,950 L1-dcache-load-misses:u # 1.37% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 31,725,851 L1-icache-loads:u # 516.709 M/sec - 297,720 L1-icache-load-misses:u # 0.94% of all L1-icache accesses - 25,188,580 dTLB-loads:u # 410.239 M/sec (5.16%) - dTLB-load-misses:u (0.00%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 6.049042045 seconds time elapsed - - 17.649315000 seconds user - 29.335859000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 0, 0, ..., 106761, 106761, - 106762]), - col_indices=tensor([ 106, 329, 1040, ..., 155, 160, 12170]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379), - nnz=106762, layout=torch.sparse_csr) -tensor([0.8344, 0.2588, 0.2246, ..., 0.5607, 0.8141, 0.9893]) -Matrix: as-caida -Shape: torch.Size([31379, 31379]) -NNZ: 106762 -Density: 0.00010842726485909405 -Time: 2.6495532989501953 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/as-caida.mtx 1000': - - 325,893 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 19,069,753 BR_RETIRED:u - - 6.023780447 seconds time elapsed - - 17.654658000 seconds user - 28.848805000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 0, 0, ..., 106761, 106761, - 106762]), - col_indices=tensor([ 106, 329, 1040, ..., 155, 160, 12170]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379), - nnz=106762, layout=torch.sparse_csr) -tensor([0.0814, 0.1132, 0.8515, ..., 0.8987, 0.5912, 0.5002]) -Matrix: as-caida -Shape: torch.Size([31379, 31379]) -NNZ: 106762 -Density: 0.00010842726485909405 -Time: 2.5444185733795166 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/as-caida.mtx 1000': - - 27,181,279 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 5,995 ITLB_WALK:u - 17,412 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 37,016,930 L1D_TLB:u - - 5.790360666 seconds time elapsed - - 17.919315000 seconds user - 30.569858000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 0, 0, ..., 106761, 106761, - 106762]), - col_indices=tensor([ 106, 329, 1040, ..., 155, 160, 12170]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379), - nnz=106762, layout=torch.sparse_csr) -tensor([0.0439, 0.1884, 0.3342, ..., 0.2027, 0.5532, 0.7245]) -Matrix: as-caida -Shape: torch.Size([31379, 31379]) -NNZ: 106762 -Density: 0.00010842726485909405 -Time: 2.620804786682129 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/as-caida.mtx 1000': - - 31,535,482 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 292,676 L1I_CACHE_REFILL:u - 471,752 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 33,119,145 L1D_CACHE:u - - 6.002311801 seconds time elapsed - - 17.427887000 seconds user - 30.063688000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 0, 0, ..., 106761, 106761, - 106762]), - col_indices=tensor([ 106, 329, 1040, ..., 155, 160, 12170]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379), - nnz=106762, layout=torch.sparse_csr) -tensor([0.1495, 0.5856, 0.8600, ..., 0.2101, 0.6229, 0.2019]) -Matrix: as-caida -Shape: torch.Size([31379, 31379]) -NNZ: 106762 -Density: 0.00010842726485909405 -Time: 2.561279296875 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/as-caida.mtx 1000': - - 540,894 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 554,700 LL_CACHE_RD:u - 191,772 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 23,711 L2D_TLB_REFILL:u - 306,195 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,755,986 L2D_CACHE:u - - 5.946428572 seconds time elapsed - - 17.396567000 seconds user - 32.141235000 seconds sys - - - diff --git a/pytorch/output/altra_10_30_dc2_1000.json b/pytorch/output/altra_10_30_dc2_1000.json deleted file mode 100644 index 9e49406..0000000 --- a/pytorch/output/altra_10_30_dc2_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [83.04, 78.44, 65.92, 53.76, 38.68, 38.68, 25.68, 22.6, 22.52, 22.32], "matrix": "dc2", "shape": [116835, 116835], "nnz": 766396, "% density": 5.614451099680581e-05, "time_s": 14.128849267959595, "power": [89.84, 89.4, 82.8, 71.32, 57.72, 51.92, 53.0, 63.8, 78.24, 78.24, 90.2, 90.36, 90.08, 88.64, 88.64, 87.64, 87.68, 87.24], "power_after": [21.4, 21.2, 21.08, 21.08, 21.28, 21.04, 20.92, 21.12, 21.08, 21.0], "task clock (msec)": 58.45, "page faults": 3471, "cycles": 76691414, "instructions": 89547095, "branch mispredictions": 329725, "branches": 19946857, "ITLB accesses": 27648951, "ITLB misses": 6857, "DTLB misses": 18047, "DTLB accesses": 37225736, "L1I cache accesses": 32434686, "L1I cache misses": 293072, "L1D cache misses": 483557, "L1D cache accesses": 34059722, "LL cache misses": 561480, "LL cache accesses": 578369, "L2D TLB accesses": 192306, "L2D TLB misses": 25364, "L2D cache misses": 317121, "L2D cache accesses": 1812330, "instructions per cycle": 1.16762868656979, "branch miss rate": 0.01653017314958442, "ITLB miss rate": 0.00024800217556174194, "DTLB miss rate": 0.00048479901109275584, "L2D TLB miss rate": 0.13189396066685385, "L1I cache miss rate": 0.00903575881696527, "L1D cache miss rate": 0.014197326683993487, "L2D cache miss rate": 0.17497972223601663, "LL cache miss rate": 0.9707989190292011} diff --git a/pytorch/output/altra_10_30_dc2_1000.output b/pytorch/output/altra_10_30_dc2_1000.output deleted file mode 100644 index f3e0bb7..0000000 --- a/pytorch/output/altra_10_30_dc2_1000.output +++ /dev/null @@ -1,173 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3394982 queued and waiting for resources -srun: job 3394982 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 1, 2, ..., 766390, 766394, - 766396]), - col_indices=tensor([ 0, 1, 2, ..., 116833, 89, - 116834]), - values=tensor([-1.0000e+00, -1.0000e+00, -1.0000e+00, ..., - 1.0331e+01, -1.0000e-03, 1.0000e-03]), - size=(116835, 116835), nnz=766396, layout=torch.sparse_csr) -tensor([0.0986, 0.6504, 0.0132, ..., 0.6525, 0.3337, 0.7557]) -Matrix: dc2 -Shape: torch.Size([116835, 116835]) -NNZ: 766396 -Density: 5.614451099680581e-05 -Time: 18.46260714530945 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/dc2.mtx 1000': - - 58.45 msec task-clock:u # 0.003 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,471 page-faults:u # 59.382 K/sec - 76,691,414 cycles:u # 1.312 GHz (41.20%) - 89,547,095 instructions:u # 1.17 insn per cycle (73.16%) - branches:u - 382,362 branch-misses:u (96.21%) - 33,271,433 L1-dcache-loads:u # 569.211 M/sec - 488,730 L1-dcache-load-misses:u # 1.47% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 31,926,596 L1-icache-loads:u # 546.204 M/sec - 304,792 L1-icache-load-misses:u # 0.95% of all L1-icache accesses - 36,392,791 dTLB-loads:u # 622.612 M/sec (31.21%) - 0 dTLB-load-misses:u (5.35%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 22.126601025 seconds time elapsed - - 103.642372000 seconds user - 1434.131491000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 1, 2, ..., 766390, 766394, - 766396]), - col_indices=tensor([ 0, 1, 2, ..., 116833, 89, - 116834]), - values=tensor([-1.0000e+00, -1.0000e+00, -1.0000e+00, ..., - 1.0331e+01, -1.0000e-03, 1.0000e-03]), - size=(116835, 116835), nnz=766396, layout=torch.sparse_csr) -tensor([0.5605, 0.9374, 0.4444, ..., 0.5937, 0.3099, 0.2252]) -Matrix: dc2 -Shape: torch.Size([116835, 116835]) -NNZ: 766396 -Density: 5.614451099680581e-05 -Time: 13.607120752334595 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/dc2.mtx 1000': - - 329,725 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 19,946,857 BR_RETIRED:u - - 17.131143957 seconds time elapsed - - 96.945305000 seconds user - 1045.242697000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 1, 2, ..., 766390, 766394, - 766396]), - col_indices=tensor([ 0, 1, 2, ..., 116833, 89, - 116834]), - values=tensor([-1.0000e+00, -1.0000e+00, -1.0000e+00, ..., - 1.0331e+01, -1.0000e-03, 1.0000e-03]), - size=(116835, 116835), nnz=766396, layout=torch.sparse_csr) -tensor([0.8954, 0.9777, 0.8042, ..., 0.2069, 0.7063, 0.8479]) -Matrix: dc2 -Shape: torch.Size([116835, 116835]) -NNZ: 766396 -Density: 5.614451099680581e-05 -Time: 17.22396969795227 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/dc2.mtx 1000': - - 27,648,951 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 6,857 ITLB_WALK:u - 18,047 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 37,225,736 L1D_TLB:u - - 20.911480243 seconds time elapsed - - 107.392462000 seconds user - 1329.272154000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 1, 2, ..., 766390, 766394, - 766396]), - col_indices=tensor([ 0, 1, 2, ..., 116833, 89, - 116834]), - values=tensor([-1.0000e+00, -1.0000e+00, -1.0000e+00, ..., - 1.0331e+01, -1.0000e-03, 1.0000e-03]), - size=(116835, 116835), nnz=766396, layout=torch.sparse_csr) -tensor([0.9293, 0.9606, 0.8914, ..., 0.2407, 0.2843, 0.5174]) -Matrix: dc2 -Shape: torch.Size([116835, 116835]) -NNZ: 766396 -Density: 5.614451099680581e-05 -Time: 13.233965873718262 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/dc2.mtx 1000': - - 32,434,686 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 293,072 L1I_CACHE_REFILL:u - 483,557 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 34,059,722 L1D_CACHE:u - - 16.956477005 seconds time elapsed - - 88.393687000 seconds user - 1037.101858000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 1, 2, ..., 766390, 766394, - 766396]), - col_indices=tensor([ 0, 1, 2, ..., 116833, 89, - 116834]), - values=tensor([-1.0000e+00, -1.0000e+00, -1.0000e+00, ..., - 1.0331e+01, -1.0000e-03, 1.0000e-03]), - size=(116835, 116835), nnz=766396, layout=torch.sparse_csr) -tensor([0.8850, 0.9552, 0.7029, ..., 0.3357, 0.0248, 0.5395]) -Matrix: dc2 -Shape: torch.Size([116835, 116835]) -NNZ: 766396 -Density: 5.614451099680581e-05 -Time: 13.873224973678589 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/dc2.mtx 1000': - - 561,480 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 578,369 LL_CACHE_RD:u - 192,306 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 25,364 L2D_TLB_REFILL:u - 317,121 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,812,330 L2D_CACHE:u - - 17.467787426 seconds time elapsed - - 92.463054000 seconds user - 1072.584062000 seconds sys - - - diff --git a/pytorch/output/altra_10_30_de2010_1000.json b/pytorch/output/altra_10_30_de2010_1000.json deleted file mode 100644 index a648acc..0000000 --- a/pytorch/output/altra_10_30_de2010_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [28.56, 28.04, 23.8, 23.08, 22.12, 21.16, 21.16, 21.0, 20.96, 20.72], "matrix": "de2010", "shape": [24115, 24115], "nnz": 116056, "% density": 0.0001995689928120616, "time_s": 2.713265895843506, "power": [33.24, 30.84, 29.96, 27.68, 25.8, 25.8, 31.16], "power_after": [20.6, 20.48, 20.24, 20.32, 20.2, 20.36, 20.4, 20.4, 20.36, 20.36], "task clock (msec)": 48.96, "page faults": 3285, "cycles": 48563060, "instructions": 73465190, "branch mispredictions": 326361, "branches": 19599354, "ITLB accesses": 26666488, "ITLB misses": 6643, "DTLB misses": 17347, "DTLB accesses": 35986736, "L1I cache accesses": 32502068, "L1I cache misses": 302739, "L1D cache misses": 480619, "L1D cache accesses": 34031072, "LL cache misses": 552815, "LL cache accesses": 567373, "L2D TLB accesses": 188248, "L2D TLB misses": 23165, "L2D cache misses": 308211, "L2D cache accesses": 1787647, "instructions per cycle": 1.5127792606149613, "branch miss rate": 0.016651620252381788, "ITLB miss rate": 0.0002491141690649327, "DTLB miss rate": 0.0004820387155978803, "L2D TLB miss rate": 0.12305575623645404, "L1I cache miss rate": 0.00931445346800702, "L1D cache miss rate": 0.014122946229845479, "L2D cache miss rate": 0.17241155552522394, "LL cache miss rate": 0.9743413944618443} diff --git a/pytorch/output/altra_10_30_de2010_1000.output b/pytorch/output/altra_10_30_de2010_1000.output deleted file mode 100644 index 7b8a0c5..0000000 --- a/pytorch/output/altra_10_30_de2010_1000.output +++ /dev/null @@ -1,168 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3394985 queued and waiting for resources -srun: job 3394985 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 13, 21, ..., 116047, 116051, - 116056]), - col_indices=tensor([ 250, 251, 757, ..., 23334, 23553, 24050]), - values=tensor([ 14900., 33341., 20255., ..., 164227., 52413., - 16949.]), size=(24115, 24115), nnz=116056, - layout=torch.sparse_csr) -tensor([0.6055, 0.8789, 0.0482, ..., 0.0736, 0.1316, 0.6744]) -Matrix: de2010 -Shape: torch.Size([24115, 24115]) -NNZ: 116056 -Density: 0.0001995689928120616 -Time: 2.6956887245178223 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/de2010.mtx 1000': - - 48.96 msec task-clock:u # 0.008 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,285 page-faults:u # 67.090 K/sec - 48,563,060 cycles:u # 0.992 GHz (59.76%) - 73,465,190 instructions:u # 1.51 insn per cycle (78.23%) - branches:u - 369,314 branch-misses:u (98.16%) - 31,769,641 L1-dcache-loads:u # 648.836 M/sec - 479,594 L1-dcache-load-misses:u # 1.51% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 30,338,929 L1-icache-loads:u # 619.616 M/sec - 282,162 L1-icache-load-misses:u # 0.93% of all L1-icache accesses - 55,516,925 dTLB-loads:u # 1.134 G/sec (23.54%) - 12,345 dTLB-load-misses:u # 0.02% of all dTLB cache accesses (3.47%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 6.017085179 seconds time elapsed - - 17.484355000 seconds user - 28.678064000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 13, 21, ..., 116047, 116051, - 116056]), - col_indices=tensor([ 250, 251, 757, ..., 23334, 23553, 24050]), - values=tensor([ 14900., 33341., 20255., ..., 164227., 52413., - 16949.]), size=(24115, 24115), nnz=116056, - layout=torch.sparse_csr) -tensor([0.2815, 0.8196, 0.3706, ..., 0.1328, 0.4062, 0.9113]) -Matrix: de2010 -Shape: torch.Size([24115, 24115]) -NNZ: 116056 -Density: 0.0001995689928120616 -Time: 2.7908551692962646 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/de2010.mtx 1000': - - 326,361 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 19,599,354 BR_RETIRED:u - - 6.215591535 seconds time elapsed - - 18.097112000 seconds user - 27.831633000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 13, 21, ..., 116047, 116051, - 116056]), - col_indices=tensor([ 250, 251, 757, ..., 23334, 23553, 24050]), - values=tensor([ 14900., 33341., 20255., ..., 164227., 52413., - 16949.]), size=(24115, 24115), nnz=116056, - layout=torch.sparse_csr) -tensor([0.9002, 0.0843, 0.5558, ..., 0.3931, 0.8070, 0.7414]) -Matrix: de2010 -Shape: torch.Size([24115, 24115]) -NNZ: 116056 -Density: 0.0001995689928120616 -Time: 2.819589376449585 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/de2010.mtx 1000': - - 26,666,488 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 6,643 ITLB_WALK:u - 17,347 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 35,986,736 L1D_TLB:u - - 6.243883495 seconds time elapsed - - 17.783312000 seconds user - 31.714619000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 13, 21, ..., 116047, 116051, - 116056]), - col_indices=tensor([ 250, 251, 757, ..., 23334, 23553, 24050]), - values=tensor([ 14900., 33341., 20255., ..., 164227., 52413., - 16949.]), size=(24115, 24115), nnz=116056, - layout=torch.sparse_csr) -tensor([0.9109, 0.6392, 0.7899, ..., 0.0945, 0.3298, 0.6865]) -Matrix: de2010 -Shape: torch.Size([24115, 24115]) -NNZ: 116056 -Density: 0.0001995689928120616 -Time: 2.747800827026367 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/de2010.mtx 1000': - - 32,502,068 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 302,739 L1I_CACHE_REFILL:u - 480,619 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 34,031,072 L1D_CACHE:u - - 6.126767063 seconds time elapsed - - 17.702029000 seconds user - 29.137072000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 13, 21, ..., 116047, 116051, - 116056]), - col_indices=tensor([ 250, 251, 757, ..., 23334, 23553, 24050]), - values=tensor([ 14900., 33341., 20255., ..., 164227., 52413., - 16949.]), size=(24115, 24115), nnz=116056, - layout=torch.sparse_csr) -tensor([0.7083, 0.6766, 0.7649, ..., 0.3027, 0.9885, 0.8086]) -Matrix: de2010 -Shape: torch.Size([24115, 24115]) -NNZ: 116056 -Density: 0.0001995689928120616 -Time: 2.795116901397705 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/de2010.mtx 1000': - - 552,815 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 567,373 LL_CACHE_RD:u - 188,248 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 23,165 L2D_TLB_REFILL:u - 308,211 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,787,647 L2D_CACHE:u - - 6.041792624 seconds time elapsed - - 17.791735000 seconds user - 29.790006000 seconds sys - - - diff --git a/pytorch/output/altra_10_30_email-Enron_1000.json b/pytorch/output/altra_10_30_email-Enron_1000.json deleted file mode 100644 index a11e02c..0000000 --- a/pytorch/output/altra_10_30_email-Enron_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [28.96, 27.92, 27.24, 23.0, 22.28, 22.28, 21.6, 20.8, 20.68, 20.76], "matrix": "email-Enron", "shape": [36692, 36692], "nnz": 367662, "% density": 0.0002730901120626302, "time_s": 12.818164587020874, "power": [84.24, 82.72, 82.72, 72.0, 60.2, 51.88, 52.4, 59.36, 72.08, 83.88, 86.48, 84.28, 82.28, 81.12, 80.96, 80.96, 81.16], "power_after": [20.92, 20.92, 20.92, 20.92, 21.0, 20.96, 20.88, 20.84, 20.88, 20.68], "task clock (msec)": 48.76, "page faults": 3281, "cycles": 45495589, "instructions": 79104832, "branch mispredictions": 335574, "branches": 20121415, "ITLB accesses": 26011880, "ITLB misses": 5842, "DTLB misses": 16448, "DTLB accesses": 35000292, "L1I cache accesses": 32193112, "L1I cache misses": 310304, "L1D cache misses": 495806, "L1D cache accesses": 33829187, "LL cache misses": 546628, "LL cache accesses": 570044, "L2D TLB accesses": 196794, "L2D TLB misses": 24071, "L2D cache misses": 316028, "L2D cache accesses": 1836018, "instructions per cycle": 1.7387362981496954, "branch miss rate": 0.016677455338006797, "ITLB miss rate": 0.00022458968748125855, "DTLB miss rate": 0.000469938936509444, "L2D TLB miss rate": 0.1223157210077543, "L1I cache miss rate": 0.009638832058236556, "L1D cache miss rate": 0.014656160669779029, "L2D cache miss rate": 0.1721268527868463, "LL cache miss rate": 0.9589224691427328} diff --git a/pytorch/output/altra_10_30_email-Enron_1000.output b/pytorch/output/altra_10_30_email-Enron_1000.output deleted file mode 100644 index b10868f..0000000 --- a/pytorch/output/altra_10_30_email-Enron_1000.output +++ /dev/null @@ -1,163 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3394986 queued and waiting for resources -srun: job 3394986 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 1, 71, ..., 367660, 367661, - 367662]), - col_indices=tensor([ 1, 0, 2, ..., 36690, 36689, 8203]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(36692, 36692), - nnz=367662, layout=torch.sparse_csr) -tensor([0.9906, 0.9401, 0.5661, ..., 0.4491, 0.7550, 0.2452]) -Matrix: email-Enron -Shape: torch.Size([36692, 36692]) -NNZ: 367662 -Density: 0.0002730901120626302 -Time: 12.80848503112793 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/email-Enron.mtx 1000': - - 48.76 msec task-clock:u # 0.003 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,281 page-faults:u # 67.289 K/sec - 45,495,589 cycles:u # 0.933 GHz (57.79%) - 79,104,832 instructions:u # 1.74 insn per cycle (81.70%) - branches:u - 372,161 branch-misses:u - 32,089,348 L1-dcache-loads:u # 658.113 M/sec - 467,576 L1-dcache-load-misses:u # 1.46% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 30,688,995 L1-icache-loads:u # 629.393 M/sec - 289,698 L1-icache-load-misses:u # 0.94% of all L1-icache accesses - 47,006,355 dTLB-loads:u # 964.042 M/sec (22.12%) - dTLB-load-misses:u (0.00%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 16.331438990 seconds time elapsed - - 76.869141000 seconds user - 999.179638000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 1, 71, ..., 367660, 367661, - 367662]), - col_indices=tensor([ 1, 0, 2, ..., 36690, 36689, 8203]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(36692, 36692), - nnz=367662, layout=torch.sparse_csr) -tensor([0.7565, 0.5273, 0.1038, ..., 0.9432, 0.1309, 0.5542]) -Matrix: email-Enron -Shape: torch.Size([36692, 36692]) -NNZ: 367662 -Density: 0.0002730901120626302 -Time: 26.91536283493042 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/email-Enron.mtx 1000': - - 335,574 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 20,121,415 BR_RETIRED:u - - 30.559245388 seconds time elapsed - - 126.799314000 seconds user - 2081.777635000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 1, 71, ..., 367660, 367661, - 367662]), - col_indices=tensor([ 1, 0, 2, ..., 36690, 36689, 8203]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(36692, 36692), - nnz=367662, layout=torch.sparse_csr) -tensor([0.2321, 0.0702, 0.2538, ..., 0.6254, 0.6308, 0.5317]) -Matrix: email-Enron -Shape: torch.Size([36692, 36692]) -NNZ: 367662 -Density: 0.0002730901120626302 -Time: 14.841739892959595 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/email-Enron.mtx 1000': - - 26,011,880 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 5,842 ITLB_WALK:u - 16,448 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 35,000,292 L1D_TLB:u - - 18.443612527 seconds time elapsed - - 80.694133000 seconds user - 1159.740575000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 1, 71, ..., 367660, 367661, - 367662]), - col_indices=tensor([ 1, 0, 2, ..., 36690, 36689, 8203]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(36692, 36692), - nnz=367662, layout=torch.sparse_csr) -tensor([0.7091, 0.9447, 0.0959, ..., 0.0090, 0.7012, 0.6025]) -Matrix: email-Enron -Shape: torch.Size([36692, 36692]) -NNZ: 367662 -Density: 0.0002730901120626302 -Time: 10.863199234008789 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/email-Enron.mtx 1000': - - 32,193,112 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 310,304 L1I_CACHE_REFILL:u - 495,806 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 33,829,187 L1D_CACHE:u - - 14.426841778 seconds time elapsed - - 70.728541000 seconds user - 853.184507000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 1, 71, ..., 367660, 367661, - 367662]), - col_indices=tensor([ 1, 0, 2, ..., 36690, 36689, 8203]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(36692, 36692), - nnz=367662, layout=torch.sparse_csr) -tensor([0.8267, 0.6185, 0.8015, ..., 0.8593, 0.4881, 0.8599]) -Matrix: email-Enron -Shape: torch.Size([36692, 36692]) -NNZ: 367662 -Density: 0.0002730901120626302 -Time: 12.076026678085327 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/email-Enron.mtx 1000': - - 546,628 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 570,044 LL_CACHE_RD:u - 196,794 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 24,071 L2D_TLB_REFILL:u - 316,028 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,836,018 L2D_CACHE:u - - 15.581045199 seconds time elapsed - - 77.345591000 seconds user - 942.987439000 seconds sys - - - diff --git a/pytorch/output/altra_10_30_p2p-Gnutella04_1000.json b/pytorch/output/altra_10_30_p2p-Gnutella04_1000.json deleted file mode 100644 index b6700fa..0000000 --- a/pytorch/output/altra_10_30_p2p-Gnutella04_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [16.12, 16.36, 16.8, 16.76, 16.6, 16.48, 16.44, 16.28, 16.28, 16.16], "matrix": "p2p-Gnutella04", "shape": [10879, 10879], "nnz": 39994, "% density": 0.0003379223282393842, "time_s": 1.0642461776733398, "power": [26.6, 27.52, 27.52, 31.16, 28.48], "power_after": [16.28, 16.4, 16.32, 16.12, 16.24, 16.0, 16.0, 16.24, 16.52, 17.04], "task clock (msec)": 50.59, "page faults": 3303, "cycles": 51318459, "instructions": 74705078, "branch mispredictions": 328853, "branches": 19620312, "ITLB accesses": 27939682, "ITLB misses": 5470, "DTLB misses": 17679, "DTLB accesses": 37425602, "L1I cache accesses": 30276633, "L1I cache misses": 291467, "L1D cache misses": 479061, "L1D cache accesses": 31689326, "LL cache misses": 529426, "LL cache accesses": 550033, "L2D TLB accesses": 171913, "L2D TLB misses": 20624, "L2D cache misses": 296662, "L2D cache accesses": 1714211, "instructions per cycle": 1.455715535028049, "branch miss rate": 0.01676084457780284, "ITLB miss rate": 0.0001957788925443031, "DTLB miss rate": 0.00047237717111404113, "L2D TLB miss rate": 0.11996765805959991, "L1I cache miss rate": 0.009626797008769106, "L1D cache miss rate": 0.015117424712661923, "L2D cache miss rate": 0.17306037588138215, "LL cache miss rate": 0.9625349751742168} diff --git a/pytorch/output/altra_10_30_p2p-Gnutella04_1000.output b/pytorch/output/altra_10_30_p2p-Gnutella04_1000.output deleted file mode 100644 index f653bb6..0000000 --- a/pytorch/output/altra_10_30_p2p-Gnutella04_1000.output +++ /dev/null @@ -1,158 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3394992 queued and waiting for resources -srun: job 3394992 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 10, 20, ..., 39994, 39994, 39994]), - col_indices=tensor([ 1, 2, 3, ..., 9711, 10875, 10876]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(10879, 10879), - nnz=39994, layout=torch.sparse_csr) -tensor([0.1181, 0.8387, 0.0554, ..., 0.8107, 0.4393, 0.9489]) -Matrix: p2p-Gnutella04 -Shape: torch.Size([10879, 10879]) -NNZ: 39994 -Density: 0.0003379223282393842 -Time: 1.061662197113037 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella04.mtx 1000': - - 50.59 msec task-clock:u # 0.012 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,303 page-faults:u # 65.291 K/sec - 51,318,459 cycles:u # 1.014 GHz (59.34%) - 74,705,078 instructions:u # 1.46 insn per cycle (83.02%) - branches:u - 366,825 branch-misses:u - 31,809,194 L1-dcache-loads:u # 628.781 M/sec - 466,198 L1-dcache-load-misses:u # 1.47% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 30,390,161 L1-icache-loads:u # 600.731 M/sec - 296,270 L1-icache-load-misses:u # 0.97% of all L1-icache accesses - 61,518,375 dTLB-loads:u # 1.216 G/sec (17.94%) - dTLB-load-misses:u (0.00%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 4.302241563 seconds time elapsed - - 16.122298000 seconds user - 29.141140000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 10, 20, ..., 39994, 39994, 39994]), - col_indices=tensor([ 1, 2, 3, ..., 9711, 10875, 10876]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(10879, 10879), - nnz=39994, layout=torch.sparse_csr) -tensor([0.7249, 0.8723, 0.3843, ..., 0.2264, 0.4891, 0.9107]) -Matrix: p2p-Gnutella04 -Shape: torch.Size([10879, 10879]) -NNZ: 39994 -Density: 0.0003379223282393842 -Time: 1.0079431533813477 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella04.mtx 1000': - - 328,853 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 19,620,312 BR_RETIRED:u - - 4.241400567 seconds time elapsed - - 15.325937000 seconds user - 28.223386000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 10, 20, ..., 39994, 39994, 39994]), - col_indices=tensor([ 1, 2, 3, ..., 9711, 10875, 10876]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(10879, 10879), - nnz=39994, layout=torch.sparse_csr) -tensor([0.7608, 0.2449, 0.5322, ..., 0.5547, 0.8659, 0.8437]) -Matrix: p2p-Gnutella04 -Shape: torch.Size([10879, 10879]) -NNZ: 39994 -Density: 0.0003379223282393842 -Time: 1.1017234325408936 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella04.mtx 1000': - - 27,939,682 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 5,470 ITLB_WALK:u - 17,679 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 37,425,602 L1D_TLB:u - - 4.296820500 seconds time elapsed - - 15.875162000 seconds user - 28.803412000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 10, 20, ..., 39994, 39994, 39994]), - col_indices=tensor([ 1, 2, 3, ..., 9711, 10875, 10876]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(10879, 10879), - nnz=39994, layout=torch.sparse_csr) -tensor([0.9980, 0.9991, 0.6749, ..., 0.4225, 0.7297, 0.3717]) -Matrix: p2p-Gnutella04 -Shape: torch.Size([10879, 10879]) -NNZ: 39994 -Density: 0.0003379223282393842 -Time: 1.0812580585479736 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella04.mtx 1000': - - 30,276,633 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 291,467 L1I_CACHE_REFILL:u - 479,061 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 31,689,326 L1D_CACHE:u - - 4.500137840 seconds time elapsed - - 15.794710000 seconds user - 27.773851000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 10, 20, ..., 39994, 39994, 39994]), - col_indices=tensor([ 1, 2, 3, ..., 9711, 10875, 10876]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(10879, 10879), - nnz=39994, layout=torch.sparse_csr) -tensor([0.8707, 0.5871, 0.5970, ..., 0.8826, 0.4673, 0.4994]) -Matrix: p2p-Gnutella04 -Shape: torch.Size([10879, 10879]) -NNZ: 39994 -Density: 0.0003379223282393842 -Time: 0.9900743961334229 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella04.mtx 1000': - - 529,426 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 550,033 LL_CACHE_RD:u - 171,913 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 20,624 L2D_TLB_REFILL:u - 296,662 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,714,211 L2D_CACHE:u - - 4.284402033 seconds time elapsed - - 15.584671000 seconds user - 27.523772000 seconds sys - - - diff --git a/pytorch/output/altra_10_30_p2p-Gnutella24_1000.json b/pytorch/output/altra_10_30_p2p-Gnutella24_1000.json deleted file mode 100644 index 09c3c83..0000000 --- a/pytorch/output/altra_10_30_p2p-Gnutella24_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [16.12, 16.12, 16.12, 16.36, 16.56, 16.52, 17.04, 16.76, 16.64, 16.92], "matrix": "p2p-Gnutella24", "shape": [26518, 26518], "nnz": 65369, "% density": 9.295875717624285e-05, "time_s": 1.6947758197784424, "power": [25.2, 25.2, 26.6, 26.28, 26.48], "power_after": [16.4, 16.6, 16.6, 16.64, 16.8, 16.48, 16.44, 16.16, 16.12, 16.2], "task clock (msec)": 66.78, "page faults": 3520, "cycles": 28858055, "instructions": 64429843, "branch mispredictions": 331167, "branches": 19518210, "ITLB accesses": 26964483, "ITLB misses": 4666, "DTLB misses": 14001, "DTLB accesses": 36143905, "L1I cache accesses": 31901160, "L1I cache misses": 302516, "L1D cache misses": 475663, "L1D cache accesses": 33507563, "LL cache misses": 558546, "LL cache accesses": 578676, "L2D TLB accesses": 187549, "L2D TLB misses": 22990, "L2D cache misses": 321826, "L2D cache accesses": 1816571, "instructions per cycle": 2.2326467601506756, "branch miss rate": 0.016967078435983628, "ITLB miss rate": 0.00017304244253449992, "DTLB miss rate": 0.00038736821602425086, "L2D TLB miss rate": 0.12258129875392564, "L1I cache miss rate": 0.009482915354802146, "L1D cache miss rate": 0.01419569068630864, "L2D cache miss rate": 0.1771612560147663, "LL cache miss rate": 0.9652136947099932} diff --git a/pytorch/output/altra_10_30_p2p-Gnutella24_1000.output b/pytorch/output/altra_10_30_p2p-Gnutella24_1000.output deleted file mode 100644 index 838ff15..0000000 --- a/pytorch/output/altra_10_30_p2p-Gnutella24_1000.output +++ /dev/null @@ -1,158 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3394989 queued and waiting for resources -srun: job 3394989 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 9, 9, ..., 65369, 65369, 65369]), - col_indices=tensor([ 1, 2, 3, ..., 15065, 9401, 26517]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(26518, 26518), - nnz=65369, layout=torch.sparse_csr) -tensor([0.2470, 0.4231, 0.1036, ..., 0.7937, 0.3241, 0.7116]) -Matrix: p2p-Gnutella24 -Shape: torch.Size([26518, 26518]) -NNZ: 65369 -Density: 9.295875717624285e-05 -Time: 1.6974337100982666 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella24.mtx 1000': - - 66.78 msec task-clock:u # 0.013 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,520 page-faults:u # 52.713 K/sec - 28,858,055 cycles:u # 0.432 GHz (26.93%) - 64,429,843 instructions:u # 2.23 insn per cycle (67.63%) - branches:u - 296,857 branch-misses:u (84.08%) - 33,646,348 L1-dcache-loads:u # 503.866 M/sec - 493,998 L1-dcache-load-misses:u # 1.47% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 32,070,415 L1-icache-loads:u # 480.266 M/sec - 305,993 L1-icache-load-misses:u # 0.95% of all L1-icache accesses - 46,903,081 dTLB-loads:u # 702.391 M/sec (46.16%) - 114,272 dTLB-load-misses:u # 0.24% of all dTLB cache accesses (32.45%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 5.106933083 seconds time elapsed - - 16.391614000 seconds user - 28.913912000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 9, 9, ..., 65369, 65369, 65369]), - col_indices=tensor([ 1, 2, 3, ..., 15065, 9401, 26517]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(26518, 26518), - nnz=65369, layout=torch.sparse_csr) -tensor([0.2307, 0.4662, 0.3789, ..., 0.0144, 0.6300, 0.7829]) -Matrix: p2p-Gnutella24 -Shape: torch.Size([26518, 26518]) -NNZ: 65369 -Density: 9.295875717624285e-05 -Time: 1.6379659175872803 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella24.mtx 1000': - - 331,167 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 19,518,210 BR_RETIRED:u - - 5.017894585 seconds time elapsed - - 16.446505000 seconds user - 31.004338000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 9, 9, ..., 65369, 65369, 65369]), - col_indices=tensor([ 1, 2, 3, ..., 15065, 9401, 26517]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(26518, 26518), - nnz=65369, layout=torch.sparse_csr) -tensor([0.7309, 0.0314, 0.4424, ..., 0.7434, 0.2124, 0.1432]) -Matrix: p2p-Gnutella24 -Shape: torch.Size([26518, 26518]) -NNZ: 65369 -Density: 9.295875717624285e-05 -Time: 1.7232718467712402 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella24.mtx 1000': - - 26,964,483 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 4,666 ITLB_WALK:u - 14,001 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 36,143,905 L1D_TLB:u - - 5.053286721 seconds time elapsed - - 16.447780000 seconds user - 28.580949000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 9, 9, ..., 65369, 65369, 65369]), - col_indices=tensor([ 1, 2, 3, ..., 15065, 9401, 26517]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(26518, 26518), - nnz=65369, layout=torch.sparse_csr) -tensor([0.5695, 0.5025, 0.1946, ..., 0.7428, 0.9634, 0.4327]) -Matrix: p2p-Gnutella24 -Shape: torch.Size([26518, 26518]) -NNZ: 65369 -Density: 9.295875717624285e-05 -Time: 1.644775629043579 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella24.mtx 1000': - - 31,901,160 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 302,516 L1I_CACHE_REFILL:u - 475,663 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 33,507,563 L1D_CACHE:u - - 4.978338941 seconds time elapsed - - 16.455298000 seconds user - 30.249373000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 9, 9, ..., 65369, 65369, 65369]), - col_indices=tensor([ 1, 2, 3, ..., 15065, 9401, 26517]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(26518, 26518), - nnz=65369, layout=torch.sparse_csr) -tensor([0.0969, 0.1950, 0.8456, ..., 0.3315, 0.1512, 0.3182]) -Matrix: p2p-Gnutella24 -Shape: torch.Size([26518, 26518]) -NNZ: 65369 -Density: 9.295875717624285e-05 -Time: 1.752812385559082 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella24.mtx 1000': - - 558,546 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 578,676 LL_CACHE_RD:u - 187,549 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 22,990 L2D_TLB_REFILL:u - 321,826 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,816,571 L2D_CACHE:u - - 4.952297819 seconds time elapsed - - 16.648691000 seconds user - 27.005944000 seconds sys - - - diff --git a/pytorch/output/altra_10_30_p2p-Gnutella25_1000.json b/pytorch/output/altra_10_30_p2p-Gnutella25_1000.json deleted file mode 100644 index 446cb18..0000000 --- a/pytorch/output/altra_10_30_p2p-Gnutella25_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [16.0, 16.4, 16.4, 16.28, 16.48, 16.6, 16.48, 16.56, 16.88, 16.92], "matrix": "p2p-Gnutella25", "shape": [22687, 22687], "nnz": 54705, "% density": 0.00010628522108964806, "time_s": 1.4688231945037842, "power": [23.04, 29.0, 30.24, 27.96, 28.04], "power_after": [16.52, 16.68, 16.88, 17.12, 17.08, 17.04, 16.84, 16.72, 16.84, 16.84], "task clock (msec)": 48.61, "page faults": 3308, "cycles": 60072179, "instructions": 70991785, "branch mispredictions": 331765, "branches": 19906014, "ITLB accesses": 28194337, "ITLB misses": 5083, "DTLB misses": 17916, "DTLB accesses": 37944713, "L1I cache accesses": 31162212, "L1I cache misses": 270684, "L1D cache misses": 465467, "L1D cache accesses": 32857500, "LL cache misses": 541118, "LL cache accesses": 564199, "L2D TLB accesses": 194022, "L2D TLB misses": 23932, "L2D cache misses": 311476, "L2D cache accesses": 1783574, "instructions per cycle": 1.1817747613250387, "branch miss rate": 0.016666571218125335, "ITLB miss rate": 0.00018028443087702328, "DTLB miss rate": 0.00047216064066685654, "L2D TLB miss rate": 0.12334683695663379, "L1I cache miss rate": 0.008686289663904475, "L1D cache miss rate": 0.014166232975728525, "L2D cache miss rate": 0.17463587157022922, "LL cache miss rate": 0.9590906754531646} diff --git a/pytorch/output/altra_10_30_p2p-Gnutella25_1000.output b/pytorch/output/altra_10_30_p2p-Gnutella25_1000.output deleted file mode 100644 index 1026977..0000000 --- a/pytorch/output/altra_10_30_p2p-Gnutella25_1000.output +++ /dev/null @@ -1,158 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3394994 queued and waiting for resources -srun: job 3394994 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 9, 9, ..., 54704, 54704, 54705]), - col_indices=tensor([ 1, 2, 3, ..., 17949, 22685, 144]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(22687, 22687), - nnz=54705, layout=torch.sparse_csr) -tensor([0.1465, 0.4354, 0.7334, ..., 0.2837, 0.5913, 0.9525]) -Matrix: p2p-Gnutella25 -Shape: torch.Size([22687, 22687]) -NNZ: 54705 -Density: 0.00010628522108964806 -Time: 1.4786670207977295 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella25.mtx 1000': - - 48.61 msec task-clock:u # 0.010 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,308 page-faults:u # 68.054 K/sec - 60,072,179 cycles:u # 1.236 GHz (53.26%) - 70,991,785 instructions:u # 1.18 insn per cycle (71.54%) - branches:u - 371,197 branch-misses:u - 32,964,378 L1-dcache-loads:u # 678.165 M/sec - 465,448 L1-dcache-load-misses:u # 1.41% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 31,435,424 L1-icache-loads:u # 646.710 M/sec - 293,561 L1-icache-load-misses:u # 0.93% of all L1-icache accesses - 56,761,270 dTLB-loads:u # 1.168 G/sec (30.54%) - dTLB-load-misses:u (0.00%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 4.700046411 seconds time elapsed - - 16.235801000 seconds user - 28.396327000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 9, 9, ..., 54704, 54704, 54705]), - col_indices=tensor([ 1, 2, 3, ..., 17949, 22685, 144]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(22687, 22687), - nnz=54705, layout=torch.sparse_csr) -tensor([0.7780, 0.3388, 0.1540, ..., 0.2989, 0.3682, 0.9160]) -Matrix: p2p-Gnutella25 -Shape: torch.Size([22687, 22687]) -NNZ: 54705 -Density: 0.00010628522108964806 -Time: 1.4235138893127441 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella25.mtx 1000': - - 331,765 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 19,906,014 BR_RETIRED:u - - 4.757340585 seconds time elapsed - - 16.412311000 seconds user - 29.238029000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 9, 9, ..., 54704, 54704, 54705]), - col_indices=tensor([ 1, 2, 3, ..., 17949, 22685, 144]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(22687, 22687), - nnz=54705, layout=torch.sparse_csr) -tensor([0.4944, 0.8057, 0.8211, ..., 0.5137, 0.3388, 0.6316]) -Matrix: p2p-Gnutella25 -Shape: torch.Size([22687, 22687]) -NNZ: 54705 -Density: 0.00010628522108964806 -Time: 1.4664146900177002 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella25.mtx 1000': - - 28,194,337 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 5,083 ITLB_WALK:u - 17,916 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 37,944,713 L1D_TLB:u - - 4.844329421 seconds time elapsed - - 16.081022000 seconds user - 28.021902000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 9, 9, ..., 54704, 54704, 54705]), - col_indices=tensor([ 1, 2, 3, ..., 17949, 22685, 144]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(22687, 22687), - nnz=54705, layout=torch.sparse_csr) -tensor([0.0963, 0.5806, 0.0397, ..., 0.1604, 0.5700, 0.8103]) -Matrix: p2p-Gnutella25 -Shape: torch.Size([22687, 22687]) -NNZ: 54705 -Density: 0.00010628522108964806 -Time: 1.3717434406280518 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella25.mtx 1000': - - 31,162,212 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 270,684 L1I_CACHE_REFILL:u - 465,467 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 32,857,500 L1D_CACHE:u - - 4.598461782 seconds time elapsed - - 15.609727000 seconds user - 30.606837000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 9, 9, ..., 54704, 54704, 54705]), - col_indices=tensor([ 1, 2, 3, ..., 17949, 22685, 144]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(22687, 22687), - nnz=54705, layout=torch.sparse_csr) -tensor([0.9137, 0.5009, 0.7507, ..., 0.6623, 0.8760, 0.2991]) -Matrix: p2p-Gnutella25 -Shape: torch.Size([22687, 22687]) -NNZ: 54705 -Density: 0.00010628522108964806 -Time: 1.4291880130767822 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella25.mtx 1000': - - 541,118 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 564,199 LL_CACHE_RD:u - 194,022 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 23,932 L2D_TLB_REFILL:u - 311,476 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,783,574 L2D_CACHE:u - - 4.792239951 seconds time elapsed - - 15.902307000 seconds user - 28.747620000 seconds sys - - - diff --git a/pytorch/output/altra_10_30_p2p-Gnutella30_1000.json b/pytorch/output/altra_10_30_p2p-Gnutella30_1000.json deleted file mode 100644 index fe2d5fc..0000000 --- a/pytorch/output/altra_10_30_p2p-Gnutella30_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [31.96, 22.0, 22.16, 22.16, 21.84, 22.08, 22.4, 22.08, 22.0, 21.48], "matrix": "p2p-Gnutella30", "shape": [36682, 36682], "nnz": 88328, "% density": 6.564359899804003e-05, "time_s": 3.504030466079712, "power": [54.2, 64.16, 67.64, 67.64, 65.92, 58.96, 59.92], "power_after": [20.72, 20.76, 20.76, 20.8, 20.8, 20.88, 20.92, 21.04, 21.04, 21.12], "task clock (msec)": 56.52, "page faults": 3194, "cycles": 58074747, "instructions": 90036443, "branch mispredictions": 327895, "branches": 20553601, "ITLB accesses": 26120611, "ITLB misses": 7531, "DTLB misses": 19097, "DTLB accesses": 35744928, "L1I cache accesses": 31819981, "L1I cache misses": 284493, "L1D cache misses": 486709, "L1D cache accesses": 33545755, "LL cache misses": 544742, "LL cache accesses": 558323, "L2D TLB accesses": 190574, "L2D TLB misses": 23746, "L2D cache misses": 305844, "L2D cache accesses": 1736964, "instructions per cycle": 1.5503544595725918, "branch miss rate": 0.015953165579111903, "ITLB miss rate": 0.00028831637973552763, "DTLB miss rate": 0.0005342576155140109, "L2D TLB miss rate": 0.12460251660772194, "L1I cache miss rate": 0.008940703012990485, "L1D cache miss rate": 0.014508810429218243, "L2D cache miss rate": 0.17607964241055082, "LL cache miss rate": 0.9756753707083534} diff --git a/pytorch/output/altra_10_30_p2p-Gnutella30_1000.output b/pytorch/output/altra_10_30_p2p-Gnutella30_1000.output deleted file mode 100644 index b8705d3..0000000 --- a/pytorch/output/altra_10_30_p2p-Gnutella30_1000.output +++ /dev/null @@ -1,158 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3394991 queued and waiting for resources -srun: job 3394991 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 10, 10, ..., 88328, 88328, 88328]), - col_indices=tensor([ 1, 2, 3, ..., 36675, 36676, 36677]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(36682, 36682), - nnz=88328, layout=torch.sparse_csr) -tensor([0.3046, 0.0725, 0.4580, ..., 0.0593, 0.5121, 0.2116]) -Matrix: p2p-Gnutella30 -Shape: torch.Size([36682, 36682]) -NNZ: 88328 -Density: 6.564359899804003e-05 -Time: 3.6646029949188232 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella30.mtx 1000': - - 56.52 msec task-clock:u # 0.008 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,194 page-faults:u # 56.515 K/sec - 58,074,747 cycles:u # 1.028 GHz (51.20%) - 90,036,443 instructions:u # 1.55 insn per cycle (89.06%) - branches:u - 363,262 branch-misses:u - 33,111,438 L1-dcache-loads:u # 585.875 M/sec - 454,665 L1-dcache-load-misses:u # 1.37% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 31,646,314 L1-icache-loads:u # 559.951 M/sec - 281,443 L1-icache-load-misses:u # 0.89% of all L1-icache accesses - 43,495,524 dTLB-loads:u # 769.611 M/sec (11.87%) - dTLB-load-misses:u (0.00%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 7.033463989 seconds time elapsed - - 34.670765000 seconds user - 307.031553000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 10, 10, ..., 88328, 88328, 88328]), - col_indices=tensor([ 1, 2, 3, ..., 36675, 36676, 36677]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(36682, 36682), - nnz=88328, layout=torch.sparse_csr) -tensor([0.9700, 0.1728, 0.2199, ..., 0.6107, 0.3357, 0.2661]) -Matrix: p2p-Gnutella30 -Shape: torch.Size([36682, 36682]) -NNZ: 88328 -Density: 6.564359899804003e-05 -Time: 2.3380045890808105 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella30.mtx 1000': - - 327,895 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 20,553,601 BR_RETIRED:u - - 5.895917276 seconds time elapsed - - 31.121063000 seconds user - 208.127447000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 10, 10, ..., 88328, 88328, 88328]), - col_indices=tensor([ 1, 2, 3, ..., 36675, 36676, 36677]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(36682, 36682), - nnz=88328, layout=torch.sparse_csr) -tensor([0.9533, 0.7568, 0.8141, ..., 0.8395, 0.5617, 0.7830]) -Matrix: p2p-Gnutella30 -Shape: torch.Size([36682, 36682]) -NNZ: 88328 -Density: 6.564359899804003e-05 -Time: 4.476518869400024 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella30.mtx 1000': - - 26,120,611 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 7,531 ITLB_WALK:u - 19,097 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 35,744,928 L1D_TLB:u - - 8.109622410 seconds time elapsed - - 38.467161000 seconds user - 370.437915000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 10, 10, ..., 88328, 88328, 88328]), - col_indices=tensor([ 1, 2, 3, ..., 36675, 36676, 36677]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(36682, 36682), - nnz=88328, layout=torch.sparse_csr) -tensor([0.6886, 0.7814, 0.9957, ..., 0.8460, 0.1015, 0.8097]) -Matrix: p2p-Gnutella30 -Shape: torch.Size([36682, 36682]) -NNZ: 88328 -Density: 6.564359899804003e-05 -Time: 2.856834888458252 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella30.mtx 1000': - - 31,819,981 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 284,493 L1I_CACHE_REFILL:u - 486,709 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 33,545,755 L1D_CACHE:u - - 6.374371632 seconds time elapsed - - 30.817943000 seconds user - 247.363843000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 10, 10, ..., 88328, 88328, 88328]), - col_indices=tensor([ 1, 2, 3, ..., 36675, 36676, 36677]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(36682, 36682), - nnz=88328, layout=torch.sparse_csr) -tensor([0.8464, 0.0437, 0.1230, ..., 0.6221, 0.9268, 0.5436]) -Matrix: p2p-Gnutella30 -Shape: torch.Size([36682, 36682]) -NNZ: 88328 -Density: 6.564359899804003e-05 -Time: 4.838747978210449 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella30.mtx 1000': - - 544,742 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 558,323 LL_CACHE_RD:u - 190,574 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 23,746 L2D_TLB_REFILL:u - 305,844 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,736,964 L2D_CACHE:u - - 8.386896120 seconds time elapsed - - 39.861141000 seconds user - 395.959334000 seconds sys - - - diff --git a/pytorch/output/altra_10_30_ri2010_1000.json b/pytorch/output/altra_10_30_ri2010_1000.json deleted file mode 100644 index dbf62d4..0000000 --- a/pytorch/output/altra_10_30_ri2010_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [31.2, 31.56, 31.56, 30.84, 24.52, 23.2, 21.32, 20.76, 20.84, 20.84], "matrix": "ri2010", "shape": [25181, 25181], "nnz": 125750, "% density": 0.00019831796057928155, "time_s": 3.077709913253784, "power": [27.76, 28.28, 28.44, 28.28, 25.16, 30.44, 30.6], "power_after": [21.08, 20.88, 20.68, 20.68, 20.6, 20.56, 20.68, 20.8, 20.96, 21.24], "task clock (msec)": 64.49, "page faults": 3473, "cycles": 42783607, "instructions": 84598454, "branch mispredictions": 331326, "branches": 20438455, "ITLB accesses": 26869742, "ITLB misses": 6302, "DTLB misses": 14926, "DTLB accesses": 36876841, "L1I cache accesses": 31664385, "L1I cache misses": 301678, "L1D cache misses": 493536, "L1D cache accesses": 33219437, "LL cache misses": 552180, "LL cache accesses": 564990, "L2D TLB accesses": 167824, "L2D TLB misses": 19594, "L2D cache misses": 304114, "L2D cache accesses": 1716370, "instructions per cycle": 1.977356747877756, "branch miss rate": 0.01621091222404042, "ITLB miss rate": 0.00023453890997539165, "DTLB miss rate": 0.00040475267390718204, "L2D TLB miss rate": 0.11675326532557918, "L1I cache miss rate": 0.009527360155581737, "L1D cache miss rate": 0.014856844202386693, "L2D cache miss rate": 0.17718440662561102, "LL cache miss rate": 0.9773270323368555} diff --git a/pytorch/output/altra_10_30_ri2010_1000.output b/pytorch/output/altra_10_30_ri2010_1000.output deleted file mode 100644 index 916d6ea..0000000 --- a/pytorch/output/altra_10_30_ri2010_1000.output +++ /dev/null @@ -1,163 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3394984 queued and waiting for resources -srun: job 3394984 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 8, ..., 125742, 125747, - 125750]), - col_indices=tensor([ 25, 56, 662, ..., 21738, 22279, 23882]), - values=tensor([17171., 37318., 5284., ..., 25993., 24918., 803.]), - size=(25181, 25181), nnz=125750, layout=torch.sparse_csr) -tensor([0.5906, 0.9651, 0.2033, ..., 0.2175, 0.4484, 0.0412]) -Matrix: ri2010 -Shape: torch.Size([25181, 25181]) -NNZ: 125750 -Density: 0.00019831796057928155 -Time: 3.107008934020996 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ri2010.mtx 1000': - - 64.49 msec task-clock:u # 0.010 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,473 page-faults:u # 53.852 K/sec - 42,783,607 cycles:u # 0.663 GHz (37.27%) - 84,598,454 instructions:u # 1.98 insn per cycle (73.53%) - branches:u - 353,558 branch-misses:u (89.57%) - 33,192,964 L1-dcache-loads:u # 514.689 M/sec - 466,217 L1-dcache-load-misses:u # 1.40% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 31,727,502 L1-icache-loads:u # 491.965 M/sec - 292,570 L1-icache-load-misses:u # 0.92% of all L1-icache accesses - 38,623,737 dTLB-loads:u # 598.898 M/sec (34.88%) - 124,174 dTLB-load-misses:u # 0.32% of all dTLB cache accesses (14.74%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 6.612563197 seconds time elapsed - - 18.114584000 seconds user - 29.808542000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 8, ..., 125742, 125747, - 125750]), - col_indices=tensor([ 25, 56, 662, ..., 21738, 22279, 23882]), - values=tensor([17171., 37318., 5284., ..., 25993., 24918., 803.]), - size=(25181, 25181), nnz=125750, layout=torch.sparse_csr) -tensor([0.6092, 0.5511, 0.6052, ..., 0.8002, 0.0295, 0.2972]) -Matrix: ri2010 -Shape: torch.Size([25181, 25181]) -NNZ: 125750 -Density: 0.00019831796057928155 -Time: 2.9385879039764404 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ri2010.mtx 1000': - - 331,326 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 20,438,455 BR_RETIRED:u - - 6.446731410 seconds time elapsed - - 17.939571000 seconds user - 33.272929000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 8, ..., 125742, 125747, - 125750]), - col_indices=tensor([ 25, 56, 662, ..., 21738, 22279, 23882]), - values=tensor([17171., 37318., 5284., ..., 25993., 24918., 803.]), - size=(25181, 25181), nnz=125750, layout=torch.sparse_csr) -tensor([0.3348, 0.2974, 0.2569, ..., 0.2397, 0.1965, 0.5651]) -Matrix: ri2010 -Shape: torch.Size([25181, 25181]) -NNZ: 125750 -Density: 0.00019831796057928155 -Time: 2.972891330718994 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ri2010.mtx 1000': - - 26,869,742 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 6,302 ITLB_WALK:u - 14,926 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 36,876,841 L1D_TLB:u - - 6.376775396 seconds time elapsed - - 17.836418000 seconds user - 29.830135000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 8, ..., 125742, 125747, - 125750]), - col_indices=tensor([ 25, 56, 662, ..., 21738, 22279, 23882]), - values=tensor([17171., 37318., 5284., ..., 25993., 24918., 803.]), - size=(25181, 25181), nnz=125750, layout=torch.sparse_csr) -tensor([0.7889, 0.7395, 0.6553, ..., 0.3938, 0.2478, 0.7923]) -Matrix: ri2010 -Shape: torch.Size([25181, 25181]) -NNZ: 125750 -Density: 0.00019831796057928155 -Time: 2.9658284187316895 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ri2010.mtx 1000': - - 31,664,385 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 301,678 L1I_CACHE_REFILL:u - 493,536 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 33,219,437 L1D_CACHE:u - - 6.559158078 seconds time elapsed - - 19.008146000 seconds user - 38.233666000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 8, ..., 125742, 125747, - 125750]), - col_indices=tensor([ 25, 56, 662, ..., 21738, 22279, 23882]), - values=tensor([17171., 37318., 5284., ..., 25993., 24918., 803.]), - size=(25181, 25181), nnz=125750, layout=torch.sparse_csr) -tensor([0.1256, 0.1417, 0.9800, ..., 0.2509, 0.8121, 0.6210]) -Matrix: ri2010 -Shape: torch.Size([25181, 25181]) -NNZ: 125750 -Density: 0.00019831796057928155 -Time: 2.9228267669677734 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ri2010.mtx 1000': - - 552,180 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 564,990 LL_CACHE_RD:u - 167,824 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 19,594 L2D_TLB_REFILL:u - 304,114 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,716,370 L2D_CACHE:u - - 6.135787277 seconds time elapsed - - 18.029630000 seconds user - 28.723217000 seconds sys - - - diff --git a/pytorch/output/altra_10_30_soc-sign-Slashdot090216_1000.json b/pytorch/output/altra_10_30_soc-sign-Slashdot090216_1000.json deleted file mode 100644 index 7612d8f..0000000 --- a/pytorch/output/altra_10_30_soc-sign-Slashdot090216_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [29.88, 23.64, 23.08, 21.84, 21.4, 21.2, 21.0, 21.0, 21.16, 21.0], "matrix": "soc-sign-Slashdot090216", "shape": [81871, 81871], "nnz": 545671, "% density": 8.140867447881048e-05, "time_s": 19.113287687301636, "power": [81.08, 81.56, 71.96, 60.52, 47.16, 48.4, 53.84, 53.84, 67.4, 82.64, 90.8, 89.16, 87.96, 85.76, 84.64, 84.04, 83.64, 84.68, 84.88, 84.88, 84.64, 84.04, 83.6], "power_after": [20.72, 20.6, 20.68, 20.88, 21.2, 21.28, 21.28, 21.48, 21.56, 21.36], "task clock (msec)": 67.66, "page faults": 3317, "cycles": 41915850, "instructions": 84471787, "branch mispredictions": 344452, "branches": 20610765, "ITLB accesses": 27276117, "ITLB misses": 6358, "DTLB misses": 17361, "DTLB accesses": 36565837, "L1I cache accesses": 32022662, "L1I cache misses": 293044, "L1D cache misses": 458939, "L1D cache accesses": 33505164, "LL cache misses": 553814, "LL cache accesses": 567372, "L2D TLB accesses": 199301, "L2D TLB misses": 25193, "L2D cache misses": 313278, "L2D cache accesses": 1796299, "instructions per cycle": 2.015270762730566, "branch miss rate": 0.016712237512775483, "ITLB miss rate": 0.00023309769495416082, "DTLB miss rate": 0.0004747874361524939, "L2D TLB miss rate": 0.12640679173712124, "L1I cache miss rate": 0.009151144274014446, "L1D cache miss rate": 0.01369756017311242, "L2D cache miss rate": 0.17440192306514674, "LL cache miss rate": 0.97610386131145} diff --git a/pytorch/output/altra_10_30_soc-sign-Slashdot090216_1000.output b/pytorch/output/altra_10_30_soc-sign-Slashdot090216_1000.output deleted file mode 100644 index fdda743..0000000 --- a/pytorch/output/altra_10_30_soc-sign-Slashdot090216_1000.output +++ /dev/null @@ -1,163 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3394981 queued and waiting for resources -srun: job 3394981 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 29, 124, ..., 545669, 545669, - 545671]), - col_indices=tensor([ 1, 2, 3, ..., 81869, 81699, 81863]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(81871, 81871), - nnz=545671, layout=torch.sparse_csr) -tensor([0.6780, 0.5234, 0.1205, ..., 0.2995, 0.6275, 0.1399]) -Matrix: soc-sign-Slashdot090216 -Shape: torch.Size([81871, 81871]) -NNZ: 545671 -Density: 8.140867447881048e-05 -Time: 30.653191089630127 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/soc-sign-Slashdot090216.mtx 1000': - - 67.66 msec task-clock:u # 0.002 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,317 page-faults:u # 49.022 K/sec - 41,915,850 cycles:u # 0.619 GHz (57.88%) - 84,471,787 instructions:u # 2.02 insn per cycle (88.19%) - branches:u - 375,016 branch-misses:u - 32,438,527 L1-dcache-loads:u # 479.407 M/sec - 499,618 L1-dcache-load-misses:u # 1.54% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 30,998,693 L1-icache-loads:u # 458.127 M/sec - 306,445 L1-icache-load-misses:u # 0.99% of all L1-icache accesses - 34,294,934 dTLB-loads:u # 506.842 M/sec (18.86%) - dTLB-load-misses:u (0.00%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 34.340632995 seconds time elapsed - - 149.743244000 seconds user - 2355.852109000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 29, 124, ..., 545669, 545669, - 545671]), - col_indices=tensor([ 1, 2, 3, ..., 81869, 81699, 81863]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(81871, 81871), - nnz=545671, layout=torch.sparse_csr) -tensor([0.9875, 0.2031, 0.7260, ..., 0.5908, 0.1575, 0.7971]) -Matrix: soc-sign-Slashdot090216 -Shape: torch.Size([81871, 81871]) -NNZ: 545671 -Density: 8.140867447881048e-05 -Time: 13.671181440353394 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/soc-sign-Slashdot090216.mtx 1000': - - 344,452 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 20,610,765 BR_RETIRED:u - - 17.331425967 seconds time elapsed - - 83.136180000 seconds user - 1069.027469000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 29, 124, ..., 545669, 545669, - 545671]), - col_indices=tensor([ 1, 2, 3, ..., 81869, 81699, 81863]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(81871, 81871), - nnz=545671, layout=torch.sparse_csr) -tensor([0.2046, 0.3645, 0.7960, ..., 0.6490, 0.4098, 0.5342]) -Matrix: soc-sign-Slashdot090216 -Shape: torch.Size([81871, 81871]) -NNZ: 545671 -Density: 8.140867447881048e-05 -Time: 19.569235801696777 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/soc-sign-Slashdot090216.mtx 1000': - - 27,276,117 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 6,358 ITLB_WALK:u - 17,361 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 36,565,837 L1D_TLB:u - - 23.323243037 seconds time elapsed - - 108.830923000 seconds user - 1521.834565000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 29, 124, ..., 545669, 545669, - 545671]), - col_indices=tensor([ 1, 2, 3, ..., 81869, 81699, 81863]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(81871, 81871), - nnz=545671, layout=torch.sparse_csr) -tensor([0.4164, 0.2188, 0.5460, ..., 0.1057, 0.5277, 0.0624]) -Matrix: soc-sign-Slashdot090216 -Shape: torch.Size([81871, 81871]) -NNZ: 545671 -Density: 8.140867447881048e-05 -Time: 26.337355375289917 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/soc-sign-Slashdot090216.mtx 1000': - - 32,022,662 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 293,044 L1I_CACHE_REFILL:u - 458,939 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 33,505,164 L1D_CACHE:u - - 30.017812847 seconds time elapsed - - 131.976276000 seconds user - 2029.636174000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 29, 124, ..., 545669, 545669, - 545671]), - col_indices=tensor([ 1, 2, 3, ..., 81869, 81699, 81863]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(81871, 81871), - nnz=545671, layout=torch.sparse_csr) -tensor([0.7679, 0.9196, 0.3474, ..., 0.5624, 0.0163, 0.8596]) -Matrix: soc-sign-Slashdot090216 -Shape: torch.Size([81871, 81871]) -NNZ: 545671 -Density: 8.140867447881048e-05 -Time: 29.926054000854492 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/soc-sign-Slashdot090216.mtx 1000': - - 553,814 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 567,372 LL_CACHE_RD:u - 199,301 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 25,193 L2D_TLB_REFILL:u - 313,278 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,796,299 L2D_CACHE:u - - 33.553779692 seconds time elapsed - - 154.498461000 seconds user - 2293.574463000 seconds sys - - - diff --git a/pytorch/output/altra_10_30_soc-sign-Slashdot090221_1000.json b/pytorch/output/altra_10_30_soc-sign-Slashdot090221_1000.json deleted file mode 100644 index 9d7e9cd..0000000 --- a/pytorch/output/altra_10_30_soc-sign-Slashdot090221_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [21.92, 21.84, 20.96, 20.24, 20.28, 20.16, 19.96, 19.72, 19.88, 19.76], "matrix": "soc-sign-Slashdot090221", "shape": [82144, 82144], "nnz": 549202, "% density": 8.13917555860553e-05, "time_s": 18.79910135269165, "power": [80.48, 80.08, 69.04, 69.04, 55.0, 46.8, 49.16, 56.2, 70.84, 82.84, 86.52, 84.28, 82.56, 81.2, 80.28, 80.28, 80.04, 80.16, 80.8, 81.0, 81.92, 83.04, 82.88], "power_after": [21.0, 20.96, 21.12, 20.76, 20.72, 20.56, 20.52, 20.64, 20.88, 21.04], "task clock (msec)": 58.57, "page faults": 3259, "cycles": 74509373, "instructions": 88672751, "branch mispredictions": 342121, "branches": 20436338, "ITLB accesses": 27189335, "ITLB misses": 6437, "DTLB misses": 18156, "DTLB accesses": 36676625, "L1I cache accesses": 30721032, "L1I cache misses": 302777, "L1D cache misses": 469833, "L1D cache accesses": 32109077, "LL cache misses": 551850, "LL cache accesses": 565355, "L2D TLB accesses": 200417, "L2D TLB misses": 25536, "L2D cache misses": 304133, "L2D cache accesses": 1801849, "instructions per cycle": 1.190088540941017, "branch miss rate": 0.016740817263836603, "ITLB miss rate": 0.0002367472393127673, "DTLB miss rate": 0.0004950291909356436, "L2D TLB miss rate": 0.12741434109880898, "L1I cache miss rate": 0.009855691045795596, "L1D cache miss rate": 0.014632404413244267, "L2D cache miss rate": 0.16878939356183564, "LL cache miss rate": 0.9761123541845389} diff --git a/pytorch/output/altra_10_30_soc-sign-Slashdot090221_1000.output b/pytorch/output/altra_10_30_soc-sign-Slashdot090221_1000.output deleted file mode 100644 index 57d4bd0..0000000 --- a/pytorch/output/altra_10_30_soc-sign-Slashdot090221_1000.output +++ /dev/null @@ -1,163 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3394979 queued and waiting for resources -srun: job 3394979 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 29, 124, ..., 549200, 549200, - 549202]), - col_indices=tensor([ 1, 2, 3, ..., 82142, 81974, 82136]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(82144, 82144), - nnz=549202, layout=torch.sparse_csr) -tensor([0.4201, 0.7748, 0.6565, ..., 0.0517, 0.6958, 0.5341]) -Matrix: soc-sign-Slashdot090221 -Shape: torch.Size([82144, 82144]) -NNZ: 549202 -Density: 8.13917555860553e-05 -Time: 27.35153603553772 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/soc-sign-Slashdot090221.mtx 1000': - - 58.57 msec task-clock:u # 0.002 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,259 page-faults:u # 55.640 K/sec - 74,509,373 cycles:u # 1.272 GHz (58.00%) - 88,672,751 instructions:u # 1.19 insn per cycle (90.97%) - branches:u - 361,568 branch-misses:u - 31,594,797 L1-dcache-loads:u # 539.410 M/sec - 460,467 L1-dcache-load-misses:u # 1.46% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 30,148,838 L1-icache-loads:u # 514.724 M/sec - 282,768 L1-icache-load-misses:u # 0.94% of all L1-icache accesses - 19,757,856 dTLB-loads:u # 337.321 M/sec (11.69%) - dTLB-load-misses:u (0.00%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 31.087250856 seconds time elapsed - - 142.716222000 seconds user - 2102.420776000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 29, 124, ..., 549200, 549200, - 549202]), - col_indices=tensor([ 1, 2, 3, ..., 82142, 81974, 82136]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(82144, 82144), - nnz=549202, layout=torch.sparse_csr) -tensor([0.7637, 0.5328, 0.8286, ..., 0.7084, 0.8903, 0.1707]) -Matrix: soc-sign-Slashdot090221 -Shape: torch.Size([82144, 82144]) -NNZ: 549202 -Density: 8.13917555860553e-05 -Time: 17.188836336135864 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/soc-sign-Slashdot090221.mtx 1000': - - 342,121 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 20,436,338 BR_RETIRED:u - - 20.753346873 seconds time elapsed - - 98.605331000 seconds user - 1332.291974000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 29, 124, ..., 549200, 549200, - 549202]), - col_indices=tensor([ 1, 2, 3, ..., 82142, 81974, 82136]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(82144, 82144), - nnz=549202, layout=torch.sparse_csr) -tensor([0.9017, 0.8505, 0.0023, ..., 0.4182, 0.6895, 0.5023]) -Matrix: soc-sign-Slashdot090221 -Shape: torch.Size([82144, 82144]) -NNZ: 549202 -Density: 8.13917555860553e-05 -Time: 16.22375249862671 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/soc-sign-Slashdot090221.mtx 1000': - - 27,189,335 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 6,437 ITLB_WALK:u - 18,156 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 36,676,625 L1D_TLB:u - - 19.748749363 seconds time elapsed - - 103.049578000 seconds user - 1249.814927000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 29, 124, ..., 549200, 549200, - 549202]), - col_indices=tensor([ 1, 2, 3, ..., 82142, 81974, 82136]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(82144, 82144), - nnz=549202, layout=torch.sparse_csr) -tensor([0.4805, 0.2325, 0.2103, ..., 0.1710, 0.7638, 0.9368]) -Matrix: soc-sign-Slashdot090221 -Shape: torch.Size([82144, 82144]) -NNZ: 549202 -Density: 8.13917555860553e-05 -Time: 15.453373908996582 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/soc-sign-Slashdot090221.mtx 1000': - - 30,721,032 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 302,777 L1I_CACHE_REFILL:u - 469,833 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 32,109,077 L1D_CACHE:u - - 19.090250444 seconds time elapsed - - 94.904880000 seconds user - 1195.102767000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 29, 124, ..., 549200, 549200, - 549202]), - col_indices=tensor([ 1, 2, 3, ..., 82142, 81974, 82136]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(82144, 82144), - nnz=549202, layout=torch.sparse_csr) -tensor([0.8430, 0.9439, 0.4260, ..., 0.8172, 0.4243, 0.3834]) -Matrix: soc-sign-Slashdot090221 -Shape: torch.Size([82144, 82144]) -NNZ: 549202 -Density: 8.13917555860553e-05 -Time: 29.316507816314697 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/soc-sign-Slashdot090221.mtx 1000': - - 551,850 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 565,355 LL_CACHE_RD:u - 200,417 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 25,536 L2D_TLB_REFILL:u - 304,133 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,801,849 L2D_CACHE:u - - 32.859276963 seconds time elapsed - - 148.969816000 seconds user - 2252.321936000 seconds sys - - - diff --git a/pytorch/output/altra_10_30_soc-sign-epinions_1000.json b/pytorch/output/altra_10_30_soc-sign-epinions_1000.json deleted file mode 100644 index 8e6797d..0000000 --- a/pytorch/output/altra_10_30_soc-sign-epinions_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [20.32, 20.52, 20.52, 20.56, 20.6, 20.4, 20.76, 20.6, 20.36, 20.4], "matrix": "soc-sign-epinions", "shape": [131828, 131828], "nnz": 841372, "% density": 4.841419648464106e-05, "time_s": 22.52380871772766, "power": [81.24, 81.16, 74.84, 62.04, 51.6, 50.56, 52.4, 52.4, 68.24, 80.56, 91.44, 91.36, 90.28, 88.32, 86.4, 85.16, 83.64, 82.36, 82.96, 82.84, 82.84, 82.56, 82.44, 82.08, 83.64, 84.4], "power_after": [20.8, 20.88, 20.8, 20.92, 20.88, 20.88, 20.8, 20.84, 20.84, 20.6], "task clock (msec)": 63.9, "page faults": 3446, "cycles": 55931043, "instructions": 77907356, "branch mispredictions": 332778, "branches": 20000746, "ITLB accesses": 27000304, "ITLB misses": 6713, "DTLB misses": 18689, "DTLB accesses": 36395663, "L1I cache accesses": 32396405, "L1I cache misses": 292629, "L1D cache misses": 473799, "L1D cache accesses": 34061981, "LL cache misses": 542765, "LL cache accesses": 557193, "L2D TLB accesses": 203626, "L2D TLB misses": 24363, "L2D cache misses": 303397, "L2D cache accesses": 1772084, "instructions per cycle": 1.3929179901043505, "branch miss rate": 0.01663827939217867, "ITLB miss rate": 0.00024862683027568875, "DTLB miss rate": 0.0005134952480464499, "L2D TLB miss rate": 0.11964582126054629, "L1I cache miss rate": 0.009032761505481858, "L1D cache miss rate": 0.01390990735389113, "L2D cache miss rate": 0.171209152613533, "LL cache miss rate": 0.9741059202107708} diff --git a/pytorch/output/altra_10_30_soc-sign-epinions_1000.output b/pytorch/output/altra_10_30_soc-sign-epinions_1000.output deleted file mode 100644 index b183c4f..0000000 --- a/pytorch/output/altra_10_30_soc-sign-epinions_1000.output +++ /dev/null @@ -1,168 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3394990 queued and waiting for resources -srun: job 3394990 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 1, 2, ..., 841371, 841371, - 841372]), - col_indices=tensor([ 1, 128552, 3, ..., 131824, 131826, - 7714]), - values=tensor([-1., -1., 1., ..., 1., 1., 1.]), - size=(131828, 131828), nnz=841372, layout=torch.sparse_csr) -tensor([0.3914, 0.2076, 0.6733, ..., 0.4758, 0.6360, 0.6316]) -Matrix: soc-sign-epinions -Shape: torch.Size([131828, 131828]) -NNZ: 841372 -Density: 4.841419648464106e-05 -Time: 20.04187798500061 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/soc-sign-epinions.mtx 1000': - - 63.90 msec task-clock:u # 0.003 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,446 page-faults:u # 53.927 K/sec - 55,931,043 cycles:u # 0.875 GHz (85.43%) - 77,907,356 instructions:u # 1.39 insn per cycle - branches:u - 357,739 branch-misses:u - 33,000,188 L1-dcache-loads:u # 516.421 M/sec - 466,824 L1-dcache-load-misses:u # 1.41% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 31,503,048 L1-icache-loads:u # 492.992 M/sec - 301,112 L1-icache-load-misses:u # 0.96% of all L1-icache accesses - 34,740,872 dTLB-loads:u # 543.661 M/sec (18.37%) - 32,355 dTLB-load-misses:u # 0.09% of all dTLB cache accesses (12.00%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 23.478083368 seconds time elapsed - - 119.232326000 seconds user - 1541.081607000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 1, 2, ..., 841371, 841371, - 841372]), - col_indices=tensor([ 1, 128552, 3, ..., 131824, 131826, - 7714]), - values=tensor([-1., -1., 1., ..., 1., 1., 1.]), - size=(131828, 131828), nnz=841372, layout=torch.sparse_csr) -tensor([0.3970, 0.5643, 0.0036, ..., 0.0338, 0.0807, 0.3885]) -Matrix: soc-sign-epinions -Shape: torch.Size([131828, 131828]) -NNZ: 841372 -Density: 4.841419648464106e-05 -Time: 16.115705490112305 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/soc-sign-epinions.mtx 1000': - - 332,778 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 20,000,746 BR_RETIRED:u - - 19.765627973 seconds time elapsed - - 103.591961000 seconds user - 1250.845091000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 1, 2, ..., 841371, 841371, - 841372]), - col_indices=tensor([ 1, 128552, 3, ..., 131824, 131826, - 7714]), - values=tensor([-1., -1., 1., ..., 1., 1., 1.]), - size=(131828, 131828), nnz=841372, layout=torch.sparse_csr) -tensor([0.0049, 0.4550, 0.3166, ..., 0.3734, 0.8337, 0.5156]) -Matrix: soc-sign-epinions -Shape: torch.Size([131828, 131828]) -NNZ: 841372 -Density: 4.841419648464106e-05 -Time: 18.55180263519287 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/soc-sign-epinions.mtx 1000': - - 27,000,304 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 6,713 ITLB_WALK:u - 18,689 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 36,395,663 L1D_TLB:u - - 22.333459337 seconds time elapsed - - 109.075160000 seconds user - 1441.055730000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 1, 2, ..., 841371, 841371, - 841372]), - col_indices=tensor([ 1, 128552, 3, ..., 131824, 131826, - 7714]), - values=tensor([-1., -1., 1., ..., 1., 1., 1.]), - size=(131828, 131828), nnz=841372, layout=torch.sparse_csr) -tensor([0.0560, 0.8530, 0.8946, ..., 0.4591, 0.5391, 0.2898]) -Matrix: soc-sign-epinions -Shape: torch.Size([131828, 131828]) -NNZ: 841372 -Density: 4.841419648464106e-05 -Time: 25.587534427642822 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/soc-sign-epinions.mtx 1000': - - 32,396,405 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 292,629 L1I_CACHE_REFILL:u - 473,799 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 34,061,981 L1D_CACHE:u - - 29.367381835 seconds time elapsed - - 142.233743000 seconds user - 1962.747683000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 1, 2, ..., 841371, 841371, - 841372]), - col_indices=tensor([ 1, 128552, 3, ..., 131824, 131826, - 7714]), - values=tensor([-1., -1., 1., ..., 1., 1., 1.]), - size=(131828, 131828), nnz=841372, layout=torch.sparse_csr) -tensor([0.7002, 0.7829, 0.1511, ..., 0.3651, 0.2391, 0.7788]) -Matrix: soc-sign-epinions -Shape: torch.Size([131828, 131828]) -NNZ: 841372 -Density: 4.841419648464106e-05 -Time: 23.656178951263428 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/soc-sign-epinions.mtx 1000': - - 542,765 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 557,193 LL_CACHE_RD:u - 203,626 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 24,363 L2D_TLB_REFILL:u - 303,397 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,772,084 L2D_CACHE:u - - 27.453055481 seconds time elapsed - - 128.709934000 seconds user - 1831.887905000 seconds sys - - - diff --git a/pytorch/output/altra_10_30_sx-mathoverflow_1000.json b/pytorch/output/altra_10_30_sx-mathoverflow_1000.json deleted file mode 100644 index b80d667..0000000 --- a/pytorch/output/altra_10_30_sx-mathoverflow_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [16.28, 16.44, 16.68, 16.68, 16.84, 17.04, 16.84, 16.84, 16.72, 16.72], "matrix": "sx-mathoverflow", "shape": [24818, 24818], "nnz": 239978, "% density": 0.00038961697406616504, "time_s": 5.405760288238525, "power": [25.64, 20.44, 21.24, 22.16, 22.28, 27.04, 26.92, 26.28, 25.32], "power_after": [16.32, 16.44, 16.4, 16.4, 16.6, 16.48, 16.56, 16.6, 16.32, 16.44], "task clock (msec)": 50.36, "page faults": 3296, "cycles": 56049457, "instructions": 72333565, "branch mispredictions": 325529, "branches": 19463406, "ITLB accesses": 27374917, "ITLB misses": 5203, "DTLB misses": 16771, "DTLB accesses": 36373182, "L1I cache accesses": 31839975, "L1I cache misses": 274158, "L1D cache misses": 471992, "L1D cache accesses": 33638817, "LL cache misses": 538067, "LL cache accesses": 557981, "L2D TLB accesses": 170169, "L2D TLB misses": 21987, "L2D cache misses": 301746, "L2D cache accesses": 1735872, "instructions per cycle": 1.2905310572411077, "branch miss rate": 0.016725181604905125, "ITLB miss rate": 0.00019006450320927, "DTLB miss rate": 0.00046108146381034247, "L2D TLB miss rate": 0.12920684731061474, "L1I cache miss rate": 0.00861049671050307, "L1D cache miss rate": 0.014031171191305569, "L2D cache miss rate": 0.1738296372082734, "LL cache miss rate": 0.9643106127269566} diff --git a/pytorch/output/altra_10_30_sx-mathoverflow_1000.output b/pytorch/output/altra_10_30_sx-mathoverflow_1000.output deleted file mode 100644 index c0ac043..0000000 --- a/pytorch/output/altra_10_30_sx-mathoverflow_1000.output +++ /dev/null @@ -1,163 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3394987 queued and waiting for resources -srun: job 3394987 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 317, 416, ..., 239976, 239977, - 239978]), - col_indices=tensor([ 0, 1, 2, ..., 1483, 2179, 24817]), - values=tensor([151., 17., 6., ..., 1., 1., 1.]), - size=(24818, 24818), nnz=239978, layout=torch.sparse_csr) -tensor([0.8864, 0.5637, 0.9805, ..., 0.0234, 0.9487, 0.4860]) -Matrix: sx-mathoverflow -Shape: torch.Size([24818, 24818]) -NNZ: 239978 -Density: 0.00038961697406616504 -Time: 5.484489917755127 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/sx-mathoverflow.mtx 1000': - - 50.36 msec task-clock:u # 0.006 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,296 page-faults:u # 65.452 K/sec - 56,049,457 cycles:u # 1.113 GHz (49.66%) - 72,333,565 instructions:u # 1.29 insn per cycle (66.35%) - branches:u - 369,218 branch-misses:u (86.12%) - 33,730,437 L1-dcache-loads:u # 669.814 M/sec (93.88%) - 459,922 L1-dcache-load-misses:u # 1.36% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 31,827,672 L1-icache-loads:u # 632.030 M/sec - 295,060 L1-icache-load-misses:u # 0.93% of all L1-icache accesses - 54,366,618 dTLB-loads:u # 1.080 G/sec (35.64%) - 84,768 dTLB-load-misses:u # 0.16% of all dTLB cache accesses (25.48%) - 12,107,953 iTLB-loads:u # 240.438 M/sec (10.11%) - iTLB-load-misses:u (0.00%) - - 8.968532171 seconds time elapsed - - 20.749643000 seconds user - 28.745486000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 317, 416, ..., 239976, 239977, - 239978]), - col_indices=tensor([ 0, 1, 2, ..., 1483, 2179, 24817]), - values=tensor([151., 17., 6., ..., 1., 1., 1.]), - size=(24818, 24818), nnz=239978, layout=torch.sparse_csr) -tensor([0.5549, 0.0336, 0.9472, ..., 0.2657, 0.3394, 0.6185]) -Matrix: sx-mathoverflow -Shape: torch.Size([24818, 24818]) -NNZ: 239978 -Density: 0.00038961697406616504 -Time: 5.532417297363281 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/sx-mathoverflow.mtx 1000': - - 325,529 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 19,463,406 BR_RETIRED:u - - 8.912497962 seconds time elapsed - - 20.214519000 seconds user - 31.566513000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 317, 416, ..., 239976, 239977, - 239978]), - col_indices=tensor([ 0, 1, 2, ..., 1483, 2179, 24817]), - values=tensor([151., 17., 6., ..., 1., 1., 1.]), - size=(24818, 24818), nnz=239978, layout=torch.sparse_csr) -tensor([0.3330, 0.8843, 0.5150, ..., 0.7292, 0.0873, 0.4184]) -Matrix: sx-mathoverflow -Shape: torch.Size([24818, 24818]) -NNZ: 239978 -Density: 0.00038961697406616504 -Time: 5.457342863082886 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/sx-mathoverflow.mtx 1000': - - 27,374,917 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 5,203 ITLB_WALK:u - 16,771 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 36,373,182 L1D_TLB:u - - 8.730534933 seconds time elapsed - - 20.156482000 seconds user - 31.426118000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 317, 416, ..., 239976, 239977, - 239978]), - col_indices=tensor([ 0, 1, 2, ..., 1483, 2179, 24817]), - values=tensor([151., 17., 6., ..., 1., 1., 1.]), - size=(24818, 24818), nnz=239978, layout=torch.sparse_csr) -tensor([0.5864, 0.4449, 0.4042, ..., 0.1651, 0.7793, 0.8302]) -Matrix: sx-mathoverflow -Shape: torch.Size([24818, 24818]) -NNZ: 239978 -Density: 0.00038961697406616504 -Time: 5.449937582015991 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/sx-mathoverflow.mtx 1000': - - 31,839,975 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 274,158 L1I_CACHE_REFILL:u - 471,992 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 33,638,817 L1D_CACHE:u - - 8.845491835 seconds time elapsed - - 20.577696000 seconds user - 35.105662000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 317, 416, ..., 239976, 239977, - 239978]), - col_indices=tensor([ 0, 1, 2, ..., 1483, 2179, 24817]), - values=tensor([151., 17., 6., ..., 1., 1., 1.]), - size=(24818, 24818), nnz=239978, layout=torch.sparse_csr) -tensor([0.8880, 0.4700, 0.5542, ..., 0.8505, 0.9123, 0.5742]) -Matrix: sx-mathoverflow -Shape: torch.Size([24818, 24818]) -NNZ: 239978 -Density: 0.00038961697406616504 -Time: 5.400304794311523 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/sx-mathoverflow.mtx 1000': - - 538,067 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 557,981 LL_CACHE_RD:u - 170,169 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 21,987 L2D_TLB_REFILL:u - 301,746 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,735,872 L2D_CACHE:u - - 8.606800178 seconds time elapsed - - 21.064990000 seconds user - 34.158762000 seconds sys - - - diff --git a/pytorch/output/altra_10_30_ut2010_1000.json b/pytorch/output/altra_10_30_ut2010_1000.json deleted file mode 100644 index 42586ca..0000000 --- a/pytorch/output/altra_10_30_ut2010_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [20.36, 20.4, 20.68, 20.64, 20.92, 20.92, 20.88, 20.68, 20.68, 20.6], "matrix": "ut2010", "shape": [115406, 115406], "nnz": 572066, "% density": 4.295259032005559e-05, "time_s": 11.10523509979248, "power": [90.68, 90.68, 88.24, 72.2, 59.48, 52.0, 54.72, 64.28, 79.24, 94.08, 96.24, 93.72, 92.36, 92.36, 90.08], "power_after": [21.24, 21.28, 20.96, 21.16, 20.92, 21.04, 21.32, 21.56, 21.16, 21.24], "task clock (msec)": 52.22, "page faults": 3288, "cycles": 67463873, "instructions": 73042754, "branch mispredictions": 344635, "branches": 20775821, "ITLB accesses": 27488750, "ITLB misses": 6494, "DTLB misses": 18293, "DTLB accesses": 36697113, "L1I cache accesses": 31066176, "L1I cache misses": 298652, "L1D cache misses": 473808, "L1D cache accesses": 32572985, "LL cache misses": 547428, "LL cache accesses": 566356, "L2D TLB accesses": 162858, "L2D TLB misses": 19852, "L2D cache misses": 304056, "L2D cache accesses": 1713420, "instructions per cycle": 1.0826943481291091, "branch miss rate": 0.01658827345499367, "ITLB miss rate": 0.00023624209904051657, "DTLB miss rate": 0.0004984860798177775, "L2D TLB miss rate": 0.12189760404769799, "L1I cache miss rate": 0.009613413636747567, "L1D cache miss rate": 0.014546041758223879, "L2D cache miss rate": 0.17745561508561825, "LL cache miss rate": 0.9665793246650517} diff --git a/pytorch/output/altra_10_30_ut2010_1000.output b/pytorch/output/altra_10_30_ut2010_1000.output deleted file mode 100644 index c2f4c79..0000000 --- a/pytorch/output/altra_10_30_ut2010_1000.output +++ /dev/null @@ -1,173 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3394993 queued and waiting for resources -srun: job 3394993 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 9, ..., 572056, 572061, - 572066]), - col_indices=tensor([ 453, 1291, 1979, ..., 113521, 114509, - 114602]), - values=tensor([160642., 31335., 282373., ..., 88393., 99485., - 18651.]), size=(115406, 115406), nnz=572066, - layout=torch.sparse_csr) -tensor([0.6983, 0.2845, 0.5984, ..., 0.1182, 0.9468, 0.3161]) -Matrix: ut2010 -Shape: torch.Size([115406, 115406]) -NNZ: 572066 -Density: 4.295259032005559e-05 -Time: 8.604448795318604 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ut2010.mtx 1000': - - 52.22 msec task-clock:u # 0.004 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,288 page-faults:u # 62.965 K/sec - 67,463,873 cycles:u # 1.292 GHz (52.95%) - 73,042,754 instructions:u # 1.08 insn per cycle (71.78%) - branches:u - 376,297 branch-misses:u (87.57%) - 34,189,906 L1-dcache-loads:u # 654.731 M/sec (97.72%) - 471,636 L1-dcache-load-misses:u # 1.38% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 31,870,328 L1-icache-loads:u # 610.312 M/sec - 297,680 L1-icache-load-misses:u # 0.93% of all L1-icache accesses - 57,623,823 dTLB-loads:u # 1.103 G/sec (30.16%) - 75,454 dTLB-load-misses:u # 0.13% of all dTLB cache accesses (24.31%) - 0 iTLB-loads:u # 0.000 /sec (3.96%) - iTLB-load-misses:u (0.00%) - - 12.112100803 seconds time elapsed - - 66.253313000 seconds user - 675.855469000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 9, ..., 572056, 572061, - 572066]), - col_indices=tensor([ 453, 1291, 1979, ..., 113521, 114509, - 114602]), - values=tensor([160642., 31335., 282373., ..., 88393., 99485., - 18651.]), size=(115406, 115406), nnz=572066, - layout=torch.sparse_csr) -tensor([0.0260, 0.8569, 0.4315, ..., 0.5243, 0.8018, 0.1763]) -Matrix: ut2010 -Shape: torch.Size([115406, 115406]) -NNZ: 572066 -Density: 4.295259032005559e-05 -Time: 8.702903270721436 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ut2010.mtx 1000': - - 344,635 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 20,775,821 BR_RETIRED:u - - 12.383096073 seconds time elapsed - - 64.544546000 seconds user - 688.477174000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 9, ..., 572056, 572061, - 572066]), - col_indices=tensor([ 453, 1291, 1979, ..., 113521, 114509, - 114602]), - values=tensor([160642., 31335., 282373., ..., 88393., 99485., - 18651.]), size=(115406, 115406), nnz=572066, - layout=torch.sparse_csr) -tensor([0.7940, 0.1585, 0.6879, ..., 0.4017, 0.1738, 0.9713]) -Matrix: ut2010 -Shape: torch.Size([115406, 115406]) -NNZ: 572066 -Density: 4.295259032005559e-05 -Time: 7.38647985458374 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ut2010.mtx 1000': - - 27,488,750 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 6,494 ITLB_WALK:u - 18,293 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 36,697,113 L1D_TLB:u - - 10.936742446 seconds time elapsed - - 63.993242000 seconds user - 580.515047000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 9, ..., 572056, 572061, - 572066]), - col_indices=tensor([ 453, 1291, 1979, ..., 113521, 114509, - 114602]), - values=tensor([160642., 31335., 282373., ..., 88393., 99485., - 18651.]), size=(115406, 115406), nnz=572066, - layout=torch.sparse_csr) -tensor([0.2725, 0.6578, 0.8180, ..., 0.0148, 0.5094, 0.1155]) -Matrix: ut2010 -Shape: torch.Size([115406, 115406]) -NNZ: 572066 -Density: 4.295259032005559e-05 -Time: 12.719107389450073 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ut2010.mtx 1000': - - 31,066,176 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 298,652 L1I_CACHE_REFILL:u - 473,808 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 32,572,985 L1D_CACHE:u - - 16.299576479 seconds time elapsed - - 86.072431000 seconds user - 987.199923000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 9, ..., 572056, 572061, - 572066]), - col_indices=tensor([ 453, 1291, 1979, ..., 113521, 114509, - 114602]), - values=tensor([160642., 31335., 282373., ..., 88393., 99485., - 18651.]), size=(115406, 115406), nnz=572066, - layout=torch.sparse_csr) -tensor([0.1156, 0.5715, 0.3099, ..., 0.3964, 0.9672, 0.5694]) -Matrix: ut2010 -Shape: torch.Size([115406, 115406]) -NNZ: 572066 -Density: 4.295259032005559e-05 -Time: 12.682909727096558 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ut2010.mtx 1000': - - 547,428 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 566,356 LL_CACHE_RD:u - 162,858 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 19,852 L2D_TLB_REFILL:u - 304,056 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,713,420 L2D_CACHE:u - - 16.221517033 seconds time elapsed - - 79.927661000 seconds user - 988.333919000 seconds sys - - - diff --git a/pytorch/output/altra_10_30_vt2010_1000.json b/pytorch/output/altra_10_30_vt2010_1000.json deleted file mode 100644 index 116bcce..0000000 --- a/pytorch/output/altra_10_30_vt2010_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [20.88, 20.76, 20.76, 20.96, 20.92, 20.88, 20.72, 20.4, 20.4, 20.24], "matrix": "vt2010", "shape": [32580, 32580], "nnz": 155598, "% density": 0.00014658915806621921, "time_s": 3.6774682998657227, "power": [34.12, 31.52, 30.36, 27.2, 27.16, 30.64, 31.0, 31.32], "power_after": [20.44, 20.52, 20.68, 20.72, 20.68, 20.72, 20.88, 20.8, 20.88, 20.52], "task clock (msec)": 48.59, "page faults": 3274, "cycles": 55030923, "instructions": 78222423, "branch mispredictions": 323004, "branches": 19091130, "ITLB accesses": 27178617, "ITLB misses": 6398, "DTLB misses": 19770, "DTLB accesses": 36355567, "L1I cache accesses": 31341858, "L1I cache misses": 291951, "L1D cache misses": 468242, "L1D cache accesses": 32805413, "LL cache misses": 520057, "LL cache accesses": 541186, "L2D TLB accesses": 191068, "L2D TLB misses": 22725, "L2D cache misses": 288895, "L2D cache accesses": 1728320, "instructions per cycle": 1.4214266949511278, "branch miss rate": 0.01691906136514706, "ITLB miss rate": 0.00023540564996371965, "DTLB miss rate": 0.0005437956723381593, "L2D TLB miss rate": 0.11893671363074926, "L1I cache miss rate": 0.009315050817982775, "L1D cache miss rate": 0.014273315199537345, "L2D cache miss rate": 0.16715365210146269, "LL cache miss rate": 0.9609579700879181} diff --git a/pytorch/output/altra_10_30_vt2010_1000.output b/pytorch/output/altra_10_30_vt2010_1000.output deleted file mode 100644 index 416e3a7..0000000 --- a/pytorch/output/altra_10_30_vt2010_1000.output +++ /dev/null @@ -1,163 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3394988 queued and waiting for resources -srun: job 3394988 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 4, 7, ..., 155588, 155592, - 155598]), - col_indices=tensor([ 131, 561, 996, ..., 32237, 32238, 32570]), - values=tensor([79040., 7820., 15136., ..., 2828., 17986., 2482.]), - size=(32580, 32580), nnz=155598, layout=torch.sparse_csr) -tensor([0.2022, 0.3400, 0.2561, ..., 0.8370, 0.0285, 0.6506]) -Matrix: vt2010 -Shape: torch.Size([32580, 32580]) -NNZ: 155598 -Density: 0.00014658915806621921 -Time: 3.74875545501709 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/vt2010.mtx 1000': - - 48.59 msec task-clock:u # 0.007 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,274 page-faults:u # 67.376 K/sec - 55,030,923 cycles:u # 1.132 GHz (65.54%) - 78,222,423 instructions:u # 1.42 insn per cycle (83.60%) - branches:u - 369,917 branch-misses:u - 32,435,815 L1-dcache-loads:u # 667.500 M/sec - 467,963 L1-dcache-load-misses:u # 1.44% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 31,013,287 L1-icache-loads:u # 638.226 M/sec - 289,982 L1-icache-load-misses:u # 0.94% of all L1-icache accesses - 60,644,978 dTLB-loads:u # 1.248 G/sec (17.29%) - dTLB-load-misses:u (0.00%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 6.978143797 seconds time elapsed - - 18.401752000 seconds user - 28.060858000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 4, 7, ..., 155588, 155592, - 155598]), - col_indices=tensor([ 131, 561, 996, ..., 32237, 32238, 32570]), - values=tensor([79040., 7820., 15136., ..., 2828., 17986., 2482.]), - size=(32580, 32580), nnz=155598, layout=torch.sparse_csr) -tensor([0.3381, 0.0423, 0.5363, ..., 0.0429, 0.4077, 0.4744]) -Matrix: vt2010 -Shape: torch.Size([32580, 32580]) -NNZ: 155598 -Density: 0.00014658915806621921 -Time: 3.7925527095794678 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/vt2010.mtx 1000': - - 323,004 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 19,091,130 BR_RETIRED:u - - 7.233250772 seconds time elapsed - - 19.111768000 seconds user - 32.178633000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 4, 7, ..., 155588, 155592, - 155598]), - col_indices=tensor([ 131, 561, 996, ..., 32237, 32238, 32570]), - values=tensor([79040., 7820., 15136., ..., 2828., 17986., 2482.]), - size=(32580, 32580), nnz=155598, layout=torch.sparse_csr) -tensor([0.7962, 0.6492, 0.2778, ..., 0.5407, 0.1159, 0.3587]) -Matrix: vt2010 -Shape: torch.Size([32580, 32580]) -NNZ: 155598 -Density: 0.00014658915806621921 -Time: 3.668635129928589 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/vt2010.mtx 1000': - - 27,178,617 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 6,398 ITLB_WALK:u - 19,770 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 36,355,567 L1D_TLB:u - - 6.925944164 seconds time elapsed - - 18.970654000 seconds user - 30.786317000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 4, 7, ..., 155588, 155592, - 155598]), - col_indices=tensor([ 131, 561, 996, ..., 32237, 32238, 32570]), - values=tensor([79040., 7820., 15136., ..., 2828., 17986., 2482.]), - size=(32580, 32580), nnz=155598, layout=torch.sparse_csr) -tensor([0.8340, 0.3434, 0.3449, ..., 0.9828, 0.6683, 0.0312]) -Matrix: vt2010 -Shape: torch.Size([32580, 32580]) -NNZ: 155598 -Density: 0.00014658915806621921 -Time: 3.623232126235962 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/vt2010.mtx 1000': - - 31,341,858 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 291,951 L1I_CACHE_REFILL:u - 468,242 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 32,805,413 L1D_CACHE:u - - 6.941260499 seconds time elapsed - - 18.410270000 seconds user - 27.908787000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 4, 7, ..., 155588, 155592, - 155598]), - col_indices=tensor([ 131, 561, 996, ..., 32237, 32238, 32570]), - values=tensor([79040., 7820., 15136., ..., 2828., 17986., 2482.]), - size=(32580, 32580), nnz=155598, layout=torch.sparse_csr) -tensor([0.2754, 0.3661, 0.9484, ..., 0.7285, 0.5354, 0.4116]) -Matrix: vt2010 -Shape: torch.Size([32580, 32580]) -NNZ: 155598 -Density: 0.00014658915806621921 -Time: 3.7337992191314697 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/vt2010.mtx 1000': - - 520,057 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 541,186 LL_CACHE_RD:u - 191,068 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 22,725 L2D_TLB_REFILL:u - 288,895 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,728,320 L2D_CACHE:u - - 7.164825085 seconds time elapsed - - 18.193885000 seconds user - 30.023194000 seconds sys - - - diff --git a/pytorch/output_HPC.json b/pytorch/output_HPC.json deleted file mode 100644 index 9be58d5..0000000 --- a/pytorch/output_HPC.json +++ /dev/null @@ -1,1406 +0,0 @@ -[ - { - "architecture": "altra", - "iterations": 1000, - "baseline_time_s": 10, - "baseline_delay_s": 30, - "power_before": [ - 80.64, - 75.2, - 61.4, - 49.84, - 38.12, - 38.12, - 24.16, - 22.68, - 22.36, - 22.2 - ], - "matrix": "ASIC_680k", - "shape": [ - 682862, - 682862 - ], - "nnz": 3871773, - "% density": 8.303171256088674e-06, - "time_s": 41.51614689826965, - "power": [ - 92.4, - 92.44, - 84.28, - 73.04, - 59.64, - 53.28, - 56.32, - 65.88, - 80.28, - 93.76, - 96.16, - 94.44, - 94.44, - 94.2, - 93.92, - 92.48, - 92.16, - 91.84, - 92.08, - 91.84, - 91.68, - 90.68, - 90.88, - 90.28, - 90.28, - 92.44, - 92.52, - 92.84, - 90.0, - 89.64, - 88.16, - 87.28, - 88.12, - 88.24, - 88.08, - 85.72, - 85.12, - 85.12, - 81.72, - 82.52, - 83.84, - 86.32, - 88.8, - 91.0, - 90.2 - ], - "power_after": [ - 21.92, - 21.88, - 21.92, - 21.88, - 21.88, - 21.72, - 21.72, - 21.72, - 21.72, - 21.44 - ], - "task clock (msec)": 55.74, - "page faults": 3266, - "cycles": 51085608, - "instructions": 88049969, - "branch mispredictions": 332704, - "branches": 20219525, - "ITLB accesses": 27856157, - "ITLB misses": 6496, - "DTLB misses": 17046, - "DTLB accesses": 37522360, - "L1I cache accesses": 31475230, - "L1I cache misses": 277921, - "L1D cache misses": 462005, - "L1D cache accesses": 33126938, - "LL cache misses": 558923, - "LL cache accesses": 571263, - "L2D TLB accesses": 190627, - "L2D TLB misses": 24234, - "L2D cache misses": 314815, - "L2D cache accesses": 1760110, - "instructions per cycle": 1.7235768046452535, - "branch miss rate": 0.01645459030318467, - "ITLB miss rate": 0.00023319799640704206, - "DTLB miss rate": 0.0004542891225392006, - "L2D TLB miss rate": 0.12712784652751186, - "L1I cache miss rate": 0.008829832220447635, - "L1D cache miss rate": 0.013946504805243395, - "L2D cache miss rate": 0.17886098027964162, - "LL cache miss rate": 0.978398741035215, - "size": 466300511044 - }, - { - "architecture": "altra", - "iterations": 1000, - "baseline_time_s": 10, - "baseline_delay_s": 30, - "power_before": [ - 31.36, - 30.64, - 31.12, - 24.52, - 24.16, - 23.12, - 22.08, - 21.28, - 21.16, - 20.88 - ], - "matrix": "rma10", - "shape": [ - 46835, - 46835 - ], - "nnz": 2374001, - "% density": 0.0010822805369125833, - "time_s": 68.86891412734985, - "power": [ - 81.8, - 81.32, - 75.08, - 63.48, - 51.92, - 51.96, - 51.8, - 65.0, - 65.0, - 75.12, - 82.68, - 82.32, - 82.08, - 82.76, - 82.8, - 83.6, - 83.36, - 83.08, - 82.88, - 83.0, - 83.32, - 83.32, - 83.36, - 84.64, - 84.56, - 84.24, - 83.52, - 83.4, - 83.36, - 83.36, - 83.72, - 84.16, - 83.24, - 82.76, - 82.76, - 82.96, - 82.36, - 82.24, - 81.64, - 81.6, - 81.4, - 81.6, - 81.88, - 82.32, - 83.04, - 83.48, - 83.48, - 84.32, - 84.04, - 84.32, - 83.16, - 82.44, - 81.96, - 81.4, - 81.8, - 82.08, - 81.8, - 81.84, - 82.04, - 82.04, - 82.08, - 82.44, - 82.6, - 82.84, - 83.8, - 84.24, - 84.6, - 85.4, - 85.6, - 86.0, - 85.72, - 85.36 - ], - "power_after": [ - 21.96, - 21.88, - 21.96, - 21.96, - 22.0, - 21.68, - 21.44, - 21.16, - 21.04, - 20.92 - ], - "task clock (msec)": 58.3, - "page faults": 3281, - "cycles": 81319364, - "instructions": 90830397, - "branch mispredictions": 342237, - "branches": 20641135, - "ITLB accesses": 27974213, - "ITLB misses": 6660, - "DTLB misses": 18441, - "DTLB accesses": 37780346, - "L1I cache accesses": 31166891, - "L1I cache misses": 291301, - "L1D cache misses": 477186, - "L1D cache accesses": 32682323, - "LL cache misses": 538552, - "LL cache accesses": 552543, - "L2D TLB accesses": 202351, - "L2D TLB misses": 24178, - "L2D cache misses": 298051, - "L2D cache accesses": 1775481, - "instructions per cycle": 1.1169590185186398, - "branch miss rate": 0.01658033824205888, - "ITLB miss rate": 0.00023807640272132053, - "DTLB miss rate": 0.00048811093471722044, - "L2D TLB miss rate": 0.11948544855226809, - "L1I cache miss rate": 0.00934648887500521, - "L1D cache miss rate": 0.014600736918241704, - "L2D cache miss rate": 0.1678705657790762, - "LL cache miss rate": 0.9746788937693537, - "size": 2193517225 - }, - { - "architecture": "altra", - "iterations": 1000, - "baseline_time_s": 10, - "baseline_delay_s": 30, - "power_before": [ - 16.44, - 16.44, - 16.44, - 16.84, - 16.72, - 16.6, - 16.72, - 16.84, - 16.68, - 16.84 - ], - "matrix": "p2p-Gnutella30", - "shape": [ - 36682, - 36682 - ], - "nnz": 88328, - "% density": 6.564359899804003e-05, - "time_s": 2.896674871444702, - "power": [ - 56.32, - 68.24, - 71.76, - 59.48, - 47.6, - 48.76, - 52.6 - ], - "power_after": [ - 16.92, - 17.0, - 16.96, - 16.8, - 16.48, - 16.52, - 16.52, - 16.52, - 16.24, - 16.36 - ], - "task clock (msec)": 56.47, - "page faults": 3222, - "cycles": 69105836, - "instructions": 89065155, - "branch mispredictions": 333669, - "branches": 20078755, - "ITLB accesses": 26015038, - "ITLB misses": 5212, - "DTLB misses": 17039, - "DTLB accesses": 35296010, - "L1I cache accesses": 31837486, - "L1I cache misses": 293353, - "L1D cache misses": 462358, - "L1D cache accesses": 33478540, - "LL cache misses": 546516, - "LL cache accesses": 559865, - "L2D TLB accesses": 190400, - "L2D TLB misses": 23787, - "L2D cache misses": 307032, - "L2D cache accesses": 1768186, - "instructions per cycle": 1.288822480926213, - "branch miss rate": 0.016618012421586895, - "ITLB miss rate": 0.00020034566161310238, - "DTLB miss rate": 0.00048274578344691083, - "L2D TLB miss rate": 0.12493172268907562, - "L1I cache miss rate": 0.009214075508348869, - "L1D cache miss rate": 0.013810578358554464, - "L2D cache miss rate": 0.17364236567872385, - "LL cache miss rate": 0.9761567520741607, - "size": 1345569124 - }, - { - "architecture": "altra", - "iterations": 1000, - "baseline_time_s": 10, - "baseline_delay_s": 30, - "power_before": [ - 32.08, - 31.8, - 28.68, - 27.6, - 22.96, - 22.08, - 21.0, - 20.84, - 20.68, - 20.72 - ], - "matrix": "va2010", - "shape": [ - 285762, - 285762 - ], - "nnz": 1402128, - "% density": 1.717033263003816e-05, - "time_s": 14.632386922836304, - "power": [ - 85.16, - 83.48, - 76.96, - 67.44, - 54.04, - 51.4, - 54.24, - 66.76, - 83.2, - 96.44, - 96.44, - 95.84, - 94.24, - 92.36, - 91.2, - 89.32, - 87.48, - 88.68, - 88.24 - ], - "power_after": [ - 21.12, - 21.0, - 21.16, - 21.4, - 21.32, - 21.36, - 21.36, - 21.12, - 20.76, - 20.84 - ], - "task clock (msec)": 57.32, - "page faults": 3280, - "cycles": 39497791, - "instructions": 64385555, - "branch mispredictions": 332792, - "branches": 19983954, - "ITLB accesses": 27156853, - "ITLB misses": 6466, - "DTLB misses": 18244, - "DTLB accesses": 36466301, - "L1I cache accesses": 30929971, - "L1I cache misses": 291811, - "L1D cache misses": 473063, - "L1D cache accesses": 32462905, - "LL cache misses": 544953, - "LL cache accesses": 565172, - "L2D TLB accesses": 183225, - "L2D TLB misses": 23924, - "L2D cache misses": 301362, - "L2D cache accesses": 1756590, - "instructions per cycle": 1.6301052127193645, - "branch miss rate": 0.01665296067034582, - "ITLB miss rate": 0.00023809828038616994, - "DTLB miss rate": 0.000500297521264907, - "L2D TLB miss rate": 0.13057170145995362, - "L1I cache miss rate": 0.009434570759862659, - "L1D cache miss rate": 0.014572417348354991, - "L2D cache miss rate": 0.17156080815671274, - "LL cache miss rate": 0.964225050073252, - "size": 81659920644 - }, - { - "architecture": "altra", - "iterations": 1000, - "baseline_time_s": 10, - "baseline_delay_s": 30, - "power_before": [ - 16.6, - 16.64, - 17.04, - 17.08, - 16.92, - 17.24, - 16.88, - 16.36, - 16.4, - 16.4 - ], - "matrix": "ri2010", - "shape": [ - 25181, - 25181 - ], - "nnz": 125750, - "% density": 0.00019831796057928155, - "time_s": 2.970583200454712, - "power": [ - 23.04, - 23.28, - 23.76, - 24.12, - 21.4, - 26.28, - 26.36 - ], - "power_after": [ - 16.16, - 16.16, - 16.52, - 16.48, - 16.52, - 16.44, - 16.36, - 16.48, - 16.76, - 16.6 - ], - "task clock (msec)": 52.61, - "page faults": 3292, - "cycles": 42915672, - "instructions": 71002596, - "branch mispredictions": 344300, - "branches": 20224759, - "ITLB accesses": 26039851, - "ITLB misses": 5035, - "DTLB misses": 16402, - "DTLB accesses": 34820806, - "L1I cache accesses": 31878105, - "L1I cache misses": 299057, - "L1D cache misses": 471869, - "L1D cache accesses": 33450518, - "LL cache misses": 530093, - "LL cache accesses": 551126, - "L2D TLB accesses": 188315, - "L2D TLB misses": 22856, - "L2D cache misses": 299885, - "L2D cache accesses": 1763155, - "instructions per cycle": 1.6544677664607, - "branch miss rate": 0.01702368863826758, - "ITLB miss rate": 0.00019335748119296073, - "DTLB miss rate": 0.0004710402165877493, - "L2D TLB miss rate": 0.12137110692191275, - "L1I cache miss rate": 0.009381266546427399, - "L1D cache miss rate": 0.014106478111938357, - "L2D cache miss rate": 0.1700843090936418, - "LL cache miss rate": 0.9618363132931489, - "size": 634082761 - }, - { - "architecture": "altra", - "iterations": 1000, - "baseline_time_s": 10, - "baseline_delay_s": 30, - "power_before": [ - 30.08, - 25.12, - 24.68, - 23.68, - 22.84, - 21.96, - 21.08, - 20.96, - 20.8, - 20.96 - ], - "matrix": "p2p-Gnutella04", - "shape": [ - 10879, - 10879 - ], - "nnz": 39994, - "% density": 0.0003379223282393842, - "time_s": 0.9992897510528564, - "power": [ - 29.48, - 30.52, - 31.88, - 31.24, - 34.32 - ], - "power_after": [ - 20.4, - 20.6, - 20.64, - 20.76, - 20.92, - 20.84, - 20.88, - 20.88, - 20.88, - 20.84 - ], - "task clock (msec)": 52.68, - "page faults": 3272, - "cycles": 63019732, - "instructions": 73518898, - "branch mispredictions": 333423, - "branches": 19435905, - "ITLB accesses": 27447537, - "ITLB misses": 6417, - "DTLB misses": 18300, - "DTLB accesses": 37569384, - "L1I cache accesses": 30830481, - "L1I cache misses": 290545, - "L1D cache misses": 473875, - "L1D cache accesses": 32284772, - "LL cache misses": 529403, - "LL cache accesses": 549794, - "L2D TLB accesses": 198306, - "L2D TLB misses": 24497, - "L2D cache misses": 298519, - "L2D cache accesses": 1772795, - "instructions per cycle": 1.1666012480027683, - "branch miss rate": 0.017155002558409294, - "ITLB miss rate": 0.00023379146915805232, - "DTLB miss rate": 0.000487098750408045, - "L2D TLB miss rate": 0.12353131019737174, - "L1I cache miss rate": 0.009423952873132274, - "L1D cache miss rate": 0.014677972636758903, - "L2D cache miss rate": 0.16838890001381998, - "LL cache miss rate": 0.9629115632400499, - "size": 118352641 - }, - { - "architecture": "altra", - "iterations": 1000, - "baseline_time_s": 10, - "baseline_delay_s": 30, - "power_before": [ - 22.08, - 21.88, - 21.88, - 21.88, - 21.56, - 21.64, - 21.84, - 21.88, - 21.72, - 21.92 - ], - "matrix": "de2010", - "shape": [ - 24115, - 24115 - ], - "nnz": 116056, - "% density": 0.0001995689928120616, - "time_s": 2.7533018589019775, - "power": [ - 29.48, - 30.24, - 27.96, - 28.4, - 26.84, - 30.6, - 30.92 - ], - "power_after": [ - 20.84, - 21.24, - 21.2, - 21.24, - 21.28, - 20.88, - 20.68, - 20.56, - 20.52, - 20.56 - ], - "task clock (msec)": 61.38, - "page faults": 3315, - "cycles": 65013274, - "instructions": 87442627, - "branch mispredictions": 328392, - "branches": 19496396, - "ITLB accesses": 28311619, - "ITLB misses": 6963, - "DTLB misses": 17888, - "DTLB accesses": 38223408, - "L1I cache accesses": 30063404, - "L1I cache misses": 272797, - "L1D cache misses": 468341, - "L1D cache accesses": 31519623, - "LL cache misses": 538689, - "LL cache accesses": 552789, - "L2D TLB accesses": 192995, - "L2D TLB misses": 23339, - "L2D cache misses": 300578, - "L2D cache accesses": 1764035, - "instructions per cycle": 1.344996515634638, - "branch miss rate": 0.016843728451145536, - "ITLB miss rate": 0.0002459414277933028, - "DTLB miss rate": 0.00046798548156668814, - "L2D TLB miss rate": 0.12093059405684085, - "L1I cache miss rate": 0.009074055619250568, - "L1D cache miss rate": 0.01485871198395996, - "L2D cache miss rate": 0.17039231081015965, - "LL cache miss rate": 0.9744929801425137, - "size": 581533225 - }, - { - "architecture": "altra", - "iterations": 1000, - "baseline_time_s": 10, - "baseline_delay_s": 30, - "power_before": [ - 51.04, - 38.64, - 22.84, - 22.24, - 21.88, - 21.88, - 21.6, - 21.4, - 21.24, - 21.28 - ], - "matrix": "ga2010", - "shape": [ - 291086, - 291086 - ], - "nnz": 1418056, - "% density": 1.6735964475229304e-05, - "time_s": 15.249999523162842, - "power": [ - 88.88, - 89.52, - 78.6, - 64.88, - 52.64, - 52.64, - 54.76, - 60.16, - 71.44, - 86.84, - 90.72, - 89.6, - 90.56, - 90.36, - 91.68, - 91.84, - 93.4, - 93.4, - 92.72 - ], - "power_after": [ - 21.68, - 21.4, - 21.28, - 21.04, - 21.04, - 20.96, - 20.92, - 20.76, - 20.8, - 20.96 - ], - "task clock (msec)": 72.45, - "page faults": 3289, - "cycles": 24836161, - "instructions": 74134706, - "branch mispredictions": 325643, - "branches": 19697746, - "ITLB accesses": 27767290, - "ITLB misses": 5832, - "DTLB misses": 18134, - "DTLB accesses": 37063060, - "L1I cache accesses": 32135376, - "L1I cache misses": 302429, - "L1D cache misses": 484427, - "L1D cache accesses": 33639686, - "LL cache misses": 548380, - "LL cache accesses": 561312, - "L2D TLB accesses": 186006, - "L2D TLB misses": 25022, - "L2D cache misses": 304539, - "L2D cache accesses": 1750107, - "instructions per cycle": 2.9849502908279586, - "branch miss rate": 0.01653199305138771, - "ITLB miss rate": 0.00021003129941740803, - "DTLB miss rate": 0.0004892742261432272, - "L2D TLB miss rate": 0.13452254228358226, - "L1I cache miss rate": 0.009411092622659838, - "L1D cache miss rate": 0.014400461407398393, - "L2D cache miss rate": 0.17401164614506429, - "LL cache miss rate": 0.976961119662505, - "size": 84731059396 - }, - { - "architecture": "altra", - "iterations": 1000, - "baseline_time_s": 10, - "baseline_delay_s": 30, - "power_before": [ - 22.04, - 21.32, - 21.32, - 21.32, - 21.12, - 21.12, - 21.0, - 20.68, - 20.72, - 20.56 - ], - "matrix": "mac_econ_fwd500", - "shape": [ - 206500, - 206500 - ], - "nnz": 1273389, - "% density": 2.9862143765866013e-05, - "time_s": 15.046087741851807, - "power": [ - 91.88, - 91.12, - 83.92, - 72.88, - 57.76, - 51.24, - 53.12, - 62.84, - 78.32, - 91.64, - 95.8, - 95.8, - 94.08, - 92.48, - 91.6, - 89.88, - 87.36, - 87.84, - 87.32 - ], - "power_after": [ - 20.92, - 21.04, - 21.12, - 20.92, - 20.92, - 20.88, - 20.88, - 20.92, - 21.04, - 20.96 - ], - "task clock (msec)": 62.46, - "page faults": 3243, - "cycles": 57150420, - "instructions": 94155455, - "branch mispredictions": 320781, - "branches": 19491698, - "ITLB accesses": 27433101, - "ITLB misses": 7382, - "DTLB misses": 19213, - "DTLB accesses": 37123052, - "L1I cache accesses": 32027284, - "L1I cache misses": 290368, - "L1D cache misses": 471338, - "L1D cache accesses": 33366668, - "LL cache misses": 571063, - "LL cache accesses": 583554, - "L2D TLB accesses": 196434, - "L2D TLB misses": 25171, - "L2D cache misses": 329198, - "L2D cache accesses": 1814040, - "instructions per cycle": 1.6475024155553013, - "branch miss rate": 0.016457314288370363, - "ITLB miss rate": 0.0002690909788142434, - "DTLB miss rate": 0.0005175490420345827, - "L2D TLB miss rate": 0.1281397314110592, - "L1I cache miss rate": 0.009066269871650684, - "L1D cache miss rate": 0.014126013421537926, - "L2D cache miss rate": 0.1814722938854711, - "LL cache miss rate": 0.9785949543658342, - "size": 42642250000 - }, - { - "architecture": "altra", - "iterations": 1000, - "baseline_time_s": 10, - "baseline_delay_s": 30, - "power_before": [ - 94.16, - 91.68, - 78.92, - 60.88, - 46.72, - 28.36, - 22.08, - 21.64, - 21.64, - 21.64 - ], - "matrix": "mc2depi", - "shape": [ - 525825, - 525825 - ], - "nnz": 2100225, - "% density": 7.595972132902821e-06, - "time_s": 11.03979206085205, - "power": [ - 95.44, - 94.0, - 88.76, - 72.12, - 59.48, - 51.92, - 53.88, - 68.6, - 83.2, - 97.76, - 98.4, - 97.12, - 97.12, - 95.28, - 94.12 - ], - "power_after": [ - 21.48, - 21.44, - 21.28, - 21.24, - 21.16, - 21.08, - 21.24, - 21.24, - 21.24, - 21.16 - ], - "task clock (msec)": 56.14, - "page faults": 3289, - "cycles": 47515158, - "instructions": 72388154, - "branch mispredictions": 327042, - "branches": 19309026, - "ITLB accesses": 26093030, - "ITLB misses": 6189, - "DTLB misses": 17253, - "DTLB accesses": 35168741, - "L1I cache accesses": 30539322, - "L1I cache misses": 285404, - "L1D cache misses": 465747, - "L1D cache accesses": 31932803, - "LL cache misses": 530261, - "LL cache accesses": 551030, - "L2D TLB accesses": 183570, - "L2D TLB misses": 23883, - "L2D cache misses": 297006, - "L2D cache accesses": 1721848, - "instructions per cycle": 1.5234749719236964, - "branch miss rate": 0.01693726032581861, - "ITLB miss rate": 0.0002371897782664566, - "DTLB miss rate": 0.0004905776979619486, - "L2D TLB miss rate": 0.13010295799967314, - "L1I cache miss rate": 0.009345459601231487, - "L1D cache miss rate": 0.014585221347465175, - "L2D cache miss rate": 0.1724925777420539, - "LL cache miss rate": 0.9623087672177558, - "size": 276491930625 - }, - { - "architecture": "altra", - "iterations": 1000, - "baseline_time_s": 10, - "baseline_delay_s": 30, - "power_before": [ - 30.72, - 30.6, - 28.68, - 26.48, - 22.44, - 21.4, - 21.28, - 21.08, - 21.32, - 21.6 - ], - "matrix": "p2p-Gnutella24", - "shape": [ - 26518, - 26518 - ], - "nnz": 65369, - "% density": 9.295875717624285e-05, - "time_s": 1.718301773071289, - "power": [ - 31.52, - 32.48, - 33.64, - 33.88, - 33.44, - 31.52 - ], - "power_after": [ - 20.96, - 20.84, - 20.92, - 20.8, - 20.76, - 20.76, - 20.76, - 20.68, - 20.72, - 20.92 - ], - "task clock (msec)": 67.08, - "page faults": 3303, - "cycles": 61261862, - "instructions": 83757591, - "branch mispredictions": 329248, - "branches": 19953212, - "ITLB accesses": 27084694, - "ITLB misses": 7107, - "DTLB misses": 17529, - "DTLB accesses": 36684333, - "L1I cache accesses": 32158234, - "L1I cache misses": 286484, - "L1D cache misses": 474161, - "L1D cache accesses": 33730073, - "LL cache misses": 550064, - "LL cache accesses": 565245, - "L2D TLB accesses": 191046, - "L2D TLB misses": 23775, - "L2D cache misses": 307419, - "L2D cache accesses": 1772169, - "instructions per cycle": 1.3672060930828385, - "branch miss rate": 0.016501002445120115, - "ITLB miss rate": 0.0002623991247602797, - "DTLB miss rate": 0.0004778334118818516, - "L2D TLB miss rate": 0.12444646838981188, - "L1I cache miss rate": 0.008908573773049851, - "L1D cache miss rate": 0.014057514788064645, - "L2D cache miss rate": 0.1734704760099065, - "LL cache miss rate": 0.973142619572044, - "size": 703204324 - }, - { - "architecture": "altra", - "iterations": 1000, - "baseline_time_s": 10, - "baseline_delay_s": 30, - "power_before": [ - 86.48, - 72.16, - 59.36, - 41.84, - 28.44, - 22.96, - 22.92, - 22.92, - 23.04, - 23.24 - ], - "matrix": "p2p-Gnutella25", - "shape": [ - 22687, - 22687 - ], - "nnz": 54705, - "% density": 0.00010628522108964806, - "time_s": 1.431199073791504, - "power": [ - 35.16, - 36.2, - 36.72, - 37.52, - 37.52 - ], - "power_after": [ - 21.32, - 21.2, - 21.2, - 21.28, - 21.52, - 21.44, - 21.92, - 21.68, - 21.6, - 21.36 - ], - "task clock (msec)": 59.85, - "page faults": 3318, - "cycles": 76505130, - "instructions": 72343215, - "branch mispredictions": 322338, - "branches": 19784096, - "ITLB accesses": 27270404, - "ITLB misses": 6607, - "DTLB misses": 17981, - "DTLB accesses": 36751047, - "L1I cache accesses": 30620441, - "L1I cache misses": 302139, - "L1D cache misses": 471011, - "L1D cache accesses": 32141810, - "LL cache misses": 531907, - "LL cache accesses": 545159, - "L2D TLB accesses": 188244, - "L2D TLB misses": 23034, - "L2D cache misses": 293848, - "L2D cache accesses": 1757551, - "instructions per cycle": 0.945599530384433, - "branch miss rate": 0.016292783860329025, - "ITLB miss rate": 0.00024227730546272803, - "DTLB miss rate": 0.0004892649725054092, - "L2D TLB miss rate": 0.12236246573595971, - "L1I cache miss rate": 0.009867232153841285, - "L1D cache miss rate": 0.014654152955294054, - "L2D cache miss rate": 0.1671917344077071, - "LL cache miss rate": 0.9756914955086498, - "size": 514699969 - }, - { - "architecture": "altra", - "iterations": 1000, - "baseline_time_s": 10, - "baseline_delay_s": 30, - "power_before": [ - 77.2, - 64.12, - 64.12, - 48.92, - 36.2, - 21.72, - 21.88, - 22.36, - 22.36, - 22.44 - ], - "matrix": "vt2010", - "shape": [ - 32580, - 32580 - ], - "nnz": 155598, - "% density": 0.00014658915806621921, - "time_s": 3.5892834663391113, - "power": [ - 33.44, - 30.68, - 31.08, - 26.96, - 26.88, - 32.48, - 32.24 - ], - "power_after": [ - 21.24, - 21.24, - 21.36, - 21.36, - 21.2, - 21.04, - 20.84, - 20.72, - 20.72, - 20.56 - ], - "task clock (msec)": 55.26, - "page faults": 3297, - "cycles": 49276491, - "instructions": 64763517, - "branch mispredictions": 340611, - "branches": 20355849, - "ITLB accesses": 27946393, - "ITLB misses": 6805, - "DTLB misses": 17877, - "DTLB accesses": 38226912, - "L1I cache accesses": 31946141, - "L1I cache misses": 295259, - "L1D cache misses": 468136, - "L1D cache accesses": 33395666, - "LL cache misses": 527109, - "LL cache accesses": 540409, - "L2D TLB accesses": 192519, - "L2D TLB misses": 24204, - "L2D cache misses": 290933, - "L2D cache accesses": 1743452, - "instructions per cycle": 1.3142883286880147, - "branch miss rate": 0.016732831924622747, - "ITLB miss rate": 0.00024350190738389746, - "DTLB miss rate": 0.0004676548291423592, - "L2D TLB miss rate": 0.1257226559456469, - "L1I cache miss rate": 0.009242399574959616, - "L1D cache miss rate": 0.014017866869311724, - "L2D cache miss rate": 0.16687181522634406, - "LL cache miss rate": 0.9753890109158063, - "size": 1061456400 - }, - { - "architecture": "altra", - "iterations": 1000, - "baseline_time_s": 10, - "baseline_delay_s": 30, - "power_before": [ - 93.52, - 87.76, - 71.6, - 58.32, - 39.6, - 26.24, - 26.24, - 22.16, - 22.24, - 22.24 - ], - "matrix": "ut2010", - "shape": [ - 115406, - 115406 - ], - "nnz": 572066, - "% density": 4.295259032005559e-05, - "time_s": 8.478580713272095, - "power": [ - 89.68, - 88.92, - 80.84, - 68.96, - 56.64, - 54.52, - 55.88, - 70.44, - 85.36, - 85.36, - 98.2, - 96.52 - ], - "power_after": [ - 21.24, - 21.32, - 21.16, - 21.44, - 21.68, - 21.76, - 21.72, - 22.0, - 21.72, - 21.72 - ], - "task clock (msec)": 53.84, - "page faults": 3291, - "cycles": 66389970, - "instructions": 74935543, - "branch mispredictions": 330515, - "branches": 19475058, - "ITLB accesses": 26125490, - "ITLB misses": 6431, - "DTLB misses": 13728, - "DTLB accesses": 35274185, - "L1I cache accesses": 30428652, - "L1I cache misses": 288897, - "L1D cache misses": 475615, - "L1D cache accesses": 31855716, - "LL cache misses": 553829, - "LL cache accesses": 574192, - "L2D TLB accesses": 181148, - "L2D TLB misses": 23202, - "L2D cache misses": 307806, - "L2D cache accesses": 1767037, - "instructions per cycle": 1.1287178319255153, - "branch miss rate": 0.016971194642911976, - "ITLB miss rate": 0.00024615806248992844, - "DTLB miss rate": 0.0003891797925309968, - "L2D TLB miss rate": 0.12808311435952924, - "L1I cache miss rate": 0.009494242465949527, - "L1D cache miss rate": 0.014930287550278261, - "L2D cache miss rate": 0.17419329646181717, - "LL cache miss rate": 0.9645362526820297, - "size": 13318544836 - }, - { - "architecture": "altra", - "iterations": 1000, - "baseline_time_s": 10, - "baseline_delay_s": 30, - "power_before": [ - 20.72, - 20.8, - 20.96, - 21.08, - 21.4, - 21.48, - 21.48, - 21.36, - 21.08, - 21.04 - ], - "matrix": "fl2010", - "shape": [ - 484481, - 484481 - ], - "nnz": 2346294, - "% density": 9.99606174861054e-06, - "time_s": 14.43001127243042, - "power": [ - 93.04, - 93.04, - 89.16, - 77.68, - 62.92, - 55.12, - 53.84, - 64.72, - 77.04, - 89.56, - 94.4, - 94.76, - 93.52, - 93.52, - 96.04, - 97.12, - 96.44, - 93.88, - 93.72 - ], - "power_after": [ - 21.08, - 21.28, - 21.28, - 21.36, - 21.08, - 21.24, - 21.08, - 20.8, - 21.04, - 20.88 - ], - "task clock (msec)": 61.6, - "page faults": 3276, - "cycles": 41408849, - "instructions": 49118917, - "branch mispredictions": 331330, - "branches": 19331189, - "ITLB accesses": 27367982, - "ITLB misses": 6160, - "DTLB misses": 17157, - "DTLB accesses": 36828216, - "L1I cache accesses": 30147304, - "L1I cache misses": 280082, - "L1D cache misses": 454022, - "L1D cache accesses": 31595140, - "LL cache misses": 536056, - "LL cache accesses": 550006, - "L2D TLB accesses": 185998, - "L2D TLB misses": 23735, - "L2D cache misses": 296648, - "L2D cache accesses": 1723525, - "instructions per cycle": 1.1861937287848787, - "branch miss rate": 0.017139659645353425, - "ITLB miss rate": 0.00022508053388810325, - "DTLB miss rate": 0.00046586562867992305, - "L2D TLB miss rate": 0.12760889902041958, - "L1I cache miss rate": 0.009290449321770198, - "L1D cache miss rate": 0.014369994878959232, - "L2D cache miss rate": 0.172117027603313, - "LL cache miss rate": 0.97463664032756, - "size": 234721839361 - }, - { - "architecture": "altra", - "iterations": 1000, - "baseline_time_s": 10, - "baseline_delay_s": 30, - "power_before": [ - 37.56, - 23.12, - 22.32, - 22.28, - 22.28, - 21.96, - 21.76, - 21.72, - 21.88, - 21.84 - ], - "matrix": "tn2010", - "shape": [ - 240116, - 240116 - ], - "nnz": 1193966, - "% density": 2.070855328296721e-05, - "time_s": 16.282614707946777, - "power": [ - 85.48, - 85.84, - 79.28, - 70.16, - 55.52, - 49.48, - 49.48, - 60.48, - 76.32, - 88.88, - 91.0, - 91.0, - 90.68, - 88.32, - 86.92, - 86.4, - 88.08, - 86.8, - 87.32, - 87.8 - ], - "power_after": [ - 21.68, - 21.48, - 21.44, - 21.36, - 21.52, - 21.4, - 21.4, - 21.32, - 21.2, - 21.04 - ], - "task clock (msec)": 68.11, - "page faults": 3486, - "cycles": 70427921, - "instructions": 85638293, - "branch mispredictions": 333780, - "branches": 19402540, - "ITLB accesses": 26935483, - "ITLB misses": 5639, - "DTLB misses": 16688, - "DTLB accesses": 36421540, - "L1I cache accesses": 33029213, - "L1I cache misses": 302558, - "L1D cache misses": 481598, - "L1D cache accesses": 34668833, - "LL cache misses": 551659, - "LL cache accesses": 564579, - "L2D TLB accesses": 188346, - "L2D TLB misses": 24479, - "L2D cache misses": 311796, - "L2D cache accesses": 1767924, - "instructions per cycle": 1.215970765344614, - "branch miss rate": 0.017202902300420462, - "ITLB miss rate": 0.0002093521025778524, - "DTLB miss rate": 0.00045819040051573877, - "L2D TLB miss rate": 0.12996824992301403, - "L1I cache miss rate": 0.00916031514284037, - "L1D cache miss rate": 0.013891381922200843, - "L2D cache miss rate": 0.17636278482559206, - "LL cache miss rate": 0.9771156915152707, - "size": 57655693456 - } -] \ No newline at end of file diff --git a/pytorch/output_HPC/altra_10_30_ASIC_680k_1000.json b/pytorch/output_HPC/altra_10_30_ASIC_680k_1000.json deleted file mode 100644 index 851f7e0..0000000 --- a/pytorch/output_HPC/altra_10_30_ASIC_680k_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [80.64, 75.2, 61.4, 49.84, 38.12, 38.12, 24.16, 22.68, 22.36, 22.2], "matrix": "ASIC_680k", "shape": [682862, 682862], "nnz": 3871773, "% density": 8.303171256088674e-06, "time_s": 41.51614689826965, "power": [92.4, 92.44, 84.28, 73.04, 59.64, 53.28, 56.32, 65.88, 80.28, 93.76, 96.16, 94.44, 94.44, 94.2, 93.92, 92.48, 92.16, 91.84, 92.08, 91.84, 91.68, 90.68, 90.88, 90.28, 90.28, 92.44, 92.52, 92.84, 90.0, 89.64, 88.16, 87.28, 88.12, 88.24, 88.08, 85.72, 85.12, 85.12, 81.72, 82.52, 83.84, 86.32, 88.8, 91.0, 90.2], "power_after": [21.92, 21.88, 21.92, 21.88, 21.88, 21.72, 21.72, 21.72, 21.72, 21.44], "task clock (msec)": 55.74, "page faults": 3266, "cycles": 51085608, "instructions": 88049969, "branch mispredictions": 332704, "branches": 20219525, "ITLB accesses": 27856157, "ITLB misses": 6496, "DTLB misses": 17046, "DTLB accesses": 37522360, "L1I cache accesses": 31475230, "L1I cache misses": 277921, "L1D cache misses": 462005, "L1D cache accesses": 33126938, "LL cache misses": 558923, "LL cache accesses": 571263, "L2D TLB accesses": 190627, "L2D TLB misses": 24234, "L2D cache misses": 314815, "L2D cache accesses": 1760110, "instructions per cycle": 1.7235768046452535, "branch miss rate": 0.01645459030318467, "ITLB miss rate": 0.00023319799640704206, "DTLB miss rate": 0.0004542891225392006, "L2D TLB miss rate": 0.12712784652751186, "L1I cache miss rate": 0.008829832220447635, "L1D cache miss rate": 0.013946504805243395, "L2D cache miss rate": 0.17886098027964162, "LL cache miss rate": 0.978398741035215} diff --git a/pytorch/output_HPC/altra_10_30_ASIC_680k_1000.output b/pytorch/output_HPC/altra_10_30_ASIC_680k_1000.output deleted file mode 100644 index 90798af..0000000 --- a/pytorch/output_HPC/altra_10_30_ASIC_680k_1000.output +++ /dev/null @@ -1,173 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3395287 queued and waiting for resources -srun: job 3395287 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 4, ..., 3871767, - 3871770, 3871773]), - col_indices=tensor([ 0, 11698, 11699, ..., 169456, 645874, - 682861]), - values=tensor([ 3.8333e-04, -3.3333e-04, -5.0000e-05, ..., - 0.0000e+00, 0.0000e+00, 7.9289e-02]), - size=(682862, 682862), nnz=3871773, layout=torch.sparse_csr) -tensor([0.9283, 0.0381, 0.0668, ..., 0.8379, 0.4193, 0.2544]) -Matrix: ASIC_680k -Shape: torch.Size([682862, 682862]) -NNZ: 3871773 -Density: 8.303171256088674e-06 -Time: 29.317893266677856 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ASIC_680k.mtx 1000': - - 55.74 msec task-clock:u # 0.002 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,266 page-faults:u # 58.589 K/sec - 51,085,608 cycles:u # 0.916 GHz (47.05%) - 88,049,969 instructions:u # 1.72 insn per cycle (92.14%) - branches:u - 360,079 branch-misses:u - 31,381,953 L1-dcache-loads:u # 562.963 M/sec - 471,072 L1-dcache-load-misses:u # 1.50% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 29,944,756 L1-icache-loads:u # 537.181 M/sec - 283,203 L1-icache-load-misses:u # 0.95% of all L1-icache accesses - 20,217,238 dTLB-loads:u # 362.679 M/sec (11.38%) - dTLB-load-misses:u (0.00%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 33.488240295 seconds time elapsed - - 222.678572000 seconds user - 2205.889153000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 4, ..., 3871767, - 3871770, 3871773]), - col_indices=tensor([ 0, 11698, 11699, ..., 169456, 645874, - 682861]), - values=tensor([ 3.8333e-04, -3.3333e-04, -5.0000e-05, ..., - 0.0000e+00, 0.0000e+00, 7.9289e-02]), - size=(682862, 682862), nnz=3871773, layout=torch.sparse_csr) -tensor([0.3482, 0.5546, 0.8398, ..., 0.6137, 0.0654, 0.9075]) -Matrix: ASIC_680k -Shape: torch.Size([682862, 682862]) -NNZ: 3871773 -Density: 8.303171256088674e-06 -Time: 38.4066903591156 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ASIC_680k.mtx 1000': - - 332,704 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 20,219,525 BR_RETIRED:u - - 42.582064532 seconds time elapsed - - 238.965431000 seconds user - 2914.615754000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 4, ..., 3871767, - 3871770, 3871773]), - col_indices=tensor([ 0, 11698, 11699, ..., 169456, 645874, - 682861]), - values=tensor([ 3.8333e-04, -3.3333e-04, -5.0000e-05, ..., - 0.0000e+00, 0.0000e+00, 7.9289e-02]), - size=(682862, 682862), nnz=3871773, layout=torch.sparse_csr) -tensor([0.2581, 0.2884, 0.9465, ..., 0.4833, 0.3421, 0.4862]) -Matrix: ASIC_680k -Shape: torch.Size([682862, 682862]) -NNZ: 3871773 -Density: 8.303171256088674e-06 -Time: 34.74818539619446 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ASIC_680k.mtx 1000': - - 27,856,157 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 6,496 ITLB_WALK:u - 17,046 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 37,522,360 L1D_TLB:u - - 39.019872270 seconds time elapsed - - 239.678206000 seconds user - 2622.552757000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 4, ..., 3871767, - 3871770, 3871773]), - col_indices=tensor([ 0, 11698, 11699, ..., 169456, 645874, - 682861]), - values=tensor([ 3.8333e-04, -3.3333e-04, -5.0000e-05, ..., - 0.0000e+00, 0.0000e+00, 7.9289e-02]), - size=(682862, 682862), nnz=3871773, layout=torch.sparse_csr) -tensor([0.8603, 0.0423, 0.3724, ..., 0.4873, 0.6469, 0.9634]) -Matrix: ASIC_680k -Shape: torch.Size([682862, 682862]) -NNZ: 3871773 -Density: 8.303171256088674e-06 -Time: 33.05097770690918 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ASIC_680k.mtx 1000': - - 31,475,230 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 277,921 L1I_CACHE_REFILL:u - 462,005 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 33,126,938 L1D_CACHE:u - - 37.399374202 seconds time elapsed - - 239.238852000 seconds user - 2492.385966000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 4, ..., 3871767, - 3871770, 3871773]), - col_indices=tensor([ 0, 11698, 11699, ..., 169456, 645874, - 682861]), - values=tensor([ 3.8333e-04, -3.3333e-04, -5.0000e-05, ..., - 0.0000e+00, 0.0000e+00, 7.9289e-02]), - size=(682862, 682862), nnz=3871773, layout=torch.sparse_csr) -tensor([0.1993, 0.2167, 0.6338, ..., 0.0614, 0.0230, 0.4851]) -Matrix: ASIC_680k -Shape: torch.Size([682862, 682862]) -NNZ: 3871773 -Density: 8.303171256088674e-06 -Time: 32.37103772163391 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ASIC_680k.mtx 1000': - - 558,923 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 571,263 LL_CACHE_RD:u - 190,627 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 24,234 L2D_TLB_REFILL:u - 314,815 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,760,110 L2D_CACHE:u - - 36.644016288 seconds time elapsed - - 233.933818000 seconds user - 2439.284669000 seconds sys - - - diff --git a/pytorch/output_HPC/altra_10_30_de2010_1000.json b/pytorch/output_HPC/altra_10_30_de2010_1000.json deleted file mode 100644 index f8e0800..0000000 --- a/pytorch/output_HPC/altra_10_30_de2010_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [22.08, 21.88, 21.88, 21.88, 21.56, 21.64, 21.84, 21.88, 21.72, 21.92], "matrix": "de2010", "shape": [24115, 24115], "nnz": 116056, "% density": 0.0001995689928120616, "time_s": 2.7533018589019775, "power": [29.48, 30.24, 27.96, 28.4, 26.84, 30.6, 30.92], "power_after": [20.84, 21.24, 21.2, 21.24, 21.28, 20.88, 20.68, 20.56, 20.52, 20.56], "task clock (msec)": 61.38, "page faults": 3315, "cycles": 65013274, "instructions": 87442627, "branch mispredictions": 328392, "branches": 19496396, "ITLB accesses": 28311619, "ITLB misses": 6963, "DTLB misses": 17888, "DTLB accesses": 38223408, "L1I cache accesses": 30063404, "L1I cache misses": 272797, "L1D cache misses": 468341, "L1D cache accesses": 31519623, "LL cache misses": 538689, "LL cache accesses": 552789, "L2D TLB accesses": 192995, "L2D TLB misses": 23339, "L2D cache misses": 300578, "L2D cache accesses": 1764035, "instructions per cycle": 1.344996515634638, "branch miss rate": 0.016843728451145536, "ITLB miss rate": 0.0002459414277933028, "DTLB miss rate": 0.00046798548156668814, "L2D TLB miss rate": 0.12093059405684085, "L1I cache miss rate": 0.009074055619250568, "L1D cache miss rate": 0.01485871198395996, "L2D cache miss rate": 0.17039231081015965, "LL cache miss rate": 0.9744929801425137} diff --git a/pytorch/output_HPC/altra_10_30_de2010_1000.output b/pytorch/output_HPC/altra_10_30_de2010_1000.output deleted file mode 100644 index b553459..0000000 --- a/pytorch/output_HPC/altra_10_30_de2010_1000.output +++ /dev/null @@ -1,168 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3395278 queued and waiting for resources -srun: job 3395278 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 13, 21, ..., 116047, 116051, - 116056]), - col_indices=tensor([ 250, 251, 757, ..., 23334, 23553, 24050]), - values=tensor([ 14900., 33341., 20255., ..., 164227., 52413., - 16949.]), size=(24115, 24115), nnz=116056, - layout=torch.sparse_csr) -tensor([0.3547, 0.6554, 0.2142, ..., 0.8854, 0.1041, 0.2243]) -Matrix: de2010 -Shape: torch.Size([24115, 24115]) -NNZ: 116056 -Density: 0.0001995689928120616 -Time: 2.74495267868042 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/de2010.mtx 1000': - - 61.38 msec task-clock:u # 0.010 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,315 page-faults:u # 54.008 K/sec - 65,013,274 cycles:u # 1.059 GHz (90.47%) - 87,442,627 instructions:u # 1.34 insn per cycle - branches:u - 369,052 branch-misses:u - 31,570,549 L1-dcache-loads:u # 514.350 M/sec - 477,402 L1-dcache-load-misses:u # 1.51% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 30,354,192 L1-icache-loads:u # 494.533 M/sec - 294,845 L1-icache-load-misses:u # 0.97% of all L1-icache accesses - 0 dTLB-loads:u # 0.000 /sec (3.92%) - dTLB-load-misses:u (0.00%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 6.232986287 seconds time elapsed - - 17.354331000 seconds user - 29.036034000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 13, 21, ..., 116047, 116051, - 116056]), - col_indices=tensor([ 250, 251, 757, ..., 23334, 23553, 24050]), - values=tensor([ 14900., 33341., 20255., ..., 164227., 52413., - 16949.]), size=(24115, 24115), nnz=116056, - layout=torch.sparse_csr) -tensor([0.3177, 0.9122, 0.6465, ..., 0.5489, 0.2254, 0.7965]) -Matrix: de2010 -Shape: torch.Size([24115, 24115]) -NNZ: 116056 -Density: 0.0001995689928120616 -Time: 2.7603256702423096 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/de2010.mtx 1000': - - 328,392 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 19,496,396 BR_RETIRED:u - - 6.149991615 seconds time elapsed - - 17.630426000 seconds user - 30.586756000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 13, 21, ..., 116047, 116051, - 116056]), - col_indices=tensor([ 250, 251, 757, ..., 23334, 23553, 24050]), - values=tensor([ 14900., 33341., 20255., ..., 164227., 52413., - 16949.]), size=(24115, 24115), nnz=116056, - layout=torch.sparse_csr) -tensor([0.7815, 0.6240, 0.3715, ..., 0.5116, 0.5969, 0.4241]) -Matrix: de2010 -Shape: torch.Size([24115, 24115]) -NNZ: 116056 -Density: 0.0001995689928120616 -Time: 2.7978765964508057 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/de2010.mtx 1000': - - 28,311,619 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 6,963 ITLB_WALK:u - 17,888 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 38,223,408 L1D_TLB:u - - 6.151843492 seconds time elapsed - - 17.202045000 seconds user - 28.014218000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 13, 21, ..., 116047, 116051, - 116056]), - col_indices=tensor([ 250, 251, 757, ..., 23334, 23553, 24050]), - values=tensor([ 14900., 33341., 20255., ..., 164227., 52413., - 16949.]), size=(24115, 24115), nnz=116056, - layout=torch.sparse_csr) -tensor([0.9638, 0.0929, 0.0479, ..., 0.1500, 0.3117, 0.9664]) -Matrix: de2010 -Shape: torch.Size([24115, 24115]) -NNZ: 116056 -Density: 0.0001995689928120616 -Time: 2.684640884399414 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/de2010.mtx 1000': - - 30,063,404 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 272,797 L1I_CACHE_REFILL:u - 468,341 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 31,519,623 L1D_CACHE:u - - 5.874324363 seconds time elapsed - - 17.629166000 seconds user - 29.998701000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 13, 21, ..., 116047, 116051, - 116056]), - col_indices=tensor([ 250, 251, 757, ..., 23334, 23553, 24050]), - values=tensor([ 14900., 33341., 20255., ..., 164227., 52413., - 16949.]), size=(24115, 24115), nnz=116056, - layout=torch.sparse_csr) -tensor([0.3936, 0.9167, 0.4396, ..., 0.1628, 0.6361, 0.1875]) -Matrix: de2010 -Shape: torch.Size([24115, 24115]) -NNZ: 116056 -Density: 0.0001995689928120616 -Time: 2.747934103012085 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/de2010.mtx 1000': - - 538,689 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 552,789 LL_CACHE_RD:u - 192,995 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 23,339 L2D_TLB_REFILL:u - 300,578 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,764,035 L2D_CACHE:u - - 6.102012809 seconds time elapsed - - 18.001082000 seconds user - 27.986033000 seconds sys - - - diff --git a/pytorch/output_HPC/altra_10_30_fl2010_1000.json b/pytorch/output_HPC/altra_10_30_fl2010_1000.json deleted file mode 100644 index 2085602..0000000 --- a/pytorch/output_HPC/altra_10_30_fl2010_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [20.72, 20.8, 20.96, 21.08, 21.4, 21.48, 21.48, 21.36, 21.08, 21.04], "matrix": "fl2010", "shape": [484481, 484481], "nnz": 2346294, "% density": 9.99606174861054e-06, "time_s": 14.43001127243042, "power": [93.04, 93.04, 89.16, 77.68, 62.92, 55.12, 53.84, 64.72, 77.04, 89.56, 94.4, 94.76, 93.52, 93.52, 96.04, 97.12, 96.44, 93.88, 93.72], "power_after": [21.08, 21.28, 21.28, 21.36, 21.08, 21.24, 21.08, 20.8, 21.04, 20.88], "task clock (msec)": 61.6, "page faults": 3276, "cycles": 41408849, "instructions": 49118917, "branch mispredictions": 331330, "branches": 19331189, "ITLB accesses": 27367982, "ITLB misses": 6160, "DTLB misses": 17157, "DTLB accesses": 36828216, "L1I cache accesses": 30147304, "L1I cache misses": 280082, "L1D cache misses": 454022, "L1D cache accesses": 31595140, "LL cache misses": 536056, "LL cache accesses": 550006, "L2D TLB accesses": 185998, "L2D TLB misses": 23735, "L2D cache misses": 296648, "L2D cache accesses": 1723525, "instructions per cycle": 1.1861937287848787, "branch miss rate": 0.017139659645353425, "ITLB miss rate": 0.00022508053388810325, "DTLB miss rate": 0.00046586562867992305, "L2D TLB miss rate": 0.12760889902041958, "L1I cache miss rate": 0.009290449321770198, "L1D cache miss rate": 0.014369994878959232, "L2D cache miss rate": 0.172117027603313, "LL cache miss rate": 0.97463664032756} diff --git a/pytorch/output_HPC/altra_10_30_fl2010_1000.output b/pytorch/output_HPC/altra_10_30_fl2010_1000.output deleted file mode 100644 index 42e1ded..0000000 --- a/pytorch/output_HPC/altra_10_30_fl2010_1000.output +++ /dev/null @@ -1,169 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3395283 queued and waiting for resources -srun: job 3395283 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 2, 5, ..., 2346288, - 2346292, 2346294]), - col_indices=tensor([ 1513, 5311, 947, ..., 484460, 482463, - 484022]), - values=tensor([28364., 12497., 11567., ..., 8532., 22622., 35914.]), - size=(484481, 484481), nnz=2346294, layout=torch.sparse_csr) -tensor([2.0367e-04, 1.7661e-01, 2.1772e-01, ..., 1.8646e-01, 2.2210e-01, - 4.2364e-02]) -Matrix: fl2010 -Shape: torch.Size([484481, 484481]) -NNZ: 2346294 -Density: 9.99606174861054e-06 -Time: 16.31556534767151 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/fl2010.mtx 1000': - - 61.60 msec task-clock:u # 0.003 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,276 page-faults:u # 53.185 K/sec - 41,408,849 cycles:u # 0.672 GHz (41.57%) - 49,118,917 instructions:u # 1.19 insn per cycle (67.74%) - branches:u - 344,653 branch-misses:u (91.69%) - 31,501,274 L1-dcache-loads:u # 511.418 M/sec - 477,740 L1-dcache-load-misses:u # 1.52% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 30,099,667 L1-icache-loads:u # 488.663 M/sec - 285,734 L1-icache-load-misses:u # 0.95% of all L1-icache accesses - 41,879,387 dTLB-loads:u # 679.904 M/sec (54.00%) - 99,044 dTLB-load-misses:u # 0.24% of all dTLB cache accesses (13.61%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 20.288512544 seconds time elapsed - - 134.447078000 seconds user - 1247.121046000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 2, 5, ..., 2346288, - 2346292, 2346294]), - col_indices=tensor([ 1513, 5311, 947, ..., 484460, 482463, - 484022]), - values=tensor([28364., 12497., 11567., ..., 8532., 22622., 35914.]), - size=(484481, 484481), nnz=2346294, layout=torch.sparse_csr) -tensor([0.9700, 0.5813, 0.6566, ..., 0.4126, 0.7652, 0.9833]) -Matrix: fl2010 -Shape: torch.Size([484481, 484481]) -NNZ: 2346294 -Density: 9.99606174861054e-06 -Time: 16.561575651168823 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/fl2010.mtx 1000': - - 331,330 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 19,331,189 BR_RETIRED:u - - 20.603578845 seconds time elapsed - - 136.555709000 seconds user - 1264.382740000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 2, 5, ..., 2346288, - 2346292, 2346294]), - col_indices=tensor([ 1513, 5311, 947, ..., 484460, 482463, - 484022]), - values=tensor([28364., 12497., 11567., ..., 8532., 22622., 35914.]), - size=(484481, 484481), nnz=2346294, layout=torch.sparse_csr) -tensor([0.1770, 0.8270, 0.4236, ..., 0.0091, 0.2300, 0.5084]) -Matrix: fl2010 -Shape: torch.Size([484481, 484481]) -NNZ: 2346294 -Density: 9.99606174861054e-06 -Time: 17.374610424041748 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/fl2010.mtx 1000': - - 27,367,982 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 6,160 ITLB_WALK:u - 17,157 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 36,828,216 L1D_TLB:u - - 21.377378255 seconds time elapsed - - 140.848520000 seconds user - 1326.124469000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 2, 5, ..., 2346288, - 2346292, 2346294]), - col_indices=tensor([ 1513, 5311, 947, ..., 484460, 482463, - 484022]), - values=tensor([28364., 12497., 11567., ..., 8532., 22622., 35914.]), - size=(484481, 484481), nnz=2346294, layout=torch.sparse_csr) -tensor([0.1268, 0.8786, 0.9762, ..., 0.0649, 0.4474, 0.9707]) -Matrix: fl2010 -Shape: torch.Size([484481, 484481]) -NNZ: 2346294 -Density: 9.99606174861054e-06 -Time: 16.753613471984863 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/fl2010.mtx 1000': - - 30,147,304 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 280,082 L1I_CACHE_REFILL:u - 454,022 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 31,595,140 L1D_CACHE:u - - 20.706929400 seconds time elapsed - - 139.881127000 seconds user - 1278.527504000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 2, 5, ..., 2346288, - 2346292, 2346294]), - col_indices=tensor([ 1513, 5311, 947, ..., 484460, 482463, - 484022]), - values=tensor([28364., 12497., 11567., ..., 8532., 22622., 35914.]), - size=(484481, 484481), nnz=2346294, layout=torch.sparse_csr) -tensor([0.1394, 0.8842, 0.4362, ..., 0.8265, 0.1643, 0.9034]) -Matrix: fl2010 -Shape: torch.Size([484481, 484481]) -NNZ: 2346294 -Density: 9.99606174861054e-06 -Time: 14.484151124954224 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/fl2010.mtx 1000': - - 536,056 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 550,006 LL_CACHE_RD:u - 185,998 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 23,735 L2D_TLB_REFILL:u - 296,648 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,723,525 L2D_CACHE:u - - 18.443039315 seconds time elapsed - - 135.498625000 seconds user - 1101.745145000 seconds sys - - - diff --git a/pytorch/output_HPC/altra_10_30_ga2010_1000.json b/pytorch/output_HPC/altra_10_30_ga2010_1000.json deleted file mode 100644 index 16647fa..0000000 --- a/pytorch/output_HPC/altra_10_30_ga2010_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [51.04, 38.64, 22.84, 22.24, 21.88, 21.88, 21.6, 21.4, 21.24, 21.28], "matrix": "ga2010", "shape": [291086, 291086], "nnz": 1418056, "% density": 1.6735964475229304e-05, "time_s": 15.249999523162842, "power": [88.88, 89.52, 78.6, 64.88, 52.64, 52.64, 54.76, 60.16, 71.44, 86.84, 90.72, 89.6, 90.56, 90.36, 91.68, 91.84, 93.4, 93.4, 92.72], "power_after": [21.68, 21.4, 21.28, 21.04, 21.04, 20.96, 20.92, 20.76, 20.8, 20.96], "task clock (msec)": 72.45, "page faults": 3289, "cycles": 24836161, "instructions": 74134706, "branch mispredictions": 325643, "branches": 19697746, "ITLB accesses": 27767290, "ITLB misses": 5832, "DTLB misses": 18134, "DTLB accesses": 37063060, "L1I cache accesses": 32135376, "L1I cache misses": 302429, "L1D cache misses": 484427, "L1D cache accesses": 33639686, "LL cache misses": 548380, "LL cache accesses": 561312, "L2D TLB accesses": 186006, "L2D TLB misses": 25022, "L2D cache misses": 304539, "L2D cache accesses": 1750107, "instructions per cycle": 2.9849502908279586, "branch miss rate": 0.01653199305138771, "ITLB miss rate": 0.00021003129941740803, "DTLB miss rate": 0.0004892742261432272, "L2D TLB miss rate": 0.13452254228358226, "L1I cache miss rate": 0.009411092622659838, "L1D cache miss rate": 0.014400461407398393, "L2D cache miss rate": 0.17401164614506429, "LL cache miss rate": 0.976961119662505} diff --git a/pytorch/output_HPC/altra_10_30_ga2010_1000.output b/pytorch/output_HPC/altra_10_30_ga2010_1000.output deleted file mode 100644 index 5759842..0000000 --- a/pytorch/output_HPC/altra_10_30_ga2010_1000.output +++ /dev/null @@ -1,168 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3395281 queued and waiting for resources -srun: job 3395281 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 10, ..., 1418047, - 1418054, 1418056]), - col_indices=tensor([ 1566, 1871, 1997, ..., 291064, 289820, - 290176]), - values=tensor([18760., 17851., 18847., ..., 65219., 56729., 77629.]), - size=(291086, 291086), nnz=1418056, layout=torch.sparse_csr) -tensor([0.8043, 0.7164, 0.5687, ..., 0.1275, 0.5142, 0.8456]) -Matrix: ga2010 -Shape: torch.Size([291086, 291086]) -NNZ: 1418056 -Density: 1.6735964475229304e-05 -Time: 13.566045045852661 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ga2010.mtx 1000': - - 72.45 msec task-clock:u # 0.004 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,289 page-faults:u # 45.396 K/sec - 24,836,161 cycles:u # 0.343 GHz (23.15%) - 74,134,706 instructions:u # 2.98 insn per cycle (85.49%) - branches:u - 381,828 branch-misses:u - 33,748,654 L1-dcache-loads:u # 465.814 M/sec - 497,166 L1-dcache-load-misses:u # 1.47% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 32,271,900 L1-icache-loads:u # 445.431 M/sec - 311,814 L1-icache-load-misses:u # 0.97% of all L1-icache accesses - 43,431,516 dTLB-loads:u # 599.461 M/sec (27.81%) - 33,416 dTLB-load-misses:u # 0.08% of all dTLB cache accesses (4.55%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 17.276157893 seconds time elapsed - - 100.320029000 seconds user - 1057.703228000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 10, ..., 1418047, - 1418054, 1418056]), - col_indices=tensor([ 1566, 1871, 1997, ..., 291064, 289820, - 290176]), - values=tensor([18760., 17851., 18847., ..., 65219., 56729., 77629.]), - size=(291086, 291086), nnz=1418056, layout=torch.sparse_csr) -tensor([0.6290, 0.2236, 0.0669, ..., 0.6531, 0.4280, 0.4384]) -Matrix: ga2010 -Shape: torch.Size([291086, 291086]) -NNZ: 1418056 -Density: 1.6735964475229304e-05 -Time: 17.094524145126343 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ga2010.mtx 1000': - - 325,643 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 19,697,746 BR_RETIRED:u - - 20.849795214 seconds time elapsed - - 115.280665000 seconds user - 1318.654953000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 10, ..., 1418047, - 1418054, 1418056]), - col_indices=tensor([ 1566, 1871, 1997, ..., 291064, 289820, - 290176]), - values=tensor([18760., 17851., 18847., ..., 65219., 56729., 77629.]), - size=(291086, 291086), nnz=1418056, layout=torch.sparse_csr) -tensor([0.1008, 0.2309, 0.3749, ..., 0.1568, 0.8852, 0.8182]) -Matrix: ga2010 -Shape: torch.Size([291086, 291086]) -NNZ: 1418056 -Density: 1.6735964475229304e-05 -Time: 15.106332063674927 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ga2010.mtx 1000': - - 27,767,290 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 5,832 ITLB_WALK:u - 18,134 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 37,063,060 L1D_TLB:u - - 18.753509375 seconds time elapsed - - 112.958759000 seconds user - 1167.457916000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 10, ..., 1418047, - 1418054, 1418056]), - col_indices=tensor([ 1566, 1871, 1997, ..., 291064, 289820, - 290176]), - values=tensor([18760., 17851., 18847., ..., 65219., 56729., 77629.]), - size=(291086, 291086), nnz=1418056, layout=torch.sparse_csr) -tensor([0.8347, 0.6624, 0.6196, ..., 0.2250, 0.0157, 0.1843]) -Matrix: ga2010 -Shape: torch.Size([291086, 291086]) -NNZ: 1418056 -Density: 1.6735964475229304e-05 -Time: 13.73094367980957 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ga2010.mtx 1000': - - 32,135,376 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 302,429 L1I_CACHE_REFILL:u - 484,427 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 33,639,686 L1D_CACHE:u - - 17.400567824 seconds time elapsed - - 110.027662000 seconds user - 1054.271122000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 10, ..., 1418047, - 1418054, 1418056]), - col_indices=tensor([ 1566, 1871, 1997, ..., 291064, 289820, - 290176]), - values=tensor([18760., 17851., 18847., ..., 65219., 56729., 77629.]), - size=(291086, 291086), nnz=1418056, layout=torch.sparse_csr) -tensor([0.8369, 0.3399, 0.1689, ..., 0.2081, 0.0714, 0.7388]) -Matrix: ga2010 -Shape: torch.Size([291086, 291086]) -NNZ: 1418056 -Density: 1.6735964475229304e-05 -Time: 15.809288501739502 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ga2010.mtx 1000': - - 548,380 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 561,312 LL_CACHE_RD:u - 186,006 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 25,022 L2D_TLB_REFILL:u - 304,539 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,750,107 L2D_CACHE:u - - 19.626934574 seconds time elapsed - - 116.733174000 seconds user - 1214.439657000 seconds sys - - - diff --git a/pytorch/output_HPC/altra_10_30_mac_econ_fwd500_1000.json b/pytorch/output_HPC/altra_10_30_mac_econ_fwd500_1000.json deleted file mode 100644 index e01919a..0000000 --- a/pytorch/output_HPC/altra_10_30_mac_econ_fwd500_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [22.04, 21.32, 21.32, 21.32, 21.12, 21.12, 21.0, 20.68, 20.72, 20.56], "matrix": "mac_econ_fwd500", "shape": [206500, 206500], "nnz": 1273389, "% density": 2.9862143765866013e-05, "time_s": 15.046087741851807, "power": [91.88, 91.12, 83.92, 72.88, 57.76, 51.24, 53.12, 62.84, 78.32, 91.64, 95.8, 95.8, 94.08, 92.48, 91.6, 89.88, 87.36, 87.84, 87.32], "power_after": [20.92, 21.04, 21.12, 20.92, 20.92, 20.88, 20.88, 20.92, 21.04, 20.96], "task clock (msec)": 62.46, "page faults": 3243, "cycles": 57150420, "instructions": 94155455, "branch mispredictions": 320781, "branches": 19491698, "ITLB accesses": 27433101, "ITLB misses": 7382, "DTLB misses": 19213, "DTLB accesses": 37123052, "L1I cache accesses": 32027284, "L1I cache misses": 290368, "L1D cache misses": 471338, "L1D cache accesses": 33366668, "LL cache misses": 571063, "LL cache accesses": 583554, "L2D TLB accesses": 196434, "L2D TLB misses": 25171, "L2D cache misses": 329198, "L2D cache accesses": 1814040, "instructions per cycle": 1.6475024155553013, "branch miss rate": 0.016457314288370363, "ITLB miss rate": 0.0002690909788142434, "DTLB miss rate": 0.0005175490420345827, "L2D TLB miss rate": 0.1281397314110592, "L1I cache miss rate": 0.009066269871650684, "L1D cache miss rate": 0.014126013421537926, "L2D cache miss rate": 0.1814722938854711, "LL cache miss rate": 0.9785949543658342} diff --git a/pytorch/output_HPC/altra_10_30_mac_econ_fwd500_1000.output b/pytorch/output_HPC/altra_10_30_mac_econ_fwd500_1000.output deleted file mode 100644 index 55d0202..0000000 --- a/pytorch/output_HPC/altra_10_30_mac_econ_fwd500_1000.output +++ /dev/null @@ -1,173 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3395279 queued and waiting for resources -srun: job 3395279 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 8, ..., 1273376, - 1273379, 1273389]), - col_indices=tensor([ 3, 30, 44, ..., 206363, 206408, - 206459]), - values=tensor([-3.7877e-03, -1.5420e-01, 9.5305e-04, ..., - 1.2290e-01, 2.2235e-01, -1.0000e+00]), - size=(206500, 206500), nnz=1273389, layout=torch.sparse_csr) -tensor([0.5388, 0.2921, 0.7349, ..., 0.6379, 0.9676, 0.6389]) -Matrix: mac_econ_fwd500 -Shape: torch.Size([206500, 206500]) -NNZ: 1273389 -Density: 2.9862143765866013e-05 -Time: 21.700236320495605 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/mac_econ_fwd500.mtx 1000': - - 62.46 msec task-clock:u # 0.002 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,243 page-faults:u # 51.921 K/sec - 57,150,420 cycles:u # 0.915 GHz (90.14%) - 94,155,455 instructions:u # 1.65 insn per cycle - branches:u - 373,032 branch-misses:u - 33,654,742 L1-dcache-loads:u # 538.817 M/sec - 479,068 L1-dcache-load-misses:u # 1.42% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 32,149,866 L1-icache-loads:u # 514.724 M/sec - 293,643 L1-icache-load-misses:u # 0.91% of all L1-icache accesses - 0 dTLB-loads:u # 0.000 /sec (5.14%) - dTLB-load-misses:u (0.00%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 25.310174677 seconds time elapsed - - 125.287203000 seconds user - 1680.798909000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 8, ..., 1273376, - 1273379, 1273389]), - col_indices=tensor([ 3, 30, 44, ..., 206363, 206408, - 206459]), - values=tensor([-3.7877e-03, -1.5420e-01, 9.5305e-04, ..., - 1.2290e-01, 2.2235e-01, -1.0000e+00]), - size=(206500, 206500), nnz=1273389, layout=torch.sparse_csr) -tensor([0.6433, 0.3677, 0.3308, ..., 0.5364, 0.2509, 0.4204]) -Matrix: mac_econ_fwd500 -Shape: torch.Size([206500, 206500]) -NNZ: 1273389 -Density: 2.9862143765866013e-05 -Time: 16.171404361724854 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/mac_econ_fwd500.mtx 1000': - - 320,781 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 19,491,698 BR_RETIRED:u - - 19.988421837 seconds time elapsed - - 112.429117000 seconds user - 1245.246161000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 8, ..., 1273376, - 1273379, 1273389]), - col_indices=tensor([ 3, 30, 44, ..., 206363, 206408, - 206459]), - values=tensor([-3.7877e-03, -1.5420e-01, 9.5305e-04, ..., - 1.2290e-01, 2.2235e-01, -1.0000e+00]), - size=(206500, 206500), nnz=1273389, layout=torch.sparse_csr) -tensor([0.9344, 0.9844, 0.2313, ..., 0.8634, 0.6912, 0.9693]) -Matrix: mac_econ_fwd500 -Shape: torch.Size([206500, 206500]) -NNZ: 1273389 -Density: 2.9862143765866013e-05 -Time: 11.788637161254883 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/mac_econ_fwd500.mtx 1000': - - 27,433,101 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 7,382 ITLB_WALK:u - 19,213 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 37,123,052 L1D_TLB:u - - 15.542834153 seconds time elapsed - - 99.681401000 seconds user - 906.856853000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 8, ..., 1273376, - 1273379, 1273389]), - col_indices=tensor([ 3, 30, 44, ..., 206363, 206408, - 206459]), - values=tensor([-3.7877e-03, -1.5420e-01, 9.5305e-04, ..., - 1.2290e-01, 2.2235e-01, -1.0000e+00]), - size=(206500, 206500), nnz=1273389, layout=torch.sparse_csr) -tensor([0.2037, 0.6417, 0.9786, ..., 0.8187, 0.4933, 0.1289]) -Matrix: mac_econ_fwd500 -Shape: torch.Size([206500, 206500]) -NNZ: 1273389 -Density: 2.9862143765866013e-05 -Time: 13.596147060394287 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/mac_econ_fwd500.mtx 1000': - - 32,027,284 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 290,368 L1I_CACHE_REFILL:u - 471,338 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 33,366,668 L1D_CACHE:u - - 17.325855116 seconds time elapsed - - 101.368582000 seconds user - 1053.826259000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 8, ..., 1273376, - 1273379, 1273389]), - col_indices=tensor([ 3, 30, 44, ..., 206363, 206408, - 206459]), - values=tensor([-3.7877e-03, -1.5420e-01, 9.5305e-04, ..., - 1.2290e-01, 2.2235e-01, -1.0000e+00]), - size=(206500, 206500), nnz=1273389, layout=torch.sparse_csr) -tensor([0.2072, 0.8681, 0.4768, ..., 0.4873, 0.8997, 0.8601]) -Matrix: mac_econ_fwd500 -Shape: torch.Size([206500, 206500]) -NNZ: 1273389 -Density: 2.9862143765866013e-05 -Time: 14.157796382904053 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/mac_econ_fwd500.mtx 1000': - - 571,063 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 583,554 LL_CACHE_RD:u - 196,434 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 25,171 L2D_TLB_REFILL:u - 329,198 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,814,040 L2D_CACHE:u - - 17.958287837 seconds time elapsed - - 104.145071000 seconds user - 1089.962121000 seconds sys - - - diff --git a/pytorch/output_HPC/altra_10_30_mc2depi_1000.json b/pytorch/output_HPC/altra_10_30_mc2depi_1000.json deleted file mode 100644 index 803b668..0000000 --- a/pytorch/output_HPC/altra_10_30_mc2depi_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [94.16, 91.68, 78.92, 60.88, 46.72, 28.36, 22.08, 21.64, 21.64, 21.64], "matrix": "mc2depi", "shape": [525825, 525825], "nnz": 2100225, "% density": 7.595972132902821e-06, "time_s": 11.03979206085205, "power": [95.44, 94.0, 88.76, 72.12, 59.48, 51.92, 53.88, 68.6, 83.2, 97.76, 98.4, 97.12, 97.12, 95.28, 94.12], "power_after": [21.48, 21.44, 21.28, 21.24, 21.16, 21.08, 21.24, 21.24, 21.24, 21.16], "task clock (msec)": 56.14, "page faults": 3289, "cycles": 47515158, "instructions": 72388154, "branch mispredictions": 327042, "branches": 19309026, "ITLB accesses": 26093030, "ITLB misses": 6189, "DTLB misses": 17253, "DTLB accesses": 35168741, "L1I cache accesses": 30539322, "L1I cache misses": 285404, "L1D cache misses": 465747, "L1D cache accesses": 31932803, "LL cache misses": 530261, "LL cache accesses": 551030, "L2D TLB accesses": 183570, "L2D TLB misses": 23883, "L2D cache misses": 297006, "L2D cache accesses": 1721848, "instructions per cycle": 1.5234749719236964, "branch miss rate": 0.01693726032581861, "ITLB miss rate": 0.0002371897782664566, "DTLB miss rate": 0.0004905776979619486, "L2D TLB miss rate": 0.13010295799967314, "L1I cache miss rate": 0.009345459601231487, "L1D cache miss rate": 0.014585221347465175, "L2D cache miss rate": 0.1724925777420539, "LL cache miss rate": 0.9623087672177558} diff --git a/pytorch/output_HPC/altra_10_30_mc2depi_1000.output b/pytorch/output_HPC/altra_10_30_mc2depi_1000.output deleted file mode 100644 index f96b0b7..0000000 --- a/pytorch/output_HPC/altra_10_30_mc2depi_1000.output +++ /dev/null @@ -1,168 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3395280 queued and waiting for resources -srun: job 3395280 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 2, 5, ..., 2100220, - 2100223, 2100225]), - col_indices=tensor([ 0, 1, 1, ..., 525824, 525821, - 525824]), - values=tensor([-2025., 2025., -2026., ..., 2025., 1024., -1024.]), - size=(525825, 525825), nnz=2100225, layout=torch.sparse_csr) -tensor([0.7162, 0.9445, 0.3087, ..., 0.2863, 0.2977, 0.0994]) -Matrix: mc2depi -Shape: torch.Size([525825, 525825]) -NNZ: 2100225 -Density: 7.595972132902821e-06 -Time: 14.228392839431763 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/mc2depi.mtx 1000': - - 56.14 msec task-clock:u # 0.003 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,289 page-faults:u # 58.584 K/sec - 47,515,158 cycles:u # 0.846 GHz (55.54%) - 72,388,154 instructions:u # 1.52 insn per cycle (79.69%) - branches:u - 369,139 branch-misses:u - 32,820,508 L1-dcache-loads:u # 584.601 M/sec - 483,558 L1-dcache-load-misses:u # 1.47% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 31,317,848 L1-icache-loads:u # 557.836 M/sec - 288,398 L1-icache-load-misses:u # 0.92% of all L1-icache accesses - 39,511,659 dTLB-loads:u # 703.784 M/sec (36.64%) - 0 dTLB-load-misses:u (3.47%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 18.186987302 seconds time elapsed - - 124.639912000 seconds user - 1088.590740000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 2, 5, ..., 2100220, - 2100223, 2100225]), - col_indices=tensor([ 0, 1, 1, ..., 525824, 525821, - 525824]), - values=tensor([-2025., 2025., -2026., ..., 2025., 1024., -1024.]), - size=(525825, 525825), nnz=2100225, layout=torch.sparse_csr) -tensor([0.4954, 0.2907, 0.0979, ..., 0.0742, 0.4519, 0.0278]) -Matrix: mc2depi -Shape: torch.Size([525825, 525825]) -NNZ: 2100225 -Density: 7.595972132902821e-06 -Time: 11.948119163513184 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/mc2depi.mtx 1000': - - 327,042 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 19,309,026 BR_RETIRED:u - - 15.715674756 seconds time elapsed - - 115.898749000 seconds user - 910.018676000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 2, 5, ..., 2100220, - 2100223, 2100225]), - col_indices=tensor([ 0, 1, 1, ..., 525824, 525821, - 525824]), - values=tensor([-2025., 2025., -2026., ..., 2025., 1024., -1024.]), - size=(525825, 525825), nnz=2100225, layout=torch.sparse_csr) -tensor([0.1402, 0.9048, 0.8859, ..., 0.9542, 0.3509, 0.0695]) -Matrix: mc2depi -Shape: torch.Size([525825, 525825]) -NNZ: 2100225 -Density: 7.595972132902821e-06 -Time: 14.170094966888428 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/mc2depi.mtx 1000': - - 26,093,030 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 6,189 ITLB_WALK:u - 17,253 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 35,168,741 L1D_TLB:u - - 18.132605509 seconds time elapsed - - 121.020111000 seconds user - 1090.508165000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 2, 5, ..., 2100220, - 2100223, 2100225]), - col_indices=tensor([ 0, 1, 1, ..., 525824, 525821, - 525824]), - values=tensor([-2025., 2025., -2026., ..., 2025., 1024., -1024.]), - size=(525825, 525825), nnz=2100225, layout=torch.sparse_csr) -tensor([0.1192, 0.6084, 0.4643, ..., 0.3445, 0.4658, 0.7085]) -Matrix: mc2depi -Shape: torch.Size([525825, 525825]) -NNZ: 2100225 -Density: 7.595972132902821e-06 -Time: 13.925398826599121 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/mc2depi.mtx 1000': - - 30,539,322 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 285,404 L1I_CACHE_REFILL:u - 465,747 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 31,932,803 L1D_CACHE:u - - 17.812911214 seconds time elapsed - - 119.918777000 seconds user - 1067.928403000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 2, 5, ..., 2100220, - 2100223, 2100225]), - col_indices=tensor([ 0, 1, 1, ..., 525824, 525821, - 525824]), - values=tensor([-2025., 2025., -2026., ..., 2025., 1024., -1024.]), - size=(525825, 525825), nnz=2100225, layout=torch.sparse_csr) -tensor([0.2075, 0.7442, 0.4477, ..., 0.0794, 0.0859, 0.8652]) -Matrix: mc2depi -Shape: torch.Size([525825, 525825]) -NNZ: 2100225 -Density: 7.595972132902821e-06 -Time: 12.866743564605713 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/mc2depi.mtx 1000': - - 530,261 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 551,030 LL_CACHE_RD:u - 183,570 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 23,883 L2D_TLB_REFILL:u - 297,006 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,721,848 L2D_CACHE:u - - 16.812811712 seconds time elapsed - - 117.780323000 seconds user - 986.834040000 seconds sys - - - diff --git a/pytorch/output_HPC/altra_10_30_p2p-Gnutella04_1000.json b/pytorch/output_HPC/altra_10_30_p2p-Gnutella04_1000.json deleted file mode 100644 index 670a9f3..0000000 --- a/pytorch/output_HPC/altra_10_30_p2p-Gnutella04_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [30.08, 25.12, 24.68, 23.68, 22.84, 21.96, 21.08, 20.96, 20.8, 20.96], "matrix": "p2p-Gnutella04", "shape": [10879, 10879], "nnz": 39994, "% density": 0.0003379223282393842, "time_s": 0.9992897510528564, "power": [29.48, 30.52, 31.88, 31.24, 34.32], "power_after": [20.4, 20.6, 20.64, 20.76, 20.92, 20.84, 20.88, 20.88, 20.88, 20.84], "task clock (msec)": 52.68, "page faults": 3272, "cycles": 63019732, "instructions": 73518898, "branch mispredictions": 333423, "branches": 19435905, "ITLB accesses": 27447537, "ITLB misses": 6417, "DTLB misses": 18300, "DTLB accesses": 37569384, "L1I cache accesses": 30830481, "L1I cache misses": 290545, "L1D cache misses": 473875, "L1D cache accesses": 32284772, "LL cache misses": 529403, "LL cache accesses": 549794, "L2D TLB accesses": 198306, "L2D TLB misses": 24497, "L2D cache misses": 298519, "L2D cache accesses": 1772795, "instructions per cycle": 1.1666012480027683, "branch miss rate": 0.017155002558409294, "ITLB miss rate": 0.00023379146915805232, "DTLB miss rate": 0.000487098750408045, "L2D TLB miss rate": 0.12353131019737174, "L1I cache miss rate": 0.009423952873132274, "L1D cache miss rate": 0.014677972636758903, "L2D cache miss rate": 0.16838890001381998, "LL cache miss rate": 0.9629115632400499} diff --git a/pytorch/output_HPC/altra_10_30_p2p-Gnutella04_1000.output b/pytorch/output_HPC/altra_10_30_p2p-Gnutella04_1000.output deleted file mode 100644 index 0cece88..0000000 --- a/pytorch/output_HPC/altra_10_30_p2p-Gnutella04_1000.output +++ /dev/null @@ -1,158 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3395271 queued and waiting for resources -srun: job 3395271 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 10, 20, ..., 39994, 39994, 39994]), - col_indices=tensor([ 1, 2, 3, ..., 9711, 10875, 10876]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(10879, 10879), - nnz=39994, layout=torch.sparse_csr) -tensor([0.3559, 0.4732, 0.3024, ..., 0.9176, 0.7712, 0.4949]) -Matrix: p2p-Gnutella04 -Shape: torch.Size([10879, 10879]) -NNZ: 39994 -Density: 0.0003379223282393842 -Time: 1.0082497596740723 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella04.mtx 1000': - - 52.68 msec task-clock:u # 0.012 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,272 page-faults:u # 62.105 K/sec - 63,019,732 cycles:u # 1.196 GHz (70.67%) - 73,518,898 instructions:u # 1.17 insn per cycle (85.80%) - branches:u - 359,236 branch-misses:u (99.44%) - 31,459,751 L1-dcache-loads:u # 597.131 M/sec - 460,969 L1-dcache-load-misses:u # 1.47% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 29,975,208 L1-icache-loads:u # 568.954 M/sec - 281,710 L1-icache-load-misses:u # 0.94% of all L1-icache accesses - 59,589,523 dTLB-loads:u # 1.131 G/sec (17.10%) - 0 dTLB-load-misses:u (1.27%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 4.456867719 seconds time elapsed - - 16.389568000 seconds user - 29.247355000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 10, 20, ..., 39994, 39994, 39994]), - col_indices=tensor([ 1, 2, 3, ..., 9711, 10875, 10876]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(10879, 10879), - nnz=39994, layout=torch.sparse_csr) -tensor([0.0123, 0.4107, 0.7785, ..., 0.7964, 0.7541, 0.4153]) -Matrix: p2p-Gnutella04 -Shape: torch.Size([10879, 10879]) -NNZ: 39994 -Density: 0.0003379223282393842 -Time: 1.030029058456421 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella04.mtx 1000': - - 333,423 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 19,435,905 BR_RETIRED:u - - 4.359656946 seconds time elapsed - - 16.490532000 seconds user - 28.366462000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 10, 20, ..., 39994, 39994, 39994]), - col_indices=tensor([ 1, 2, 3, ..., 9711, 10875, 10876]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(10879, 10879), - nnz=39994, layout=torch.sparse_csr) -tensor([0.1898, 0.0740, 0.4564, ..., 0.7987, 0.1017, 0.5949]) -Matrix: p2p-Gnutella04 -Shape: torch.Size([10879, 10879]) -NNZ: 39994 -Density: 0.0003379223282393842 -Time: 1.004878044128418 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella04.mtx 1000': - - 27,447,537 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 6,417 ITLB_WALK:u - 18,300 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 37,569,384 L1D_TLB:u - - 4.355627133 seconds time elapsed - - 15.883078000 seconds user - 27.120829000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 10, 20, ..., 39994, 39994, 39994]), - col_indices=tensor([ 1, 2, 3, ..., 9711, 10875, 10876]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(10879, 10879), - nnz=39994, layout=torch.sparse_csr) -tensor([0.1682, 0.9350, 0.9210, ..., 0.3758, 0.2263, 0.1068]) -Matrix: p2p-Gnutella04 -Shape: torch.Size([10879, 10879]) -NNZ: 39994 -Density: 0.0003379223282393842 -Time: 1.0207850933074951 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella04.mtx 1000': - - 30,830,481 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 290,545 L1I_CACHE_REFILL:u - 473,875 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 32,284,772 L1D_CACHE:u - - 4.427088851 seconds time elapsed - - 15.711555000 seconds user - 29.627091000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 10, 20, ..., 39994, 39994, 39994]), - col_indices=tensor([ 1, 2, 3, ..., 9711, 10875, 10876]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(10879, 10879), - nnz=39994, layout=torch.sparse_csr) -tensor([0.9351, 0.3836, 0.0822, ..., 0.9798, 0.3726, 0.7394]) -Matrix: p2p-Gnutella04 -Shape: torch.Size([10879, 10879]) -NNZ: 39994 -Density: 0.0003379223282393842 -Time: 1.041510820388794 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella04.mtx 1000': - - 529,403 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 549,794 LL_CACHE_RD:u - 198,306 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 24,497 L2D_TLB_REFILL:u - 298,519 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,772,795 L2D_CACHE:u - - 4.454107604 seconds time elapsed - - 16.577921000 seconds user - 29.390427000 seconds sys - - - diff --git a/pytorch/output_HPC/altra_10_30_p2p-Gnutella24_1000.json b/pytorch/output_HPC/altra_10_30_p2p-Gnutella24_1000.json deleted file mode 100644 index 141a46e..0000000 --- a/pytorch/output_HPC/altra_10_30_p2p-Gnutella24_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [30.72, 30.6, 28.68, 26.48, 22.44, 21.4, 21.28, 21.08, 21.32, 21.6], "matrix": "p2p-Gnutella24", "shape": [26518, 26518], "nnz": 65369, "% density": 9.295875717624285e-05, "time_s": 1.718301773071289, "power": [31.52, 32.48, 33.64, 33.88, 33.44, 31.52], "power_after": [20.96, 20.84, 20.92, 20.8, 20.76, 20.76, 20.76, 20.68, 20.72, 20.92], "task clock (msec)": 67.08, "page faults": 3303, "cycles": 61261862, "instructions": 83757591, "branch mispredictions": 329248, "branches": 19953212, "ITLB accesses": 27084694, "ITLB misses": 7107, "DTLB misses": 17529, "DTLB accesses": 36684333, "L1I cache accesses": 32158234, "L1I cache misses": 286484, "L1D cache misses": 474161, "L1D cache accesses": 33730073, "LL cache misses": 550064, "LL cache accesses": 565245, "L2D TLB accesses": 191046, "L2D TLB misses": 23775, "L2D cache misses": 307419, "L2D cache accesses": 1772169, "instructions per cycle": 1.3672060930828385, "branch miss rate": 0.016501002445120115, "ITLB miss rate": 0.0002623991247602797, "DTLB miss rate": 0.0004778334118818516, "L2D TLB miss rate": 0.12444646838981188, "L1I cache miss rate": 0.008908573773049851, "L1D cache miss rate": 0.014057514788064645, "L2D cache miss rate": 0.1734704760099065, "LL cache miss rate": 0.973142619572044} diff --git a/pytorch/output_HPC/altra_10_30_p2p-Gnutella24_1000.output b/pytorch/output_HPC/altra_10_30_p2p-Gnutella24_1000.output deleted file mode 100644 index 94adedd..0000000 --- a/pytorch/output_HPC/altra_10_30_p2p-Gnutella24_1000.output +++ /dev/null @@ -1,158 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3395289 queued and waiting for resources -srun: job 3395289 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 9, 9, ..., 65369, 65369, 65369]), - col_indices=tensor([ 1, 2, 3, ..., 15065, 9401, 26517]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(26518, 26518), - nnz=65369, layout=torch.sparse_csr) -tensor([0.3210, 0.3418, 0.9584, ..., 0.8929, 0.9807, 0.5532]) -Matrix: p2p-Gnutella24 -Shape: torch.Size([26518, 26518]) -NNZ: 65369 -Density: 9.295875717624285e-05 -Time: 1.6565663814544678 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella24.mtx 1000': - - 67.08 msec task-clock:u # 0.013 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,303 page-faults:u # 49.241 K/sec - 61,261,862 cycles:u # 0.913 GHz (49.19%) - 83,757,591 instructions:u # 1.37 insn per cycle (88.30%) - branches:u - 364,692 branch-misses:u - 31,954,743 L1-dcache-loads:u # 476.379 M/sec - 490,953 L1-dcache-load-misses:u # 1.54% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 30,490,915 L1-icache-loads:u # 454.556 M/sec - 291,964 L1-icache-load-misses:u # 0.96% of all L1-icache accesses - 32,131,046 dTLB-loads:u # 479.007 M/sec (19.20%) - dTLB-load-misses:u (0.00%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 5.107407925 seconds time elapsed - - 16.045361000 seconds user - 30.574855000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 9, 9, ..., 65369, 65369, 65369]), - col_indices=tensor([ 1, 2, 3, ..., 15065, 9401, 26517]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(26518, 26518), - nnz=65369, layout=torch.sparse_csr) -tensor([0.4851, 0.2524, 0.2134, ..., 0.5976, 0.0089, 0.2284]) -Matrix: p2p-Gnutella24 -Shape: torch.Size([26518, 26518]) -NNZ: 65369 -Density: 9.295875717624285e-05 -Time: 1.6902527809143066 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella24.mtx 1000': - - 329,248 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 19,953,212 BR_RETIRED:u - - 4.990707186 seconds time elapsed - - 16.713526000 seconds user - 27.761595000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 9, 9, ..., 65369, 65369, 65369]), - col_indices=tensor([ 1, 2, 3, ..., 15065, 9401, 26517]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(26518, 26518), - nnz=65369, layout=torch.sparse_csr) -tensor([0.1844, 0.9003, 0.0155, ..., 0.5184, 0.1445, 0.3588]) -Matrix: p2p-Gnutella24 -Shape: torch.Size([26518, 26518]) -NNZ: 65369 -Density: 9.295875717624285e-05 -Time: 1.6478993892669678 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella24.mtx 1000': - - 27,084,694 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 7,107 ITLB_WALK:u - 17,529 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 36,684,333 L1D_TLB:u - - 5.010572757 seconds time elapsed - - 16.570396000 seconds user - 27.387405000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 9, 9, ..., 65369, 65369, 65369]), - col_indices=tensor([ 1, 2, 3, ..., 15065, 9401, 26517]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(26518, 26518), - nnz=65369, layout=torch.sparse_csr) -tensor([0.2313, 0.8375, 0.3065, ..., 0.2374, 0.2281, 0.2100]) -Matrix: p2p-Gnutella24 -Shape: torch.Size([26518, 26518]) -NNZ: 65369 -Density: 9.295875717624285e-05 -Time: 1.637598991394043 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella24.mtx 1000': - - 32,158,234 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 286,484 L1I_CACHE_REFILL:u - 474,161 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 33,730,073 L1D_CACHE:u - - 4.963121627 seconds time elapsed - - 16.730431000 seconds user - 29.869416000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 9, 9, ..., 65369, 65369, 65369]), - col_indices=tensor([ 1, 2, 3, ..., 15065, 9401, 26517]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(26518, 26518), - nnz=65369, layout=torch.sparse_csr) -tensor([0.5006, 0.8470, 0.3527, ..., 0.3901, 0.3581, 0.1154]) -Matrix: p2p-Gnutella24 -Shape: torch.Size([26518, 26518]) -NNZ: 65369 -Density: 9.295875717624285e-05 -Time: 1.6584653854370117 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella24.mtx 1000': - - 550,064 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 565,245 LL_CACHE_RD:u - 191,046 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 23,775 L2D_TLB_REFILL:u - 307,419 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,772,169 L2D_CACHE:u - - 5.019317303 seconds time elapsed - - 16.518292000 seconds user - 30.069880000 seconds sys - - - diff --git a/pytorch/output_HPC/altra_10_30_p2p-Gnutella25_1000.json b/pytorch/output_HPC/altra_10_30_p2p-Gnutella25_1000.json deleted file mode 100644 index 69eef6f..0000000 --- a/pytorch/output_HPC/altra_10_30_p2p-Gnutella25_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [86.48, 72.16, 59.36, 41.84, 28.44, 22.96, 22.92, 22.92, 23.04, 23.24], "matrix": "p2p-Gnutella25", "shape": [22687, 22687], "nnz": 54705, "% density": 0.00010628522108964806, "time_s": 1.431199073791504, "power": [35.16, 36.2, 36.72, 37.52, 37.52], "power_after": [21.32, 21.2, 21.2, 21.28, 21.52, 21.44, 21.92, 21.68, 21.6, 21.36], "task clock (msec)": 59.85, "page faults": 3318, "cycles": 76505130, "instructions": 72343215, "branch mispredictions": 322338, "branches": 19784096, "ITLB accesses": 27270404, "ITLB misses": 6607, "DTLB misses": 17981, "DTLB accesses": 36751047, "L1I cache accesses": 30620441, "L1I cache misses": 302139, "L1D cache misses": 471011, "L1D cache accesses": 32141810, "LL cache misses": 531907, "LL cache accesses": 545159, "L2D TLB accesses": 188244, "L2D TLB misses": 23034, "L2D cache misses": 293848, "L2D cache accesses": 1757551, "instructions per cycle": 0.945599530384433, "branch miss rate": 0.016292783860329025, "ITLB miss rate": 0.00024227730546272803, "DTLB miss rate": 0.0004892649725054092, "L2D TLB miss rate": 0.12236246573595971, "L1I cache miss rate": 0.009867232153841285, "L1D cache miss rate": 0.014654152955294054, "L2D cache miss rate": 0.1671917344077071, "LL cache miss rate": 0.9756914955086498} diff --git a/pytorch/output_HPC/altra_10_30_p2p-Gnutella25_1000.output b/pytorch/output_HPC/altra_10_30_p2p-Gnutella25_1000.output deleted file mode 100644 index 04422e4..0000000 --- a/pytorch/output_HPC/altra_10_30_p2p-Gnutella25_1000.output +++ /dev/null @@ -1,158 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3395288 queued and waiting for resources -srun: job 3395288 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 9, 9, ..., 54704, 54704, 54705]), - col_indices=tensor([ 1, 2, 3, ..., 17949, 22685, 144]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(22687, 22687), - nnz=54705, layout=torch.sparse_csr) -tensor([0.9962, 0.2550, 0.9564, ..., 0.7113, 0.6635, 0.3831]) -Matrix: p2p-Gnutella25 -Shape: torch.Size([22687, 22687]) -NNZ: 54705 -Density: 0.00010628522108964806 -Time: 1.4832944869995117 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella25.mtx 1000': - - 59.85 msec task-clock:u # 0.012 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,318 page-faults:u # 55.439 K/sec - 76,505,130 cycles:u # 1.278 GHz (43.11%) - 72,343,215 instructions:u # 0.95 insn per cycle (62.06%) - branches:u - 371,337 branch-misses:u (77.63%) - 33,969,604 L1-dcache-loads:u # 567.579 M/sec (88.85%) - 472,023 L1-dcache-load-misses:u # 1.39% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 31,728,689 L1-icache-loads:u # 530.137 M/sec - 299,356 L1-icache-load-misses:u # 0.94% of all L1-icache accesses - 50,921,898 dTLB-loads:u # 850.825 M/sec (39.93%) - 90,542 dTLB-load-misses:u # 0.18% of all dTLB cache accesses (36.53%) - 11,563,883 iTLB-loads:u # 193.214 M/sec (20.26%) - iTLB-load-misses:u (0.00%) - - 4.953668960 seconds time elapsed - - 16.652653000 seconds user - 30.408692000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 9, 9, ..., 54704, 54704, 54705]), - col_indices=tensor([ 1, 2, 3, ..., 17949, 22685, 144]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(22687, 22687), - nnz=54705, layout=torch.sparse_csr) -tensor([0.9968, 0.7101, 0.9319, ..., 0.2871, 0.7386, 0.8934]) -Matrix: p2p-Gnutella25 -Shape: torch.Size([22687, 22687]) -NNZ: 54705 -Density: 0.00010628522108964806 -Time: 1.3799591064453125 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella25.mtx 1000': - - 322,338 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 19,784,096 BR_RETIRED:u - - 4.633544255 seconds time elapsed - - 16.572749000 seconds user - 26.228349000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 9, 9, ..., 54704, 54704, 54705]), - col_indices=tensor([ 1, 2, 3, ..., 17949, 22685, 144]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(22687, 22687), - nnz=54705, layout=torch.sparse_csr) -tensor([0.3551, 0.8297, 0.9950, ..., 0.9625, 0.7129, 0.2173]) -Matrix: p2p-Gnutella25 -Shape: torch.Size([22687, 22687]) -NNZ: 54705 -Density: 0.00010628522108964806 -Time: 1.400240182876587 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella25.mtx 1000': - - 27,270,404 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 6,607 ITLB_WALK:u - 17,981 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 36,751,047 L1D_TLB:u - - 4.696092090 seconds time elapsed - - 15.781810000 seconds user - 28.383624000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 9, 9, ..., 54704, 54704, 54705]), - col_indices=tensor([ 1, 2, 3, ..., 17949, 22685, 144]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(22687, 22687), - nnz=54705, layout=torch.sparse_csr) -tensor([0.3600, 0.0388, 0.5262, ..., 0.5849, 0.3707, 0.1514]) -Matrix: p2p-Gnutella25 -Shape: torch.Size([22687, 22687]) -NNZ: 54705 -Density: 0.00010628522108964806 -Time: 1.4545772075653076 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella25.mtx 1000': - - 30,620,441 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 302,139 L1I_CACHE_REFILL:u - 471,011 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 32,141,810 L1D_CACHE:u - - 4.897499310 seconds time elapsed - - 16.207163000 seconds user - 32.246890000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 9, 9, ..., 54704, 54704, 54705]), - col_indices=tensor([ 1, 2, 3, ..., 17949, 22685, 144]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(22687, 22687), - nnz=54705, layout=torch.sparse_csr) -tensor([0.1220, 0.8435, 0.7035, ..., 0.2109, 0.0289, 0.0715]) -Matrix: p2p-Gnutella25 -Shape: torch.Size([22687, 22687]) -NNZ: 54705 -Density: 0.00010628522108964806 -Time: 1.4200170040130615 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella25.mtx 1000': - - 531,907 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 545,159 LL_CACHE_RD:u - 188,244 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 23,034 L2D_TLB_REFILL:u - 293,848 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,757,551 L2D_CACHE:u - - 4.683262937 seconds time elapsed - - 16.111909000 seconds user - 29.660483000 seconds sys - - - diff --git a/pytorch/output_HPC/altra_10_30_p2p-Gnutella30_1000.json b/pytorch/output_HPC/altra_10_30_p2p-Gnutella30_1000.json deleted file mode 100644 index 2cd2684..0000000 --- a/pytorch/output_HPC/altra_10_30_p2p-Gnutella30_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [16.44, 16.44, 16.44, 16.84, 16.72, 16.6, 16.72, 16.84, 16.68, 16.84], "matrix": "p2p-Gnutella30", "shape": [36682, 36682], "nnz": 88328, "% density": 6.564359899804003e-05, "time_s": 2.896674871444702, "power": [56.32, 68.24, 71.76, 59.48, 47.6, 48.76, 52.6], "power_after": [16.92, 17.0, 16.96, 16.8, 16.48, 16.52, 16.52, 16.52, 16.24, 16.36], "task clock (msec)": 56.47, "page faults": 3222, "cycles": 69105836, "instructions": 89065155, "branch mispredictions": 333669, "branches": 20078755, "ITLB accesses": 26015038, "ITLB misses": 5212, "DTLB misses": 17039, "DTLB accesses": 35296010, "L1I cache accesses": 31837486, "L1I cache misses": 293353, "L1D cache misses": 462358, "L1D cache accesses": 33478540, "LL cache misses": 546516, "LL cache accesses": 559865, "L2D TLB accesses": 190400, "L2D TLB misses": 23787, "L2D cache misses": 307032, "L2D cache accesses": 1768186, "instructions per cycle": 1.288822480926213, "branch miss rate": 0.016618012421586895, "ITLB miss rate": 0.00020034566161310238, "DTLB miss rate": 0.00048274578344691083, "L2D TLB miss rate": 0.12493172268907562, "L1I cache miss rate": 0.009214075508348869, "L1D cache miss rate": 0.013810578358554464, "L2D cache miss rate": 0.17364236567872385, "LL cache miss rate": 0.9761567520741607} diff --git a/pytorch/output_HPC/altra_10_30_p2p-Gnutella30_1000.output b/pytorch/output_HPC/altra_10_30_p2p-Gnutella30_1000.output deleted file mode 100644 index f3fd1c7..0000000 --- a/pytorch/output_HPC/altra_10_30_p2p-Gnutella30_1000.output +++ /dev/null @@ -1,158 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3395282 queued and waiting for resources -srun: job 3395282 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 10, 10, ..., 88328, 88328, 88328]), - col_indices=tensor([ 1, 2, 3, ..., 36675, 36676, 36677]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(36682, 36682), - nnz=88328, layout=torch.sparse_csr) -tensor([0.0302, 0.1334, 0.4142, ..., 0.9516, 0.6030, 0.3883]) -Matrix: p2p-Gnutella30 -Shape: torch.Size([36682, 36682]) -NNZ: 88328 -Density: 6.564359899804003e-05 -Time: 2.790724277496338 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella30.mtx 1000': - - 56.47 msec task-clock:u # 0.009 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,222 page-faults:u # 57.061 K/sec - 69,105,836 cycles:u # 1.224 GHz (53.55%) - 89,065,155 instructions:u # 1.29 insn per cycle (92.79%) - branches:u - 367,525 branch-misses:u - 32,122,654 L1-dcache-loads:u # 568.886 M/sec - 467,921 L1-dcache-load-misses:u # 1.46% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 30,765,438 L1-icache-loads:u # 544.850 M/sec - 289,327 L1-icache-load-misses:u # 0.94% of all L1-icache accesses - 24,642,710 dTLB-loads:u # 436.418 M/sec (11.11%) - dTLB-load-misses:u (0.00%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 6.334250152 seconds time elapsed - - 32.099712000 seconds user - 240.206702000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 10, 10, ..., 88328, 88328, 88328]), - col_indices=tensor([ 1, 2, 3, ..., 36675, 36676, 36677]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(36682, 36682), - nnz=88328, layout=torch.sparse_csr) -tensor([0.6147, 0.4171, 0.2258, ..., 0.0253, 0.8932, 0.8040]) -Matrix: p2p-Gnutella30 -Shape: torch.Size([36682, 36682]) -NNZ: 88328 -Density: 6.564359899804003e-05 -Time: 2.092158079147339 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella30.mtx 1000': - - 333,669 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 20,078,755 BR_RETIRED:u - - 5.557038624 seconds time elapsed - - 29.074016000 seconds user - 186.372846000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 10, 10, ..., 88328, 88328, 88328]), - col_indices=tensor([ 1, 2, 3, ..., 36675, 36676, 36677]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(36682, 36682), - nnz=88328, layout=torch.sparse_csr) -tensor([0.0146, 0.2151, 0.1948, ..., 0.7633, 0.4329, 0.7106]) -Matrix: p2p-Gnutella30 -Shape: torch.Size([36682, 36682]) -NNZ: 88328 -Density: 6.564359899804003e-05 -Time: 3.1269772052764893 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella30.mtx 1000': - - 26,015,038 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 5,212 ITLB_WALK:u - 17,039 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 35,296,010 L1D_TLB:u - - 6.550798214 seconds time elapsed - - 36.334689000 seconds user - 263.614426000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 10, 10, ..., 88328, 88328, 88328]), - col_indices=tensor([ 1, 2, 3, ..., 36675, 36676, 36677]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(36682, 36682), - nnz=88328, layout=torch.sparse_csr) -tensor([0.1810, 0.5208, 0.0542, ..., 0.6108, 0.4905, 0.8918]) -Matrix: p2p-Gnutella30 -Shape: torch.Size([36682, 36682]) -NNZ: 88328 -Density: 6.564359899804003e-05 -Time: 1.9065814018249512 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella30.mtx 1000': - - 31,837,486 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 293,353 L1I_CACHE_REFILL:u - 462,358 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 33,478,540 L1D_CACHE:u - - 5.319975004 seconds time elapsed - - 26.918342000 seconds user - 175.603919000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 10, 10, ..., 88328, 88328, 88328]), - col_indices=tensor([ 1, 2, 3, ..., 36675, 36676, 36677]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(36682, 36682), - nnz=88328, layout=torch.sparse_csr) -tensor([0.8456, 0.8302, 0.2078, ..., 0.8155, 0.5148, 0.5853]) -Matrix: p2p-Gnutella30 -Shape: torch.Size([36682, 36682]) -NNZ: 88328 -Density: 6.564359899804003e-05 -Time: 3.8523874282836914 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/p2p-Gnutella30.mtx 1000': - - 546,516 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 559,865 LL_CACHE_RD:u - 190,400 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 23,787 L2D_TLB_REFILL:u - 307,032 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,768,186 L2D_CACHE:u - - 7.266305868 seconds time elapsed - - 37.085321000 seconds user - 320.780766000 seconds sys - - - diff --git a/pytorch/output_HPC/altra_10_30_ri2010_1000.json b/pytorch/output_HPC/altra_10_30_ri2010_1000.json deleted file mode 100644 index 42c7e19..0000000 --- a/pytorch/output_HPC/altra_10_30_ri2010_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [16.6, 16.64, 17.04, 17.08, 16.92, 17.24, 16.88, 16.36, 16.4, 16.4], "matrix": "ri2010", "shape": [25181, 25181], "nnz": 125750, "% density": 0.00019831796057928155, "time_s": 2.970583200454712, "power": [23.04, 23.28, 23.76, 24.12, 21.4, 26.28, 26.36], "power_after": [16.16, 16.16, 16.52, 16.48, 16.52, 16.44, 16.36, 16.48, 16.76, 16.6], "task clock (msec)": 52.61, "page faults": 3292, "cycles": 42915672, "instructions": 71002596, "branch mispredictions": 344300, "branches": 20224759, "ITLB accesses": 26039851, "ITLB misses": 5035, "DTLB misses": 16402, "DTLB accesses": 34820806, "L1I cache accesses": 31878105, "L1I cache misses": 299057, "L1D cache misses": 471869, "L1D cache accesses": 33450518, "LL cache misses": 530093, "LL cache accesses": 551126, "L2D TLB accesses": 188315, "L2D TLB misses": 22856, "L2D cache misses": 299885, "L2D cache accesses": 1763155, "instructions per cycle": 1.6544677664607, "branch miss rate": 0.01702368863826758, "ITLB miss rate": 0.00019335748119296073, "DTLB miss rate": 0.0004710402165877493, "L2D TLB miss rate": 0.12137110692191275, "L1I cache miss rate": 0.009381266546427399, "L1D cache miss rate": 0.014106478111938357, "L2D cache miss rate": 0.1700843090936418, "LL cache miss rate": 0.9618363132931489} diff --git a/pytorch/output_HPC/altra_10_30_ri2010_1000.output b/pytorch/output_HPC/altra_10_30_ri2010_1000.output deleted file mode 100644 index 1f5f007..0000000 --- a/pytorch/output_HPC/altra_10_30_ri2010_1000.output +++ /dev/null @@ -1,163 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3395268 queued and waiting for resources -srun: job 3395268 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 8, ..., 125742, 125747, - 125750]), - col_indices=tensor([ 25, 56, 662, ..., 21738, 22279, 23882]), - values=tensor([17171., 37318., 5284., ..., 25993., 24918., 803.]), - size=(25181, 25181), nnz=125750, layout=torch.sparse_csr) -tensor([0.4029, 0.5373, 0.8376, ..., 0.9299, 0.3127, 0.4778]) -Matrix: ri2010 -Shape: torch.Size([25181, 25181]) -NNZ: 125750 -Density: 0.00019831796057928155 -Time: 2.9858975410461426 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ri2010.mtx 1000': - - 52.61 msec task-clock:u # 0.008 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,292 page-faults:u # 62.576 K/sec - 42,915,672 cycles:u # 0.816 GHz (55.04%) - 71,002,596 instructions:u # 1.65 insn per cycle (81.89%) - branches:u - 369,793 branch-misses:u - 33,163,106 L1-dcache-loads:u # 630.381 M/sec - 471,533 L1-dcache-load-misses:u # 1.42% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 31,640,002 L1-icache-loads:u # 601.429 M/sec - 297,919 L1-icache-load-misses:u # 0.94% of all L1-icache accesses - 48,642,108 dTLB-loads:u # 924.614 M/sec (29.77%) - 0 dTLB-load-misses:u (5.06%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 6.215745697 seconds time elapsed - - 17.600216000 seconds user - 30.777524000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 8, ..., 125742, 125747, - 125750]), - col_indices=tensor([ 25, 56, 662, ..., 21738, 22279, 23882]), - values=tensor([17171., 37318., 5284., ..., 25993., 24918., 803.]), - size=(25181, 25181), nnz=125750, layout=torch.sparse_csr) -tensor([0.8706, 0.3724, 0.8779, ..., 0.4299, 0.0920, 0.4238]) -Matrix: ri2010 -Shape: torch.Size([25181, 25181]) -NNZ: 125750 -Density: 0.00019831796057928155 -Time: 2.9231789112091064 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ri2010.mtx 1000': - - 344,300 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 20,224,759 BR_RETIRED:u - - 6.297708483 seconds time elapsed - - 17.546068000 seconds user - 26.920857000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 8, ..., 125742, 125747, - 125750]), - col_indices=tensor([ 25, 56, 662, ..., 21738, 22279, 23882]), - values=tensor([17171., 37318., 5284., ..., 25993., 24918., 803.]), - size=(25181, 25181), nnz=125750, layout=torch.sparse_csr) -tensor([0.2988, 0.0160, 0.4360, ..., 0.7543, 0.0919, 0.2321]) -Matrix: ri2010 -Shape: torch.Size([25181, 25181]) -NNZ: 125750 -Density: 0.00019831796057928155 -Time: 2.9701316356658936 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ri2010.mtx 1000': - - 26,039,851 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 5,035 ITLB_WALK:u - 16,402 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 34,820,806 L1D_TLB:u - - 6.227977259 seconds time elapsed - - 17.937381000 seconds user - 30.196552000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 8, ..., 125742, 125747, - 125750]), - col_indices=tensor([ 25, 56, 662, ..., 21738, 22279, 23882]), - values=tensor([17171., 37318., 5284., ..., 25993., 24918., 803.]), - size=(25181, 25181), nnz=125750, layout=torch.sparse_csr) -tensor([0.5797, 0.8992, 0.8317, ..., 0.0283, 0.7124, 0.2690]) -Matrix: ri2010 -Shape: torch.Size([25181, 25181]) -NNZ: 125750 -Density: 0.00019831796057928155 -Time: 2.968733072280884 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ri2010.mtx 1000': - - 31,878,105 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 299,057 L1I_CACHE_REFILL:u - 471,869 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 33,450,518 L1D_CACHE:u - - 6.278062824 seconds time elapsed - - 17.822878000 seconds user - 27.932170000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 8, ..., 125742, 125747, - 125750]), - col_indices=tensor([ 25, 56, 662, ..., 21738, 22279, 23882]), - values=tensor([17171., 37318., 5284., ..., 25993., 24918., 803.]), - size=(25181, 25181), nnz=125750, layout=torch.sparse_csr) -tensor([0.0630, 0.5194, 0.8720, ..., 0.9537, 0.3959, 0.5550]) -Matrix: ri2010 -Shape: torch.Size([25181, 25181]) -NNZ: 125750 -Density: 0.00019831796057928155 -Time: 2.9069995880126953 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ri2010.mtx 1000': - - 530,093 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 551,126 LL_CACHE_RD:u - 188,315 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 22,856 L2D_TLB_REFILL:u - 299,885 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,763,155 L2D_CACHE:u - - 6.075529293 seconds time elapsed - - 17.073983000 seconds user - 27.811966000 seconds sys - - - diff --git a/pytorch/output_HPC/altra_10_30_rma10_1000.json b/pytorch/output_HPC/altra_10_30_rma10_1000.json deleted file mode 100644 index 9987299..0000000 --- a/pytorch/output_HPC/altra_10_30_rma10_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [31.36, 30.64, 31.12, 24.52, 24.16, 23.12, 22.08, 21.28, 21.16, 20.88], "matrix": "rma10", "shape": [46835, 46835], "nnz": 2374001, "% density": 0.0010822805369125833, "time_s": 68.86891412734985, "power": [81.8, 81.32, 75.08, 63.48, 51.92, 51.96, 51.8, 65.0, 65.0, 75.12, 82.68, 82.32, 82.08, 82.76, 82.8, 83.6, 83.36, 83.08, 82.88, 83.0, 83.32, 83.32, 83.36, 84.64, 84.56, 84.24, 83.52, 83.4, 83.36, 83.36, 83.72, 84.16, 83.24, 82.76, 82.76, 82.96, 82.36, 82.24, 81.64, 81.6, 81.4, 81.6, 81.88, 82.32, 83.04, 83.48, 83.48, 84.32, 84.04, 84.32, 83.16, 82.44, 81.96, 81.4, 81.8, 82.08, 81.8, 81.84, 82.04, 82.04, 82.08, 82.44, 82.6, 82.84, 83.8, 84.24, 84.6, 85.4, 85.6, 86.0, 85.72, 85.36], "power_after": [21.96, 21.88, 21.96, 21.96, 22.0, 21.68, 21.44, 21.16, 21.04, 20.92], "task clock (msec)": 58.3, "page faults": 3281, "cycles": 81319364, "instructions": 90830397, "branch mispredictions": 342237, "branches": 20641135, "ITLB accesses": 27974213, "ITLB misses": 6660, "DTLB misses": 18441, "DTLB accesses": 37780346, "L1I cache accesses": 31166891, "L1I cache misses": 291301, "L1D cache misses": 477186, "L1D cache accesses": 32682323, "LL cache misses": 538552, "LL cache accesses": 552543, "L2D TLB accesses": 202351, "L2D TLB misses": 24178, "L2D cache misses": 298051, "L2D cache accesses": 1775481, "instructions per cycle": 1.1169590185186398, "branch miss rate": 0.01658033824205888, "ITLB miss rate": 0.00023807640272132053, "DTLB miss rate": 0.00048811093471722044, "L2D TLB miss rate": 0.11948544855226809, "L1I cache miss rate": 0.00934648887500521, "L1D cache miss rate": 0.014600736918241704, "L2D cache miss rate": 0.1678705657790762, "LL cache miss rate": 0.9746788937693537} diff --git a/pytorch/output_HPC/altra_10_30_rma10_1000.output b/pytorch/output_HPC/altra_10_30_rma10_1000.output deleted file mode 100644 index b659ed0..0000000 --- a/pytorch/output_HPC/altra_10_30_rma10_1000.output +++ /dev/null @@ -1,168 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3395286 queued and waiting for resources -srun: job 3395286 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 17, 34, ..., 2373939, - 2373970, 2374001]), - col_indices=tensor([ 0, 1, 2, ..., 46831, 46833, 46834]), - values=tensor([ 1.2636e+05, -1.6615e+07, -8.2015e+04, ..., - 8.3378e+01, 2.5138e+00, 1.2184e+03]), - size=(46835, 46835), nnz=2374001, layout=torch.sparse_csr) -tensor([0.4937, 0.5946, 0.4240, ..., 0.9888, 0.5278, 0.9155]) -Matrix: rma10 -Shape: torch.Size([46835, 46835]) -NNZ: 2374001 -Density: 0.0010822805369125833 -Time: 52.320035219192505 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/rma10.mtx 1000': - - 58.30 msec task-clock:u # 0.001 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,281 page-faults:u # 56.279 K/sec - 81,319,364 cycles:u # 1.395 GHz (62.38%) - 90,830,397 instructions:u # 1.12 insn per cycle (94.62%) - branches:u - 358,947 branch-misses:u - 32,561,141 L1-dcache-loads:u # 558.523 M/sec - 477,147 L1-dcache-load-misses:u # 1.47% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 31,044,361 L1-icache-loads:u # 532.506 M/sec - 286,125 L1-icache-load-misses:u # 0.92% of all L1-icache accesses - 29,678,379 dTLB-loads:u # 509.075 M/sec (5.72%) - dTLB-load-misses:u (0.00%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 56.145511940 seconds time elapsed - - 269.541895000 seconds user - 3993.928150000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 17, 34, ..., 2373939, - 2373970, 2374001]), - col_indices=tensor([ 0, 1, 2, ..., 46831, 46833, 46834]), - values=tensor([ 1.2636e+05, -1.6615e+07, -8.2015e+04, ..., - 8.3378e+01, 2.5138e+00, 1.2184e+03]), - size=(46835, 46835), nnz=2374001, layout=torch.sparse_csr) -tensor([0.2401, 0.9608, 0.9686, ..., 0.2643, 0.1097, 0.0695]) -Matrix: rma10 -Shape: torch.Size([46835, 46835]) -NNZ: 2374001 -Density: 0.0010822805369125833 -Time: 65.29214668273926 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/rma10.mtx 1000': - - 342,237 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 20,641,135 BR_RETIRED:u - - 69.131216008 seconds time elapsed - - 324.908899000 seconds user - 4969.165543000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 17, 34, ..., 2373939, - 2373970, 2374001]), - col_indices=tensor([ 0, 1, 2, ..., 46831, 46833, 46834]), - values=tensor([ 1.2636e+05, -1.6615e+07, -8.2015e+04, ..., - 8.3378e+01, 2.5138e+00, 1.2184e+03]), - size=(46835, 46835), nnz=2374001, layout=torch.sparse_csr) -tensor([0.5237, 0.3525, 0.2809, ..., 0.8641, 0.3894, 0.4198]) -Matrix: rma10 -Shape: torch.Size([46835, 46835]) -NNZ: 2374001 -Density: 0.0010822805369125833 -Time: 66.05637407302856 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/rma10.mtx 1000': - - 27,974,213 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 6,660 ITLB_WALK:u - 18,441 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 37,780,346 L1D_TLB:u - - 69.880637029 seconds time elapsed - - 320.759259000 seconds user - 5037.255757000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 17, 34, ..., 2373939, - 2373970, 2374001]), - col_indices=tensor([ 0, 1, 2, ..., 46831, 46833, 46834]), - values=tensor([ 1.2636e+05, -1.6615e+07, -8.2015e+04, ..., - 8.3378e+01, 2.5138e+00, 1.2184e+03]), - size=(46835, 46835), nnz=2374001, layout=torch.sparse_csr) -tensor([0.8185, 0.4278, 0.7553, ..., 0.5022, 0.1058, 0.0783]) -Matrix: rma10 -Shape: torch.Size([46835, 46835]) -NNZ: 2374001 -Density: 0.0010822805369125833 -Time: 63.55399775505066 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/rma10.mtx 1000': - - 31,166,891 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 291,301 L1I_CACHE_REFILL:u - 477,186 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 32,682,323 L1D_CACHE:u - - 67.517251505 seconds time elapsed - - 319.301754000 seconds user - 4839.755901000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 17, 34, ..., 2373939, - 2373970, 2374001]), - col_indices=tensor([ 0, 1, 2, ..., 46831, 46833, 46834]), - values=tensor([ 1.2636e+05, -1.6615e+07, -8.2015e+04, ..., - 8.3378e+01, 2.5138e+00, 1.2184e+03]), - size=(46835, 46835), nnz=2374001, layout=torch.sparse_csr) -tensor([0.8358, 0.0086, 0.1779, ..., 0.6354, 0.7134, 0.5745]) -Matrix: rma10 -Shape: torch.Size([46835, 46835]) -NNZ: 2374001 -Density: 0.0010822805369125833 -Time: 63.55393171310425 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/rma10.mtx 1000': - - 538,552 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 552,543 LL_CACHE_RD:u - 202,351 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 24,178 L2D_TLB_REFILL:u - 298,051 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,775,481 L2D_CACHE:u - - 67.538674790 seconds time elapsed - - 321.810383000 seconds user - 4836.154538000 seconds sys - - - diff --git a/pytorch/output_HPC/altra_10_30_tn2010_1000.json b/pytorch/output_HPC/altra_10_30_tn2010_1000.json deleted file mode 100644 index 3379070..0000000 --- a/pytorch/output_HPC/altra_10_30_tn2010_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [37.56, 23.12, 22.32, 22.28, 22.28, 21.96, 21.76, 21.72, 21.88, 21.84], "matrix": "tn2010", "shape": [240116, 240116], "nnz": 1193966, "% density": 2.070855328296721e-05, "time_s": 16.282614707946777, "power": [85.48, 85.84, 79.28, 70.16, 55.52, 49.48, 49.48, 60.48, 76.32, 88.88, 91.0, 91.0, 90.68, 88.32, 86.92, 86.4, 88.08, 86.8, 87.32, 87.8], "power_after": [21.68, 21.48, 21.44, 21.36, 21.52, 21.4, 21.4, 21.32, 21.2, 21.04], "task clock (msec)": 68.11, "page faults": 3486, "cycles": 70427921, "instructions": 85638293, "branch mispredictions": 333780, "branches": 19402540, "ITLB accesses": 26935483, "ITLB misses": 5639, "DTLB misses": 16688, "DTLB accesses": 36421540, "L1I cache accesses": 33029213, "L1I cache misses": 302558, "L1D cache misses": 481598, "L1D cache accesses": 34668833, "LL cache misses": 551659, "LL cache accesses": 564579, "L2D TLB accesses": 188346, "L2D TLB misses": 24479, "L2D cache misses": 311796, "L2D cache accesses": 1767924, "instructions per cycle": 1.215970765344614, "branch miss rate": 0.017202902300420462, "ITLB miss rate": 0.0002093521025778524, "DTLB miss rate": 0.00045819040051573877, "L2D TLB miss rate": 0.12996824992301403, "L1I cache miss rate": 0.00916031514284037, "L1D cache miss rate": 0.013891381922200843, "L2D cache miss rate": 0.17636278482559206, "LL cache miss rate": 0.9771156915152707} diff --git a/pytorch/output_HPC/altra_10_30_tn2010_1000.output b/pytorch/output_HPC/altra_10_30_tn2010_1000.output deleted file mode 100644 index faed868..0000000 --- a/pytorch/output_HPC/altra_10_30_tn2010_1000.output +++ /dev/null @@ -1,173 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3395275 queued and waiting for resources -srun: job 3395275 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 4, 20, ..., 1193961, - 1193963, 1193966]), - col_indices=tensor([ 1152, 1272, 1961, ..., 238254, 239142, - 240113]), - values=tensor([ 5728., 2871., 418449., ..., 10058., 33324., - 34928.]), size=(240116, 240116), nnz=1193966, - layout=torch.sparse_csr) -tensor([0.2511, 0.1104, 0.8257, ..., 0.4006, 0.1534, 0.0009]) -Matrix: tn2010 -Shape: torch.Size([240116, 240116]) -NNZ: 1193966 -Density: 2.070855328296721e-05 -Time: 12.89618182182312 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/tn2010.mtx 1000': - - 68.11 msec task-clock:u # 0.004 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,486 page-faults:u # 51.182 K/sec - 70,427,921 cycles:u # 1.034 GHz (46.81%) - 85,638,293 instructions:u # 1.22 insn per cycle (74.19%) - branches:u - 356,748 branch-misses:u (89.74%) - 34,044,117 L1-dcache-loads:u # 499.843 M/sec - 481,076 L1-dcache-load-misses:u # 1.41% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 32,553,977 L1-icache-loads:u # 477.965 M/sec - 309,127 L1-icache-load-misses:u # 0.95% of all L1-icache accesses - 41,245,978 dTLB-loads:u # 605.583 M/sec (33.60%) - 127,770 dTLB-load-misses:u # 0.31% of all dTLB cache accesses (15.43%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 16.626373547 seconds time elapsed - - 101.073288000 seconds user - 996.348020000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 4, 20, ..., 1193961, - 1193963, 1193966]), - col_indices=tensor([ 1152, 1272, 1961, ..., 238254, 239142, - 240113]), - values=tensor([ 5728., 2871., 418449., ..., 10058., 33324., - 34928.]), size=(240116, 240116), nnz=1193966, - layout=torch.sparse_csr) -tensor([0.0138, 0.1394, 0.6273, ..., 0.8681, 0.0444, 0.2705]) -Matrix: tn2010 -Shape: torch.Size([240116, 240116]) -NNZ: 1193966 -Density: 2.070855328296721e-05 -Time: 14.216531038284302 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/tn2010.mtx 1000': - - 333,780 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 19,402,540 BR_RETIRED:u - - 17.985093703 seconds time elapsed - - 106.904608000 seconds user - 1091.172933000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 4, 20, ..., 1193961, - 1193963, 1193966]), - col_indices=tensor([ 1152, 1272, 1961, ..., 238254, 239142, - 240113]), - values=tensor([ 5728., 2871., 418449., ..., 10058., 33324., - 34928.]), size=(240116, 240116), nnz=1193966, - layout=torch.sparse_csr) -tensor([0.6279, 0.1696, 0.6937, ..., 0.4267, 0.4847, 0.6447]) -Matrix: tn2010 -Shape: torch.Size([240116, 240116]) -NNZ: 1193966 -Density: 2.070855328296721e-05 -Time: 12.462992429733276 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/tn2010.mtx 1000': - - 26,935,483 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 5,639 ITLB_WALK:u - 16,688 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 36,421,540 L1D_TLB:u - - 15.984498303 seconds time elapsed - - 95.195897000 seconds user - 962.237122000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 4, 20, ..., 1193961, - 1193963, 1193966]), - col_indices=tensor([ 1152, 1272, 1961, ..., 238254, 239142, - 240113]), - values=tensor([ 5728., 2871., 418449., ..., 10058., 33324., - 34928.]), size=(240116, 240116), nnz=1193966, - layout=torch.sparse_csr) -tensor([0.4060, 0.4915, 0.8557, ..., 0.9902, 0.0548, 0.2450]) -Matrix: tn2010 -Shape: torch.Size([240116, 240116]) -NNZ: 1193966 -Density: 2.070855328296721e-05 -Time: 9.298198223114014 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/tn2010.mtx 1000': - - 33,029,213 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 302,558 L1I_CACHE_REFILL:u - 481,598 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 34,668,833 L1D_CACHE:u - - 12.985459942 seconds time elapsed - - 78.950722000 seconds user - 727.126874000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 4, 20, ..., 1193961, - 1193963, 1193966]), - col_indices=tensor([ 1152, 1272, 1961, ..., 238254, 239142, - 240113]), - values=tensor([ 5728., 2871., 418449., ..., 10058., 33324., - 34928.]), size=(240116, 240116), nnz=1193966, - layout=torch.sparse_csr) -tensor([0.0166, 0.6910, 0.0311, ..., 0.6156, 0.5689, 0.9849]) -Matrix: tn2010 -Shape: torch.Size([240116, 240116]) -NNZ: 1193966 -Density: 2.070855328296721e-05 -Time: 12.012693405151367 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/tn2010.mtx 1000': - - 551,659 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 564,579 LL_CACHE_RD:u - 188,346 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 24,479 L2D_TLB_REFILL:u - 311,796 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,767,924 L2D_CACHE:u - - 15.749851583 seconds time elapsed - - 98.008506000 seconds user - 926.127594000 seconds sys - - - diff --git a/pytorch/output_HPC/altra_10_30_ut2010_1000.json b/pytorch/output_HPC/altra_10_30_ut2010_1000.json deleted file mode 100644 index 3fd8e6b..0000000 --- a/pytorch/output_HPC/altra_10_30_ut2010_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [93.52, 87.76, 71.6, 58.32, 39.6, 26.24, 26.24, 22.16, 22.24, 22.24], "matrix": "ut2010", "shape": [115406, 115406], "nnz": 572066, "% density": 4.295259032005559e-05, "time_s": 8.478580713272095, "power": [89.68, 88.92, 80.84, 68.96, 56.64, 54.52, 55.88, 70.44, 85.36, 85.36, 98.2, 96.52], "power_after": [21.24, 21.32, 21.16, 21.44, 21.68, 21.76, 21.72, 22.0, 21.72, 21.72], "task clock (msec)": 53.84, "page faults": 3291, "cycles": 66389970, "instructions": 74935543, "branch mispredictions": 330515, "branches": 19475058, "ITLB accesses": 26125490, "ITLB misses": 6431, "DTLB misses": 13728, "DTLB accesses": 35274185, "L1I cache accesses": 30428652, "L1I cache misses": 288897, "L1D cache misses": 475615, "L1D cache accesses": 31855716, "LL cache misses": 553829, "LL cache accesses": 574192, "L2D TLB accesses": 181148, "L2D TLB misses": 23202, "L2D cache misses": 307806, "L2D cache accesses": 1767037, "instructions per cycle": 1.1287178319255153, "branch miss rate": 0.016971194642911976, "ITLB miss rate": 0.00024615806248992844, "DTLB miss rate": 0.0003891797925309968, "L2D TLB miss rate": 0.12808311435952924, "L1I cache miss rate": 0.009494242465949527, "L1D cache miss rate": 0.014930287550278261, "L2D cache miss rate": 0.17419329646181717, "LL cache miss rate": 0.9645362526820297} diff --git a/pytorch/output_HPC/altra_10_30_ut2010_1000.output b/pytorch/output_HPC/altra_10_30_ut2010_1000.output deleted file mode 100644 index d437337..0000000 --- a/pytorch/output_HPC/altra_10_30_ut2010_1000.output +++ /dev/null @@ -1,173 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3395284 queued and waiting for resources -srun: job 3395284 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 9, ..., 572056, 572061, - 572066]), - col_indices=tensor([ 453, 1291, 1979, ..., 113521, 114509, - 114602]), - values=tensor([160642., 31335., 282373., ..., 88393., 99485., - 18651.]), size=(115406, 115406), nnz=572066, - layout=torch.sparse_csr) -tensor([0.1487, 0.4275, 0.9471, ..., 0.3851, 0.0801, 0.4295]) -Matrix: ut2010 -Shape: torch.Size([115406, 115406]) -NNZ: 572066 -Density: 4.295259032005559e-05 -Time: 8.772023677825928 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ut2010.mtx 1000': - - 53.84 msec task-clock:u # 0.004 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,291 page-faults:u # 61.127 K/sec - 66,389,970 cycles:u # 1.233 GHz (67.37%) - 74,935,543 instructions:u # 1.13 insn per cycle (83.30%) - branches:u - 365,846 branch-misses:u - 31,684,169 L1-dcache-loads:u # 588.504 M/sec - 462,583 L1-dcache-load-misses:u # 1.46% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 30,260,337 L1-icache-loads:u # 562.058 M/sec - 288,196 L1-icache-load-misses:u # 0.95% of all L1-icache accesses - 57,721,334 dTLB-loads:u # 1.072 G/sec (18.54%) - dTLB-load-misses:u (0.00%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 12.179628060 seconds time elapsed - - 68.068275000 seconds user - 690.223452000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 9, ..., 572056, 572061, - 572066]), - col_indices=tensor([ 453, 1291, 1979, ..., 113521, 114509, - 114602]), - values=tensor([160642., 31335., 282373., ..., 88393., 99485., - 18651.]), size=(115406, 115406), nnz=572066, - layout=torch.sparse_csr) -tensor([0.9553, 0.9401, 0.7135, ..., 0.8664, 0.5986, 0.8459]) -Matrix: ut2010 -Shape: torch.Size([115406, 115406]) -NNZ: 572066 -Density: 4.295259032005559e-05 -Time: 8.94040060043335 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ut2010.mtx 1000': - - 330,515 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 19,475,058 BR_RETIRED:u - - 12.428594105 seconds time elapsed - - 67.011228000 seconds user - 709.528404000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 9, ..., 572056, 572061, - 572066]), - col_indices=tensor([ 453, 1291, 1979, ..., 113521, 114509, - 114602]), - values=tensor([160642., 31335., 282373., ..., 88393., 99485., - 18651.]), size=(115406, 115406), nnz=572066, - layout=torch.sparse_csr) -tensor([0.6289, 0.8171, 0.1590, ..., 0.7515, 0.5400, 0.3693]) -Matrix: ut2010 -Shape: torch.Size([115406, 115406]) -NNZ: 572066 -Density: 4.295259032005559e-05 -Time: 14.403366804122925 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ut2010.mtx 1000': - - 26,125,490 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 6,431 ITLB_WALK:u - 13,728 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 35,274,185 L1D_TLB:u - - 18.084508405 seconds time elapsed - - 95.162133000 seconds user - 1117.716009000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 9, ..., 572056, 572061, - 572066]), - col_indices=tensor([ 453, 1291, 1979, ..., 113521, 114509, - 114602]), - values=tensor([160642., 31335., 282373., ..., 88393., 99485., - 18651.]), size=(115406, 115406), nnz=572066, - layout=torch.sparse_csr) -tensor([0.8824, 0.0692, 0.7225, ..., 0.8736, 0.6854, 0.7514]) -Matrix: ut2010 -Shape: torch.Size([115406, 115406]) -NNZ: 572066 -Density: 4.295259032005559e-05 -Time: 9.64679503440857 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ut2010.mtx 1000': - - 30,428,652 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 288,897 L1I_CACHE_REFILL:u - 475,615 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 31,855,716 L1D_CACHE:u - - 13.170070008 seconds time elapsed - - 68.362809000 seconds user - 761.360459000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 9, ..., 572056, 572061, - 572066]), - col_indices=tensor([ 453, 1291, 1979, ..., 113521, 114509, - 114602]), - values=tensor([160642., 31335., 282373., ..., 88393., 99485., - 18651.]), size=(115406, 115406), nnz=572066, - layout=torch.sparse_csr) -tensor([0.9552, 0.0509, 0.7738, ..., 0.7722, 0.4417, 0.7772]) -Matrix: ut2010 -Shape: torch.Size([115406, 115406]) -NNZ: 572066 -Density: 4.295259032005559e-05 -Time: 12.372079133987427 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/ut2010.mtx 1000': - - 553,829 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 574,192 LL_CACHE_RD:u - 181,148 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 23,202 L2D_TLB_REFILL:u - 307,806 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,767,037 L2D_CACHE:u - - 15.923392394 seconds time elapsed - - 83.307253000 seconds user - 958.949992000 seconds sys - - - diff --git a/pytorch/output_HPC/altra_10_30_va2010_1000.json b/pytorch/output_HPC/altra_10_30_va2010_1000.json deleted file mode 100644 index 916bc87..0000000 --- a/pytorch/output_HPC/altra_10_30_va2010_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [32.08, 31.8, 28.68, 27.6, 22.96, 22.08, 21.0, 20.84, 20.68, 20.72], "matrix": "va2010", "shape": [285762, 285762], "nnz": 1402128, "% density": 1.717033263003816e-05, "time_s": 14.632386922836304, "power": [85.16, 83.48, 76.96, 67.44, 54.04, 51.4, 54.24, 66.76, 83.2, 96.44, 96.44, 95.84, 94.24, 92.36, 91.2, 89.32, 87.48, 88.68, 88.24], "power_after": [21.12, 21.0, 21.16, 21.4, 21.32, 21.36, 21.36, 21.12, 20.76, 20.84], "task clock (msec)": 57.32, "page faults": 3280, "cycles": 39497791, "instructions": 64385555, "branch mispredictions": 332792, "branches": 19983954, "ITLB accesses": 27156853, "ITLB misses": 6466, "DTLB misses": 18244, "DTLB accesses": 36466301, "L1I cache accesses": 30929971, "L1I cache misses": 291811, "L1D cache misses": 473063, "L1D cache accesses": 32462905, "LL cache misses": 544953, "LL cache accesses": 565172, "L2D TLB accesses": 183225, "L2D TLB misses": 23924, "L2D cache misses": 301362, "L2D cache accesses": 1756590, "instructions per cycle": 1.6301052127193645, "branch miss rate": 0.01665296067034582, "ITLB miss rate": 0.00023809828038616994, "DTLB miss rate": 0.000500297521264907, "L2D TLB miss rate": 0.13057170145995362, "L1I cache miss rate": 0.009434570759862659, "L1D cache miss rate": 0.014572417348354991, "L2D cache miss rate": 0.17156080815671274, "LL cache miss rate": 0.964225050073252} diff --git a/pytorch/output_HPC/altra_10_30_va2010_1000.output b/pytorch/output_HPC/altra_10_30_va2010_1000.output deleted file mode 100644 index 80a5055..0000000 --- a/pytorch/output_HPC/altra_10_30_va2010_1000.output +++ /dev/null @@ -1,173 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3395277 queued and waiting for resources -srun: job 3395277 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 2, 8, ..., 1402119, - 1402123, 1402128]), - col_indices=tensor([ 2006, 2464, 1166, ..., 285581, 285634, - 285760]), - values=tensor([125334., 3558., 1192., ..., 10148., 1763., - 9832.]), size=(285762, 285762), nnz=1402128, - layout=torch.sparse_csr) -tensor([0.2920, 0.3583, 0.0598, ..., 0.2208, 0.1741, 0.4955]) -Matrix: va2010 -Shape: torch.Size([285762, 285762]) -NNZ: 1402128 -Density: 1.717033263003816e-05 -Time: 14.792448997497559 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/va2010.mtx 1000': - - 57.32 msec task-clock:u # 0.003 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,280 page-faults:u # 57.220 K/sec - 39,497,791 cycles:u # 0.689 GHz (54.25%) - 64,385,555 instructions:u # 1.63 insn per cycle (81.24%) - branches:u - 362,674 branch-misses:u - 33,532,520 L1-dcache-loads:u # 584.977 M/sec - 481,355 L1-dcache-load-misses:u # 1.44% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 31,924,348 L1-icache-loads:u # 556.922 M/sec - 296,637 L1-icache-load-misses:u # 0.93% of all L1-icache accesses - 43,420,143 dTLB-loads:u # 757.467 M/sec (40.22%) - 30,923 dTLB-load-misses:u # 0.07% of all dTLB cache accesses (19.05%) - iTLB-loads:u (0.00%) - iTLB-load-misses:u (0.00%) - - 18.678937115 seconds time elapsed - - 112.979167000 seconds user - 1135.785668000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 2, 8, ..., 1402119, - 1402123, 1402128]), - col_indices=tensor([ 2006, 2464, 1166, ..., 285581, 285634, - 285760]), - values=tensor([125334., 3558., 1192., ..., 10148., 1763., - 9832.]), size=(285762, 285762), nnz=1402128, - layout=torch.sparse_csr) -tensor([0.7703, 0.7481, 0.5351, ..., 0.4663, 0.6089, 0.3679]) -Matrix: va2010 -Shape: torch.Size([285762, 285762]) -NNZ: 1402128 -Density: 1.717033263003816e-05 -Time: 14.130552530288696 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/va2010.mtx 1000': - - 332,792 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 19,983,954 BR_RETIRED:u - - 17.923156218 seconds time elapsed - - 107.999690000 seconds user - 1091.659165000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 2, 8, ..., 1402119, - 1402123, 1402128]), - col_indices=tensor([ 2006, 2464, 1166, ..., 285581, 285634, - 285760]), - values=tensor([125334., 3558., 1192., ..., 10148., 1763., - 9832.]), size=(285762, 285762), nnz=1402128, - layout=torch.sparse_csr) -tensor([0.8850, 0.1406, 0.0617, ..., 0.4325, 0.2725, 0.9292]) -Matrix: va2010 -Shape: torch.Size([285762, 285762]) -NNZ: 1402128 -Density: 1.717033263003816e-05 -Time: 13.32977032661438 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/va2010.mtx 1000': - - 27,156,853 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 6,466 ITLB_WALK:u - 18,244 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 36,466,301 L1D_TLB:u - - 17.186572497 seconds time elapsed - - 104.940187000 seconds user - 1032.527271000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 2, 8, ..., 1402119, - 1402123, 1402128]), - col_indices=tensor([ 2006, 2464, 1166, ..., 285581, 285634, - 285760]), - values=tensor([125334., 3558., 1192., ..., 10148., 1763., - 9832.]), size=(285762, 285762), nnz=1402128, - layout=torch.sparse_csr) -tensor([0.6289, 0.0403, 0.9207, ..., 0.0183, 0.4807, 0.7504]) -Matrix: va2010 -Shape: torch.Size([285762, 285762]) -NNZ: 1402128 -Density: 1.717033263003816e-05 -Time: 13.460915803909302 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/va2010.mtx 1000': - - 30,929,971 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 291,811 L1I_CACHE_REFILL:u - 473,063 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 32,462,905 L1D_CACHE:u - - 17.219448483 seconds time elapsed - - 100.274467000 seconds user - 1045.271682000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 2, 8, ..., 1402119, - 1402123, 1402128]), - col_indices=tensor([ 2006, 2464, 1166, ..., 285581, 285634, - 285760]), - values=tensor([125334., 3558., 1192., ..., 10148., 1763., - 9832.]), size=(285762, 285762), nnz=1402128, - layout=torch.sparse_csr) -tensor([0.6412, 0.1151, 0.5075, ..., 0.9251, 0.9288, 0.3560]) -Matrix: va2010 -Shape: torch.Size([285762, 285762]) -NNZ: 1402128 -Density: 1.717033263003816e-05 -Time: 15.992860555648804 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/va2010.mtx 1000': - - 544,953 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 565,172 LL_CACHE_RD:u - 183,225 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 23,924 L2D_TLB_REFILL:u - 301,362 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,756,590 L2D_CACHE:u - - 19.884223259 seconds time elapsed - - 113.211516000 seconds user - 1230.525804000 seconds sys - - - diff --git a/pytorch/output_HPC/altra_10_30_vt2010_1000.json b/pytorch/output_HPC/altra_10_30_vt2010_1000.json deleted file mode 100644 index 3b71e9f..0000000 --- a/pytorch/output_HPC/altra_10_30_vt2010_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"architecture": "altra", "iterations": 1000, "baseline_time_s": 10, "baseline_delay_s": 30, "power_before": [77.2, 64.12, 64.12, 48.92, 36.2, 21.72, 21.88, 22.36, 22.36, 22.44], "matrix": "vt2010", "shape": [32580, 32580], "nnz": 155598, "% density": 0.00014658915806621921, "time_s": 3.5892834663391113, "power": [33.44, 30.68, 31.08, 26.96, 26.88, 32.48, 32.24], "power_after": [21.24, 21.24, 21.36, 21.36, 21.2, 21.04, 20.84, 20.72, 20.72, 20.56], "task clock (msec)": 55.26, "page faults": 3297, "cycles": 49276491, "instructions": 64763517, "branch mispredictions": 340611, "branches": 20355849, "ITLB accesses": 27946393, "ITLB misses": 6805, "DTLB misses": 17877, "DTLB accesses": 38226912, "L1I cache accesses": 31946141, "L1I cache misses": 295259, "L1D cache misses": 468136, "L1D cache accesses": 33395666, "LL cache misses": 527109, "LL cache accesses": 540409, "L2D TLB accesses": 192519, "L2D TLB misses": 24204, "L2D cache misses": 290933, "L2D cache accesses": 1743452, "instructions per cycle": 1.3142883286880147, "branch miss rate": 0.016732831924622747, "ITLB miss rate": 0.00024350190738389746, "DTLB miss rate": 0.0004676548291423592, "L2D TLB miss rate": 0.1257226559456469, "L1I cache miss rate": 0.009242399574959616, "L1D cache miss rate": 0.014017866869311724, "L2D cache miss rate": 0.16687181522634406, "LL cache miss rate": 0.9753890109158063} diff --git a/pytorch/output_HPC/altra_10_30_vt2010_1000.output b/pytorch/output_HPC/altra_10_30_vt2010_1000.output deleted file mode 100644 index 45aeebf..0000000 --- a/pytorch/output_HPC/altra_10_30_vt2010_1000.output +++ /dev/null @@ -1,163 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3395285 queued and waiting for resources -srun: job 3395285 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 4, 7, ..., 155588, 155592, - 155598]), - col_indices=tensor([ 131, 561, 996, ..., 32237, 32238, 32570]), - values=tensor([79040., 7820., 15136., ..., 2828., 17986., 2482.]), - size=(32580, 32580), nnz=155598, layout=torch.sparse_csr) -tensor([0.1179, 0.2288, 0.5357, ..., 0.4845, 0.6375, 0.4513]) -Matrix: vt2010 -Shape: torch.Size([32580, 32580]) -NNZ: 155598 -Density: 0.00014658915806621921 -Time: 3.628732681274414 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/vt2010.mtx 1000': - - 55.26 msec task-clock:u # 0.008 CPUs utilized - 0 context-switches:u # 0.000 /sec - 0 cpu-migrations:u # 0.000 /sec - 3,297 page-faults:u # 59.661 K/sec - 49,276,491 cycles:u # 0.892 GHz (31.65%) - 64,763,517 instructions:u # 1.31 insn per cycle (57.73%) - branches:u - 357,693 branch-misses:u (76.18%) - 32,426,852 L1-dcache-loads:u # 586.784 M/sec (88.36%) - 469,495 L1-dcache-load-misses:u # 1.45% of all L1-dcache accesses - LLC-loads:u - LLC-load-misses:u - 30,941,957 L1-icache-loads:u # 559.914 M/sec - 279,512 L1-icache-load-misses:u # 0.90% of all L1-icache accesses - 47,128,547 dTLB-loads:u # 852.821 M/sec (46.73%) - 108,931 dTLB-load-misses:u # 0.23% of all dTLB cache accesses (32.30%) - 14,189,608 iTLB-loads:u # 256.770 M/sec (19.86%) - iTLB-load-misses:u (0.00%) - - 7.117399121 seconds time elapsed - - 18.404618000 seconds user - 29.532104000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 4, 7, ..., 155588, 155592, - 155598]), - col_indices=tensor([ 131, 561, 996, ..., 32237, 32238, 32570]), - values=tensor([79040., 7820., 15136., ..., 2828., 17986., 2482.]), - size=(32580, 32580), nnz=155598, layout=torch.sparse_csr) -tensor([0.7544, 0.0071, 0.0491, ..., 0.7236, 0.5537, 0.4901]) -Matrix: vt2010 -Shape: torch.Size([32580, 32580]) -NNZ: 155598 -Density: 0.00014658915806621921 -Time: 3.6322426795959473 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/vt2010.mtx 1000': - - 340,611 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio - 20,355,849 BR_RETIRED:u - - 7.112879848 seconds time elapsed - - 18.362004000 seconds user - 29.398677000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 4, 7, ..., 155588, 155592, - 155598]), - col_indices=tensor([ 131, 561, 996, ..., 32237, 32238, 32570]), - values=tensor([79040., 7820., 15136., ..., 2828., 17986., 2482.]), - size=(32580, 32580), nnz=155598, layout=torch.sparse_csr) -tensor([0.7651, 0.6605, 0.7128, ..., 0.7434, 0.6656, 0.3987]) -Matrix: vt2010 -Shape: torch.Size([32580, 32580]) -NNZ: 155598 -Density: 0.00014658915806621921 -Time: 3.7933311462402344 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/vt2010.mtx 1000': - - 27,946,393 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio - 6,805 ITLB_WALK:u - 17,877 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio - 38,226,912 L1D_TLB:u - - 7.235266934 seconds time elapsed - - 18.566568000 seconds user - 29.759130000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 4, 7, ..., 155588, 155592, - 155598]), - col_indices=tensor([ 131, 561, 996, ..., 32237, 32238, 32570]), - values=tensor([79040., 7820., 15136., ..., 2828., 17986., 2482.]), - size=(32580, 32580), nnz=155598, layout=torch.sparse_csr) -tensor([0.3319, 0.1241, 0.4830, ..., 0.5188, 0.8684, 0.1488]) -Matrix: vt2010 -Shape: torch.Size([32580, 32580]) -NNZ: 155598 -Density: 0.00014658915806621921 -Time: 3.662006378173828 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/vt2010.mtx 1000': - - 31,946,141 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio - 295,259 L1I_CACHE_REFILL:u - 468,136 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio - 33,395,666 L1D_CACHE:u - - 7.187008251 seconds time elapsed - - 18.275672000 seconds user - 30.724065000 seconds sys - - - -/nfshomes/vut/ampere_research/pytorch/spmv.py:20: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 4, 7, ..., 155588, 155592, - 155598]), - col_indices=tensor([ 131, 561, 996, ..., 32237, 32238, 32570]), - values=tensor([79040., 7820., 15136., ..., 2828., 17986., 2482.]), - size=(32580, 32580), nnz=155598, layout=torch.sparse_csr) -tensor([0.0055, 0.7774, 0.9046, ..., 0.5143, 0.0678, 0.4725]) -Matrix: vt2010 -Shape: torch.Size([32580, 32580]) -NNZ: 155598 -Density: 0.00014658915806621921 -Time: 3.616023063659668 seconds - - Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/vt2010.mtx 1000': - - 527,109 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio - 540,409 LL_CACHE_RD:u - 192,519 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio - 24,204 L2D_TLB_REFILL:u - 290,933 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio - 1,743,452 L2D_CACHE:u - - 7.030605378 seconds time elapsed - - 18.274323000 seconds user - 28.779020000 seconds sys - - - diff --git a/pytorch/output_cpu/altra_10_10_ASIC_680k_10000.json b/pytorch/output_cpu/altra_10_10_ASIC_680k_10000.json deleted file mode 100644 index b552439..0000000 --- a/pytorch/output_cpu/altra_10_10_ASIC_680k_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 10000, "MATRIX_FILE": "ASIC_680k", "MATRIX_SHAPE": [682862, 682862], "MATRIX_SIZE": 466300511044, "MATRIX_NNZ": 3871773, "MATRIX_DENSITY": 8.303171256088674e-06, "TIME_S": 11.77456283569336, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [20.36, 20.44, 20.48, 20.72, 20.8, 21.0, 21.32, 21.32, 21.28, 21.08], "POWER": [92.0, 91.8, 78.72, 66.68, 51.2, 46.6, 53.36, 53.36, 70.48, 90.16, 100.04, 103.68, 98.2, 95.64, 97.16, 101.4], "JOULES": 938.4206715393068, "POWER_AFTER": [20.96, 20.76, 20.76, 21.08, 21.24, 21.16, 21.28, 21.2, 21.0, 21.08]} diff --git a/pytorch/output_cpu/altra_10_10_Oregon-2_10000.json b/pytorch/output_cpu/altra_10_10_Oregon-2_10000.json deleted file mode 100644 index 796afd3..0000000 --- a/pytorch/output_cpu/altra_10_10_Oregon-2_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 10000, "MATRIX_FILE": "Oregon-2", "MATRIX_SHAPE": [11806, 11806], "MATRIX_SIZE": 139381636, "MATRIX_NNZ": 65460, "MATRIX_DENSITY": 0.0004696458003979807, "TIME_S": 0.9880795478820801, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [21.04, 21.12, 21.2, 21.12, 21.04, 20.96, 20.92, 20.88, 21.16, 21.08], "POWER": [25.92, 42.32, 42.32, 45.44, 45.4], "JOULES": 44.85881147384644, "POWER_AFTER": [20.72, 20.72, 20.84, 20.84, 20.84, 20.96, 20.92, 20.6, 20.68, 20.84]} diff --git a/pytorch/output_cpu/altra_10_10_Oregon-2_10000.output b/pytorch/output_cpu/altra_10_10_Oregon-2_10000.output deleted file mode 100644 index d84fac7..0000000 --- a/pytorch/output_cpu/altra_10_10_Oregon-2_10000.output +++ /dev/null @@ -1,23 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3471014 queued and waiting for resources -srun: job 3471014 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 583, 584, ..., 65459, 65460, 65460]), - col_indices=tensor([ 2, 23, 27, ..., 3324, 958, 841]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(11806, 11806), - nnz=65460, layout=torch.sparse_csr) -tensor([0.2158, 0.5422, 0.9585, ..., 0.6377, 0.8158, 0.5743]) -Matrix: Oregon-2 -Shape: torch.Size([11806, 11806]) -Size: 139381636 -NNZ: 65460 -Density: 0.0004696458003979807 -Time: 0.9880795478820801 seconds - diff --git a/pytorch/output_cpu/altra_10_10_as-caida_10000.json b/pytorch/output_cpu/altra_10_10_as-caida_10000.json deleted file mode 100644 index ea58106..0000000 --- a/pytorch/output_cpu/altra_10_10_as-caida_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 10000, "MATRIX_FILE": "as-caida", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 106762, "MATRIX_DENSITY": 0.00010842726485909405, "TIME_S": 1.066300630569458, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [20.64, 20.48, 20.68, 20.64, 20.32, 20.32, 20.4, 20.2, 20.52, 20.52], "POWER": [26.32, 39.88, 50.16, 50.64, 50.24], "JOULES": 53.97094367980957, "POWER_AFTER": [20.28, 20.4, 20.2, 20.32, 20.32, 20.4, 20.48, 20.28, 20.28, 20.44]} diff --git a/pytorch/output_cpu/altra_10_10_as-caida_10000.output b/pytorch/output_cpu/altra_10_10_as-caida_10000.output deleted file mode 100644 index aee2265..0000000 --- a/pytorch/output_cpu/altra_10_10_as-caida_10000.output +++ /dev/null @@ -1,24 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3470988 queued and waiting for resources -srun: job 3470988 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 0, 0, ..., 106761, 106761, - 106762]), - col_indices=tensor([ 106, 329, 1040, ..., 155, 160, 12170]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379), - nnz=106762, layout=torch.sparse_csr) -tensor([0.8877, 0.6518, 0.0601, ..., 0.0372, 0.4806, 0.8853]) -Matrix: as-caida -Shape: torch.Size([31379, 31379]) -Size: 984641641 -NNZ: 106762 -Density: 0.00010842726485909405 -Time: 1.066300630569458 seconds - diff --git a/pytorch/output_cpu/altra_10_10_dc2_10000.json b/pytorch/output_cpu/altra_10_10_dc2_10000.json deleted file mode 100644 index f445a04..0000000 --- a/pytorch/output_cpu/altra_10_10_dc2_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 10000, "MATRIX_FILE": "dc2", "MATRIX_SHAPE": [116835, 116835], "MATRIX_SIZE": 13650417225, "MATRIX_NNZ": 766396, "MATRIX_DENSITY": 5.614451099680581e-05, "TIME_S": 3.0164122581481934, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [20.44, 20.72, 20.72, 21.0, 20.84, 21.08, 20.88, 20.8, 20.8, 20.88], "POWER": [64.4, 79.8, 83.24, 75.76, 58.2, 58.2, 56.64, 60.64, 75.88, 93.68], "JOULES": 194.69750034332276, "POWER_AFTER": [21.12, 21.0, 21.12, 20.88, 20.88, 20.84, 20.96, 20.92, 20.88, 20.8]} diff --git a/pytorch/output_cpu/altra_10_10_dc2_10000.output b/pytorch/output_cpu/altra_10_10_dc2_10000.output deleted file mode 100644 index a0aa6f0..0000000 --- a/pytorch/output_cpu/altra_10_10_dc2_10000.output +++ /dev/null @@ -1,26 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3470982 queued and waiting for resources -srun: job 3470982 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 1, 2, ..., 766390, 766394, - 766396]), - col_indices=tensor([ 0, 1, 2, ..., 116833, 89, - 116834]), - values=tensor([-1.0000e+00, -1.0000e+00, -1.0000e+00, ..., - 1.0331e+01, -1.0000e-03, 1.0000e-03]), - size=(116835, 116835), nnz=766396, layout=torch.sparse_csr) -tensor([0.3305, 0.9342, 0.6954, ..., 0.1999, 0.9064, 0.6304]) -Matrix: dc2 -Shape: torch.Size([116835, 116835]) -Size: 13650417225 -NNZ: 766396 -Density: 5.614451099680581e-05 -Time: 3.0164122581481934 seconds - diff --git a/pytorch/output_cpu/altra_10_10_de2010_10000.json b/pytorch/output_cpu/altra_10_10_de2010_10000.json deleted file mode 100644 index 1dfbbb2..0000000 --- a/pytorch/output_cpu/altra_10_10_de2010_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 10000, "MATRIX_FILE": "de2010", "MATRIX_SHAPE": [24115, 24115], "MATRIX_SIZE": 581533225, "MATRIX_NNZ": 116056, "MATRIX_DENSITY": 0.0001995689928120616, "TIME_S": 1.1378686428070068, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [21.0, 20.88, 21.04, 20.8, 20.8, 20.44, 20.64, 20.48, 20.28, 20.16], "POWER": [22.84, 39.8, 49.48, 50.32, 50.28], "JOULES": 57.25203536033631, "POWER_AFTER": [20.68, 20.44, 20.68, 20.68, 20.56, 20.88, 20.92, 20.88, 21.0, 20.96]} diff --git a/pytorch/output_cpu/altra_10_10_email-Enron_10000.json b/pytorch/output_cpu/altra_10_10_email-Enron_10000.json deleted file mode 100644 index d7162df..0000000 --- a/pytorch/output_cpu/altra_10_10_email-Enron_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 10000, "MATRIX_FILE": "email-Enron", "MATRIX_SHAPE": [36692, 36692], "MATRIX_SIZE": 1346302864, "MATRIX_NNZ": 367662, "MATRIX_DENSITY": 0.0002730901120626302, "TIME_S": 1.3314027786254883, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [20.8, 20.64, 20.6, 20.6, 20.48, 20.8, 20.72, 20.72, 20.92, 20.92], "POWER": [28.4, 43.96, 54.4, 55.28, 55.08], "JOULES": 73.5336650466919, "POWER_AFTER": [20.88, 20.8, 20.8, 20.8, 20.64, 20.64, 20.64, 20.48, 20.52, 20.72]} diff --git a/pytorch/output_cpu/altra_10_10_email-Enron_10000.output b/pytorch/output_cpu/altra_10_10_email-Enron_10000.output deleted file mode 100644 index 1499d7f..0000000 --- a/pytorch/output_cpu/altra_10_10_email-Enron_10000.output +++ /dev/null @@ -1,24 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3470985 queued and waiting for resources -srun: job 3470985 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 1, 71, ..., 367660, 367661, - 367662]), - col_indices=tensor([ 1, 0, 2, ..., 36690, 36689, 8203]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(36692, 36692), - nnz=367662, layout=torch.sparse_csr) -tensor([0.7107, 0.7540, 0.8321, ..., 0.9503, 0.7781, 0.9277]) -Matrix: email-Enron -Shape: torch.Size([36692, 36692]) -Size: 1346302864 -NNZ: 367662 -Density: 0.0002730901120626302 -Time: 1.3314027786254883 seconds - diff --git a/pytorch/output_cpu/altra_10_10_fl2010_10000.json b/pytorch/output_cpu/altra_10_10_fl2010_10000.json deleted file mode 100644 index 939ef0b..0000000 --- a/pytorch/output_cpu/altra_10_10_fl2010_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 10000, "MATRIX_FILE": "fl2010", "MATRIX_SHAPE": [484481, 484481], "MATRIX_SIZE": 234721839361, "MATRIX_NNZ": 2346294, "MATRIX_DENSITY": 9.99606174861054e-06, "TIME_S": 2.924255609512329, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [20.8, 20.88, 20.72, 20.64, 20.56, 20.92, 20.92, 21.0, 20.96, 20.84], "POWER": [73.32, 93.24, 93.64, 82.2, 61.36, 61.36, 58.0], "JOULES": 176.3268253517151, "POWER_AFTER": [20.76, 20.56, 20.76, 20.72, 20.76, 20.76, 20.76, 20.88, 20.68, 20.68]} diff --git a/pytorch/output_cpu/altra_10_10_ga2010_10000.json b/pytorch/output_cpu/altra_10_10_ga2010_10000.json deleted file mode 100644 index 19767fe..0000000 --- a/pytorch/output_cpu/altra_10_10_ga2010_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 10000, "MATRIX_FILE": "ga2010", "MATRIX_SHAPE": [291086, 291086], "MATRIX_SIZE": 84731059396, "MATRIX_NNZ": 1418056, "MATRIX_DENSITY": 1.6735964475229304e-05, "TIME_S": 2.341104745864868, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [20.32, 20.28, 20.24, 20.44, 20.52, 20.8, 20.64, 20.68, 20.6, 20.36], "POWER": [33.84, 53.08, 66.2, 66.52, 67.36, 59.0], "JOULES": 154.00518000602722, "POWER_AFTER": [20.28, 20.32, 20.52, 20.6, 20.6, 20.84, 21.12, 20.96, 20.76, 20.8]} diff --git a/pytorch/output_cpu/altra_10_10_ga2010_10000.output b/pytorch/output_cpu/altra_10_10_ga2010_10000.output deleted file mode 100644 index dabd461..0000000 --- a/pytorch/output_cpu/altra_10_10_ga2010_10000.output +++ /dev/null @@ -1,25 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3470989 queued and waiting for resources -srun: job 3470989 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 10, ..., 1418047, - 1418054, 1418056]), - col_indices=tensor([ 1566, 1871, 1997, ..., 291064, 289820, - 290176]), - values=tensor([18760., 17851., 18847., ..., 65219., 56729., 77629.]), - size=(291086, 291086), nnz=1418056, layout=torch.sparse_csr) -tensor([0.0746, 0.8150, 0.2560, ..., 0.7929, 0.2552, 0.7733]) -Matrix: ga2010 -Shape: torch.Size([291086, 291086]) -Size: 84731059396 -NNZ: 1418056 -Density: 1.6735964475229304e-05 -Time: 2.341104745864868 seconds - diff --git a/pytorch/output_cpu/altra_10_10_mac_econ_fwd500_10000.json b/pytorch/output_cpu/altra_10_10_mac_econ_fwd500_10000.json deleted file mode 100644 index 7e78fe9..0000000 --- a/pytorch/output_cpu/altra_10_10_mac_econ_fwd500_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 10000, "MATRIX_FILE": "mac_econ_fwd500", "MATRIX_SHAPE": [206500, 206500], "MATRIX_SIZE": 42642250000, "MATRIX_NNZ": 1273389, "MATRIX_DENSITY": 2.9862143765866013e-05, "TIME_S": 1.6093401908874512, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [20.76, 20.72, 20.28, 20.2, 20.24, 20.56, 20.72, 21.12, 21.24, 21.0], "POWER": [48.6, 65.2, 65.2, 61.84, 62.88, 59.36], "JOULES": 99.0504337310791, "POWER_AFTER": [20.76, 20.4, 20.64, 20.68, 20.68, 20.56, 20.48, 20.68, 20.64, 20.88]} diff --git a/pytorch/output_cpu/altra_10_10_mc2depi_10000.json b/pytorch/output_cpu/altra_10_10_mc2depi_10000.json deleted file mode 100644 index f54e717..0000000 --- a/pytorch/output_cpu/altra_10_10_mc2depi_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 10000, "MATRIX_FILE": "mc2depi", "MATRIX_SHAPE": [525825, 525825], "MATRIX_SIZE": 276491930625, "MATRIX_NNZ": 2100225, "MATRIX_DENSITY": 7.595972132902821e-06, "TIME_S": 2.123237371444702, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [20.68, 20.68, 20.68, 20.64, 20.72, 20.6, 20.84, 20.76, 20.92, 20.96], "POWER": [52.52, 76.2, 82.92, 85.4, 72.28, 58.76], "JOULES": 164.92142794609072, "POWER_AFTER": [20.68, 20.72, 20.84, 20.88, 20.84, 21.16, 21.04, 21.16, 20.88, 20.88]} diff --git a/pytorch/output_cpu/altra_10_10_mc2depi_10000.output b/pytorch/output_cpu/altra_10_10_mc2depi_10000.output deleted file mode 100644 index c4f19b8..0000000 --- a/pytorch/output_cpu/altra_10_10_mc2depi_10000.output +++ /dev/null @@ -1,25 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3470981 queued and waiting for resources -srun: job 3470981 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 2, 5, ..., 2100220, - 2100223, 2100225]), - col_indices=tensor([ 0, 1, 1, ..., 525824, 525821, - 525824]), - values=tensor([-2025., 2025., -2026., ..., 2025., 1024., -1024.]), - size=(525825, 525825), nnz=2100225, layout=torch.sparse_csr) -tensor([0.8254, 0.0543, 0.1764, ..., 0.7650, 0.8254, 0.6404]) -Matrix: mc2depi -Shape: torch.Size([525825, 525825]) -Size: 276491930625 -NNZ: 2100225 -Density: 7.595972132902821e-06 -Time: 2.123237371444702 seconds - diff --git a/pytorch/output_cpu/altra_10_10_p2p-Gnutella04_10000.json b/pytorch/output_cpu/altra_10_10_p2p-Gnutella04_10000.json deleted file mode 100644 index 072f297..0000000 --- a/pytorch/output_cpu/altra_10_10_p2p-Gnutella04_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 10000, "MATRIX_FILE": "p2p-Gnutella04", "MATRIX_SHAPE": [10879, 10879], "MATRIX_SIZE": 118352641, "MATRIX_NNZ": 39994, "MATRIX_DENSITY": 0.0003379223282393842, "TIME_S": 0.9692902565002441, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [20.6, 20.48, 20.64, 20.64, 20.64, 20.56, 20.52, 20.44, 20.24, 20.12], "POWER": [25.92, 43.16, 50.56, 48.4, 49.28], "JOULES": 47.76662384033203, "POWER_AFTER": [20.4, 20.52, 20.44, 20.64, 20.72, 20.64, 20.8, 20.6, 20.6, 20.64]} diff --git a/pytorch/output_cpu/altra_10_10_p2p-Gnutella24_10000.json b/pytorch/output_cpu/altra_10_10_p2p-Gnutella24_10000.json deleted file mode 100644 index e8ce518..0000000 --- a/pytorch/output_cpu/altra_10_10_p2p-Gnutella24_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 10000, "MATRIX_FILE": "p2p-Gnutella24", "MATRIX_SHAPE": [26518, 26518], "MATRIX_SIZE": 703204324, "MATRIX_NNZ": 65369, "MATRIX_DENSITY": 9.295875717624285e-05, "TIME_S": 0.9848971366882324, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [16.32, 16.36, 16.36, 16.32, 16.56, 16.64, 16.72, 16.92, 16.76, 16.96], "POWER": [22.56, 40.8, 42.16, 42.16, 39.84], "JOULES": 39.23830192565919, "POWER_AFTER": [16.56, 16.44, 16.44, 16.68, 16.72, 16.72, 16.76, 16.68, 16.68, 16.92]} diff --git a/pytorch/output_cpu/altra_10_10_p2p-Gnutella24_10000.output b/pytorch/output_cpu/altra_10_10_p2p-Gnutella24_10000.output deleted file mode 100644 index 892cbce..0000000 --- a/pytorch/output_cpu/altra_10_10_p2p-Gnutella24_10000.output +++ /dev/null @@ -1,23 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3471012 queued and waiting for resources -srun: job 3471012 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 9, 9, ..., 65369, 65369, 65369]), - col_indices=tensor([ 1, 2, 3, ..., 15065, 9401, 26517]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(26518, 26518), - nnz=65369, layout=torch.sparse_csr) -tensor([0.6126, 0.7089, 0.2938, ..., 0.5143, 0.3903, 0.8766]) -Matrix: p2p-Gnutella24 -Shape: torch.Size([26518, 26518]) -Size: 703204324 -NNZ: 65369 -Density: 9.295875717624285e-05 -Time: 0.9848971366882324 seconds - diff --git a/pytorch/output_cpu/altra_10_10_p2p-Gnutella25_10000.json b/pytorch/output_cpu/altra_10_10_p2p-Gnutella25_10000.json deleted file mode 100644 index bef61e3..0000000 --- a/pytorch/output_cpu/altra_10_10_p2p-Gnutella25_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 10000, "MATRIX_FILE": "p2p-Gnutella25", "MATRIX_SHAPE": [22687, 22687], "MATRIX_SIZE": 514699969, "MATRIX_NNZ": 54705, "MATRIX_DENSITY": 0.00010628522108964806, "TIME_S": 1.064000129699707, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [20.4, 20.68, 20.76, 20.6, 20.64, 20.48, 20.36, 20.48, 20.52, 20.52], "POWER": [33.4, 49.92, 52.44, 52.44, 51.68], "JOULES": 55.747526702880855, "POWER_AFTER": [20.96, 20.76, 20.96, 21.08, 20.64, 20.84, 20.84, 20.56, 20.28, 20.48]} diff --git a/pytorch/output_cpu/altra_10_10_p2p-Gnutella25_10000.output b/pytorch/output_cpu/altra_10_10_p2p-Gnutella25_10000.output deleted file mode 100644 index 2320569..0000000 --- a/pytorch/output_cpu/altra_10_10_p2p-Gnutella25_10000.output +++ /dev/null @@ -1,23 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3470999 queued and waiting for resources -srun: job 3470999 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 9, 9, ..., 54704, 54704, 54705]), - col_indices=tensor([ 1, 2, 3, ..., 17949, 22685, 144]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(22687, 22687), - nnz=54705, layout=torch.sparse_csr) -tensor([0.1096, 0.4722, 0.2402, ..., 0.8482, 0.4609, 0.1028]) -Matrix: p2p-Gnutella25 -Shape: torch.Size([22687, 22687]) -Size: 514699969 -NNZ: 54705 -Density: 0.00010628522108964806 -Time: 1.064000129699707 seconds - diff --git a/pytorch/output_cpu/altra_10_10_p2p-Gnutella30_10000.json b/pytorch/output_cpu/altra_10_10_p2p-Gnutella30_10000.json deleted file mode 100644 index b444ddf..0000000 --- a/pytorch/output_cpu/altra_10_10_p2p-Gnutella30_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 10000, "MATRIX_FILE": "p2p-Gnutella30", "MATRIX_SHAPE": [36682, 36682], "MATRIX_SIZE": 1345569124, "MATRIX_NNZ": 88328, "MATRIX_DENSITY": 6.564359899804003e-05, "TIME_S": 1.022092580795288, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [20.44, 20.56, 20.76, 20.6, 20.64, 21.08, 20.76, 20.32, 20.32, 20.44], "POWER": [25.64, 36.88, 51.72, 49.6, 50.84], "JOULES": 50.723186807632445, "POWER_AFTER": [20.56, 20.68, 20.6, 20.88, 21.08, 20.76, 20.76, 20.92, 20.32, 20.24]} diff --git a/pytorch/output_cpu/altra_10_10_ri2010_10000.json b/pytorch/output_cpu/altra_10_10_ri2010_10000.json deleted file mode 100644 index 7eb8f84..0000000 --- a/pytorch/output_cpu/altra_10_10_ri2010_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 10000, "MATRIX_FILE": "ri2010", "MATRIX_SHAPE": [25181, 25181], "MATRIX_SIZE": 634082761, "MATRIX_NNZ": 125750, "MATRIX_DENSITY": 0.00019831796057928155, "TIME_S": 0.7675364017486572, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [20.64, 20.64, 20.64, 20.64, 20.8, 20.8, 20.8, 20.96, 20.92, 20.84], "POWER": [26.52, 43.16, 47.12, 46.0, 47.48], "JOULES": 36.442628355026244, "POWER_AFTER": [20.48, 20.44, 20.6, 20.64, 20.6, 20.68, 20.6, 20.8, 20.6, 20.6]} diff --git a/pytorch/output_cpu/altra_10_10_rma10_10000.json b/pytorch/output_cpu/altra_10_10_rma10_10000.json deleted file mode 100644 index 411d0b7..0000000 --- a/pytorch/output_cpu/altra_10_10_rma10_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 10000, "MATRIX_FILE": "rma10", "MATRIX_SHAPE": [46835, 46835], "MATRIX_SIZE": 2193517225, "MATRIX_NNZ": 2374001, "MATRIX_DENSITY": 0.0010822805369125833, "TIME_S": 2.688584089279175, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [20.24, 20.24, 20.4, 20.44, 20.76, 20.76, 20.68, 20.72, 20.56, 20.44], "POWER": [53.84, 65.36, 65.36, 65.6, 62.2, 50.6], "JOULES": 162.64235491752623, "POWER_AFTER": [20.28, 20.4, 20.48, 20.44, 20.4, 20.48, 20.52, 20.44, 20.44, 20.44]} diff --git a/pytorch/output_cpu/altra_10_10_soc-sign-Slashdot090216_10000.json b/pytorch/output_cpu/altra_10_10_soc-sign-Slashdot090216_10000.json deleted file mode 100644 index 9ea8d7c..0000000 --- a/pytorch/output_cpu/altra_10_10_soc-sign-Slashdot090216_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 10000, "MATRIX_FILE": "soc-sign-Slashdot090216", "MATRIX_SHAPE": [81871, 81871], "MATRIX_SIZE": 6702860641, "MATRIX_NNZ": 545671, "MATRIX_DENSITY": 8.140867447881048e-05, "TIME_S": 1.4809374809265137, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [21.16, 20.96, 20.92, 20.92, 20.76, 20.72, 21.04, 21.04, 21.08, 20.84], "POWER": [38.4, 56.52, 60.12, 59.64, 58.44], "JOULES": 87.74598638534546, "POWER_AFTER": [20.56, 20.56, 20.68, 20.52, 21.16, 21.16, 21.28, 21.0, 21.12, 20.84]} diff --git a/pytorch/output_cpu/altra_10_10_soc-sign-Slashdot090221_10000.json b/pytorch/output_cpu/altra_10_10_soc-sign-Slashdot090221_10000.json deleted file mode 100644 index ff323c2..0000000 --- a/pytorch/output_cpu/altra_10_10_soc-sign-Slashdot090221_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 10000, "MATRIX_FILE": "soc-sign-Slashdot090221", "MATRIX_SHAPE": [82144, 82144], "MATRIX_SIZE": 6747636736, "MATRIX_NNZ": 549202, "MATRIX_DENSITY": 8.13917555860553e-05, "TIME_S": 1.608903408050537, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [20.68, 20.68, 20.64, 20.28, 20.32, 20.44, 20.44, 20.44, 20.44, 20.52], "POWER": [57.2, 57.2, 72.76, 72.52, 70.32, 58.68], "JOULES": 106.05045198440551, "POWER_AFTER": [20.96, 20.76, 20.84, 20.92, 20.92, 20.96, 21.12, 21.24, 21.16, 21.04]} diff --git a/pytorch/output_cpu/altra_10_10_soc-sign-epinions_10000.json b/pytorch/output_cpu/altra_10_10_soc-sign-epinions_10000.json deleted file mode 100644 index 8b8d867..0000000 --- a/pytorch/output_cpu/altra_10_10_soc-sign-epinions_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 10000, "MATRIX_FILE": "soc-sign-epinions", "MATRIX_SHAPE": [131828, 131828], "MATRIX_SIZE": 17378621584, "MATRIX_NNZ": 841372, "MATRIX_DENSITY": 4.841419648464106e-05, "TIME_S": 4.555854320526123, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [16.4, 16.36, 16.48, 16.68, 16.32, 16.32, 16.56, 16.56, 16.64, 16.64], "POWER": [51.6, 68.68, 77.56, 77.4, 61.4, 55.08, 54.44, 65.6], "JOULES": 284.7840434265137, "POWER_AFTER": [16.92, 16.88, 17.04, 16.92, 16.84, 16.92, 16.88, 16.8, 17.12, 17.12]} diff --git a/pytorch/output_cpu/altra_10_10_sx-mathoverflow_10000.json b/pytorch/output_cpu/altra_10_10_sx-mathoverflow_10000.json deleted file mode 100644 index 0f61308..0000000 --- a/pytorch/output_cpu/altra_10_10_sx-mathoverflow_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 10000, "MATRIX_FILE": "sx-mathoverflow", "MATRIX_SHAPE": [24818, 24818], "MATRIX_SIZE": 615933124, "MATRIX_NNZ": 239978, "MATRIX_DENSITY": 0.00038961697406616504, "TIME_S": 1.0039293766021729, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [20.88, 21.0, 21.0, 20.92, 20.92, 20.8, 20.6, 20.6, 20.76, 20.92], "POWER": [29.76, 49.24, 50.6, 47.84, 47.84], "JOULES": 48.02798137664795, "POWER_AFTER": [20.96, 20.8, 20.92, 21.68, 22.4, 23.04, 23.76, 23.12, 22.6, 21.8]} diff --git a/pytorch/output_cpu/altra_10_10_sx-mathoverflow_10000.output b/pytorch/output_cpu/altra_10_10_sx-mathoverflow_10000.output deleted file mode 100644 index 458979b..0000000 --- a/pytorch/output_cpu/altra_10_10_sx-mathoverflow_10000.output +++ /dev/null @@ -1,24 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3470984 queued and waiting for resources -srun: job 3470984 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 317, 416, ..., 239976, 239977, - 239978]), - col_indices=tensor([ 0, 1, 2, ..., 1483, 2179, 24817]), - values=tensor([151., 17., 6., ..., 1., 1., 1.]), - size=(24818, 24818), nnz=239978, layout=torch.sparse_csr) -tensor([0.8169, 0.9455, 0.2378, ..., 0.7183, 0.8285, 0.9774]) -Matrix: sx-mathoverflow -Shape: torch.Size([24818, 24818]) -Size: 615933124 -NNZ: 239978 -Density: 0.00038961697406616504 -Time: 1.0039293766021729 seconds - diff --git a/pytorch/output_cpu/altra_10_10_tn2010_10000.json b/pytorch/output_cpu/altra_10_10_tn2010_10000.json deleted file mode 100644 index 4d8c5bf..0000000 --- a/pytorch/output_cpu/altra_10_10_tn2010_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 10000, "MATRIX_FILE": "tn2010", "MATRIX_SHAPE": [240116, 240116], "MATRIX_SIZE": 57655693456, "MATRIX_NNZ": 1193966, "MATRIX_DENSITY": 2.070855328296721e-05, "TIME_S": 2.2318568229675293, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [20.52, 20.52, 20.68, 20.6, 20.76, 20.84, 20.52, 20.44, 20.48, 20.4], "POWER": [47.04, 68.12, 70.92, 71.88, 71.88, 61.28], "JOULES": 157.9681861114502, "POWER_AFTER": [21.04, 20.76, 20.8, 20.72, 20.76, 20.84, 20.92, 21.04, 20.8, 20.8]} diff --git a/pytorch/output_cpu/altra_10_10_tn2010_10000.output b/pytorch/output_cpu/altra_10_10_tn2010_10000.output deleted file mode 100644 index d7a77ad..0000000 --- a/pytorch/output_cpu/altra_10_10_tn2010_10000.output +++ /dev/null @@ -1,26 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3470986 queued and waiting for resources -srun: job 3470986 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 4, 20, ..., 1193961, - 1193963, 1193966]), - col_indices=tensor([ 1152, 1272, 1961, ..., 238254, 239142, - 240113]), - values=tensor([ 5728., 2871., 418449., ..., 10058., 33324., - 34928.]), size=(240116, 240116), nnz=1193966, - layout=torch.sparse_csr) -tensor([0.2593, 0.6684, 0.1857, ..., 0.6282, 0.3314, 0.7454]) -Matrix: tn2010 -Shape: torch.Size([240116, 240116]) -Size: 57655693456 -NNZ: 1193966 -Density: 2.070855328296721e-05 -Time: 2.2318568229675293 seconds - diff --git a/pytorch/output_cpu/altra_10_10_ut2010_10000.json b/pytorch/output_cpu/altra_10_10_ut2010_10000.json deleted file mode 100644 index 6ba8d96..0000000 --- a/pytorch/output_cpu/altra_10_10_ut2010_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 10000, "MATRIX_FILE": "ut2010", "MATRIX_SHAPE": [115406, 115406], "MATRIX_SIZE": 13318544836, "MATRIX_NNZ": 572066, "MATRIX_DENSITY": 4.295259032005559e-05, "TIME_S": 1.5120632648468018, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [23.36, 22.84, 22.36, 21.92, 21.48, 21.48, 21.72, 22.08, 22.64, 23.28], "POWER": [43.48, 59.4, 65.28, 65.16, 62.16], "JOULES": 96.98985254287719, "POWER_AFTER": [22.56, 22.8, 22.24, 21.84, 21.4, 21.32, 20.96, 21.28, 21.36, 21.08]} diff --git a/pytorch/output_cpu/altra_10_10_ut2010_10000.output b/pytorch/output_cpu/altra_10_10_ut2010_10000.output deleted file mode 100644 index 5956478..0000000 --- a/pytorch/output_cpu/altra_10_10_ut2010_10000.output +++ /dev/null @@ -1,26 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3471001 queued and waiting for resources -srun: job 3471001 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 9, ..., 572056, 572061, - 572066]), - col_indices=tensor([ 453, 1291, 1979, ..., 113521, 114509, - 114602]), - values=tensor([160642., 31335., 282373., ..., 88393., 99485., - 18651.]), size=(115406, 115406), nnz=572066, - layout=torch.sparse_csr) -tensor([0.9240, 0.3751, 0.9849, ..., 0.9377, 0.9441, 0.6765]) -Matrix: ut2010 -Shape: torch.Size([115406, 115406]) -Size: 13318544836 -NNZ: 572066 -Density: 4.295259032005559e-05 -Time: 1.5120632648468018 seconds - diff --git a/pytorch/output_cpu/altra_10_10_va2010_10000.json b/pytorch/output_cpu/altra_10_10_va2010_10000.json deleted file mode 100644 index e22845a..0000000 --- a/pytorch/output_cpu/altra_10_10_va2010_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 10000, "MATRIX_FILE": "va2010", "MATRIX_SHAPE": [285762, 285762], "MATRIX_SIZE": 81659920644, "MATRIX_NNZ": 1402128, "MATRIX_DENSITY": 1.717033263003816e-05, "TIME_S": 2.1484014987945557, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [20.76, 20.72, 20.76, 20.88, 20.88, 20.96, 20.96, 20.96, 20.8, 20.6], "POWER": [65.16, 84.16, 87.88, 82.08, 64.16, 59.44], "JOULES": 155.0609850883484, "POWER_AFTER": [20.52, 20.52, 20.72, 20.56, 20.64, 20.64, 20.72, 20.92, 21.16, 21.32]} diff --git a/pytorch/output_cpu/altra_10_10_va2010_10000.output b/pytorch/output_cpu/altra_10_10_va2010_10000.output deleted file mode 100644 index de218b7..0000000 --- a/pytorch/output_cpu/altra_10_10_va2010_10000.output +++ /dev/null @@ -1,26 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3471004 queued and waiting for resources -srun: job 3471004 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 2, 8, ..., 1402119, - 1402123, 1402128]), - col_indices=tensor([ 2006, 2464, 1166, ..., 285581, 285634, - 285760]), - values=tensor([125334., 3558., 1192., ..., 10148., 1763., - 9832.]), size=(285762, 285762), nnz=1402128, - layout=torch.sparse_csr) -tensor([0.5972, 0.8492, 0.1772, ..., 0.7912, 0.0415, 0.8296]) -Matrix: va2010 -Shape: torch.Size([285762, 285762]) -Size: 81659920644 -NNZ: 1402128 -Density: 1.717033263003816e-05 -Time: 2.1484014987945557 seconds - diff --git a/pytorch/output_cpu/altra_10_10_vt2010_10000.json b/pytorch/output_cpu/altra_10_10_vt2010_10000.json deleted file mode 100644 index 78aeae3..0000000 --- a/pytorch/output_cpu/altra_10_10_vt2010_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 10000, "MATRIX_FILE": "vt2010", "MATRIX_SHAPE": [32580, 32580], "MATRIX_SIZE": 1061456400, "MATRIX_NNZ": 155598, "MATRIX_DENSITY": 0.00014658915806621921, "TIME_S": 0.8885588645935059, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [20.84, 20.84, 20.84, 20.92, 20.96, 20.88, 20.72, 20.52, 20.28, 20.28], "POWER": [24.52, 34.36, 45.4, 48.68, 47.56], "JOULES": 42.25985960006714, "POWER_AFTER": [20.36, 20.48, 20.56, 20.8, 21.08, 21.08, 21.28, 21.6, 21.68, 21.48]} diff --git a/pytorch/output_cpu/altra_10_10_vt2010_10000.output b/pytorch/output_cpu/altra_10_10_vt2010_10000.output deleted file mode 100644 index 996ee9d..0000000 --- a/pytorch/output_cpu/altra_10_10_vt2010_10000.output +++ /dev/null @@ -1,24 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3471005 queued and waiting for resources -srun: job 3471005 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 4, 7, ..., 155588, 155592, - 155598]), - col_indices=tensor([ 131, 561, 996, ..., 32237, 32238, 32570]), - values=tensor([79040., 7820., 15136., ..., 2828., 17986., 2482.]), - size=(32580, 32580), nnz=155598, layout=torch.sparse_csr) -tensor([0.7980, 0.7955, 0.8301, ..., 0.2464, 0.9642, 0.0961]) -Matrix: vt2010 -Shape: torch.Size([32580, 32580]) -Size: 1061456400 -NNZ: 155598 -Density: 0.00014658915806621921 -Time: 0.8885588645935059 seconds - diff --git a/pytorch/output_cpu/epyc_7313p_10_10_ASIC_680k_10000.json b/pytorch/output_cpu/epyc_7313p_10_10_ASIC_680k_10000.json deleted file mode 100644 index af8b627..0000000 --- a/pytorch/output_cpu/epyc_7313p_10_10_ASIC_680k_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 10000, "MATRIX_FILE": "ASIC_680k", "MATRIX_SHAPE": [682862, 682862], "MATRIX_SIZE": 466300511044, "MATRIX_NNZ": 3871773, "MATRIX_DENSITY": 8.303171256088674e-06, "TIME_S": 7.5851967334747314, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [41.12, 39.22, 38.67, 39.0, 39.11, 39.2, 39.03, 39.06, 39.93, 38.51], "POWER": [122.77], "JOULES": 931.2346029686928, "POWER_AFTER": [40.16, 38.97, 38.8, 39.29, 39.44, 38.77, 39.27, 38.71, 38.69, 38.72]} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_Oregon-2_10000.json b/pytorch/output_cpu/epyc_7313p_10_10_Oregon-2_10000.json deleted file mode 100644 index a5dcb6a..0000000 --- a/pytorch/output_cpu/epyc_7313p_10_10_Oregon-2_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 10000, "MATRIX_FILE": "Oregon-2", "MATRIX_SHAPE": [11806, 11806], "MATRIX_SIZE": 139381636, "MATRIX_NNZ": 65460, "MATRIX_DENSITY": 0.0004696458003979807, "TIME_S": 0.4882948398590088, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [41.11, 38.63, 39.35, 38.39, 39.53, 38.33, 39.4, 39.34, 42.37, 41.56], "POWER": [78.62], "JOULES": 38.38974030971527, "POWER_AFTER": [41.57, 38.36, 39.18, 38.33, 39.47, 38.52, 39.07, 38.29, 39.18, 38.38]} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_as-caida_10000.json b/pytorch/output_cpu/epyc_7313p_10_10_as-caida_10000.json deleted file mode 100644 index 827b10e..0000000 --- a/pytorch/output_cpu/epyc_7313p_10_10_as-caida_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 10000, "MATRIX_FILE": "as-caida", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 106762, "MATRIX_DENSITY": 0.00010842726485909405, "TIME_S": 0.6748511791229248, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [40.0, 38.56, 38.3, 38.46, 39.39, 38.44, 38.81, 38.3, 38.45, 38.62], "POWER": [80.47], "JOULES": 54.30527438402176, "POWER_AFTER": [40.22, 38.5, 39.18, 38.29, 39.13, 38.27, 38.85, 38.25, 38.39, 38.34]} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_dc2_10000.json b/pytorch/output_cpu/epyc_7313p_10_10_dc2_10000.json deleted file mode 100644 index b9a980e..0000000 --- a/pytorch/output_cpu/epyc_7313p_10_10_dc2_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 10000, "MATRIX_FILE": "dc2", "MATRIX_SHAPE": [116835, 116835], "MATRIX_SIZE": 13650417225, "MATRIX_NNZ": 766396, "MATRIX_DENSITY": 5.614451099680581e-05, "TIME_S": 2.0699713230133057, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [39.82, 38.4, 39.36, 38.95, 39.48, 38.39, 39.32, 38.43, 39.06, 38.43], "POWER": [98.5], "JOULES": 203.8921753168106, "POWER_AFTER": [39.63, 39.36, 38.51, 38.63, 38.49, 39.69, 38.57, 39.3, 38.49, 39.44]} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_de2010_10000.json b/pytorch/output_cpu/epyc_7313p_10_10_de2010_10000.json deleted file mode 100644 index ab2718e..0000000 --- a/pytorch/output_cpu/epyc_7313p_10_10_de2010_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 10000, "MATRIX_FILE": "de2010", "MATRIX_SHAPE": [24115, 24115], "MATRIX_SIZE": 581533225, "MATRIX_NNZ": 116056, "MATRIX_DENSITY": 0.0001995689928120616, "TIME_S": 0.5970535278320312, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [39.59, 38.0, 38.9, 38.03, 39.01, 37.93, 38.85, 37.93, 38.11, 38.06], "POWER": [79.47], "JOULES": 47.447843856811524, "POWER_AFTER": [40.4, 38.96, 38.62, 38.82, 38.75, 38.73, 38.33, 38.84, 38.46, 39.16]} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_email-Enron_10000.json b/pytorch/output_cpu/epyc_7313p_10_10_email-Enron_10000.json deleted file mode 100644 index 69cf030..0000000 --- a/pytorch/output_cpu/epyc_7313p_10_10_email-Enron_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 10000, "MATRIX_FILE": "email-Enron", "MATRIX_SHAPE": [36692, 36692], "MATRIX_SIZE": 1346302864, "MATRIX_NNZ": 367662, "MATRIX_DENSITY": 0.0002730901120626302, "TIME_S": 1.2558205127716064, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [39.74, 38.27, 38.91, 38.72, 38.52, 38.36, 38.35, 39.81, 38.58, 38.28], "POWER": [88.07], "JOULES": 110.60011255979538, "POWER_AFTER": [40.62, 39.93, 43.98, 39.02, 38.58, 38.9, 38.4, 38.52, 38.45, 38.36]} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_fl2010_10000.json b/pytorch/output_cpu/epyc_7313p_10_10_fl2010_10000.json deleted file mode 100644 index 9a15ca0..0000000 --- a/pytorch/output_cpu/epyc_7313p_10_10_fl2010_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 10000, "MATRIX_FILE": "fl2010", "MATRIX_SHAPE": [484481, 484481], "MATRIX_SIZE": 234721839361, "MATRIX_NNZ": 2346294, "MATRIX_DENSITY": 9.99606174861054e-06, "TIME_S": 3.8482837677001953, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [39.85, 44.42, 38.66, 39.02, 38.56, 38.59, 39.21, 38.44, 38.64, 38.76], "POWER": [123.85], "JOULES": 476.60994462966914, "POWER_AFTER": [41.95, 38.59, 39.04, 38.7, 38.69, 38.66, 39.8, 38.57, 39.54, 38.67]} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_ga2010_10000.json b/pytorch/output_cpu/epyc_7313p_10_10_ga2010_10000.json deleted file mode 100644 index e3f12ca..0000000 --- a/pytorch/output_cpu/epyc_7313p_10_10_ga2010_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 10000, "MATRIX_FILE": "ga2010", "MATRIX_SHAPE": [291086, 291086], "MATRIX_SIZE": 84731059396, "MATRIX_NNZ": 1418056, "MATRIX_DENSITY": 1.6735964475229304e-05, "TIME_S": 2.374833583831787, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [40.39, 38.45, 39.02, 38.45, 39.0, 38.2, 39.33, 38.48, 39.15, 40.07], "POWER": [110.89], "JOULES": 263.34529611110685, "POWER_AFTER": [39.66, 39.39, 38.41, 39.31, 38.38, 38.89, 38.48, 39.38, 38.53, 39.22]} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_mac_econ_fwd500_10000.json b/pytorch/output_cpu/epyc_7313p_10_10_mac_econ_fwd500_10000.json deleted file mode 100644 index 51d2df3..0000000 --- a/pytorch/output_cpu/epyc_7313p_10_10_mac_econ_fwd500_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 10000, "MATRIX_FILE": "mac_econ_fwd500", "MATRIX_SHAPE": [206500, 206500], "MATRIX_SIZE": 42642250000, "MATRIX_NNZ": 1273389, "MATRIX_DENSITY": 2.9862143765866013e-05, "TIME_S": 1.166548252105713, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [40.4, 38.82, 39.17, 38.67, 39.62, 39.37, 38.77, 38.83, 39.09, 38.67], "POWER": [96.59], "JOULES": 112.67689567089081, "POWER_AFTER": [39.82, 39.34, 39.56, 38.78, 38.54, 39.44, 38.58, 39.51, 38.79, 39.36]} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_mc2depi_10000.json b/pytorch/output_cpu/epyc_7313p_10_10_mc2depi_10000.json deleted file mode 100644 index b0ff3f7..0000000 --- a/pytorch/output_cpu/epyc_7313p_10_10_mc2depi_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 10000, "MATRIX_FILE": "mc2depi", "MATRIX_SHAPE": [525825, 525825], "MATRIX_SIZE": 276491930625, "MATRIX_NNZ": 2100225, "MATRIX_DENSITY": 7.595972132902821e-06, "TIME_S": 1.4909443855285645, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [40.83, 38.41, 39.81, 39.67, 44.4, 38.42, 38.51, 39.33, 39.34, 39.73], "POWER": [105.92], "JOULES": 157.92082931518556, "POWER_AFTER": [41.76, 38.57, 38.98, 38.45, 39.54, 39.61, 44.62, 39.62, 39.65, 38.4]} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_p2p-Gnutella04_10000.json b/pytorch/output_cpu/epyc_7313p_10_10_p2p-Gnutella04_10000.json deleted file mode 100644 index 08dcfed..0000000 --- a/pytorch/output_cpu/epyc_7313p_10_10_p2p-Gnutella04_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 10000, "MATRIX_FILE": "p2p-Gnutella04", "MATRIX_SHAPE": [10879, 10879], "MATRIX_SIZE": 118352641, "MATRIX_NNZ": 39994, "MATRIX_DENSITY": 0.0003379223282393842, "TIME_S": 0.3917062282562256, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [40.7, 38.58, 38.9, 38.56, 44.06, 38.54, 38.5, 38.53, 39.05, 38.74], "POWER": [78.01], "JOULES": 30.55700286626816, "POWER_AFTER": [39.85, 39.15, 38.49, 38.63, 38.95, 38.93, 38.57, 41.29, 43.97, 38.61]} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_p2p-Gnutella24_10000.json b/pytorch/output_cpu/epyc_7313p_10_10_p2p-Gnutella24_10000.json deleted file mode 100644 index 2896e4d..0000000 --- a/pytorch/output_cpu/epyc_7313p_10_10_p2p-Gnutella24_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 10000, "MATRIX_FILE": "p2p-Gnutella24", "MATRIX_SHAPE": [26518, 26518], "MATRIX_SIZE": 703204324, "MATRIX_NNZ": 65369, "MATRIX_DENSITY": 9.295875717624285e-05, "TIME_S": 0.5904901027679443, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [41.69, 38.5, 38.66, 38.85, 39.3, 38.41, 39.31, 38.3, 39.22, 38.45], "POWER": [78.57], "JOULES": 46.39480737447738, "POWER_AFTER": [39.95, 38.46, 44.94, 39.14, 39.06, 39.03, 38.93, 39.28, 38.52, 38.92]} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_p2p-Gnutella25_10000.json b/pytorch/output_cpu/epyc_7313p_10_10_p2p-Gnutella25_10000.json deleted file mode 100644 index 2f91672..0000000 --- a/pytorch/output_cpu/epyc_7313p_10_10_p2p-Gnutella25_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 10000, "MATRIX_FILE": "p2p-Gnutella25", "MATRIX_SHAPE": [22687, 22687], "MATRIX_SIZE": 514699969, "MATRIX_NNZ": 54705, "MATRIX_DENSITY": 0.00010628522108964806, "TIME_S": 0.5463590621948242, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [40.5, 38.94, 39.13, 38.5, 38.9, 39.01, 38.41, 38.75, 38.72, 38.83], "POWER": [80.15], "JOULES": 43.79067883491516, "POWER_AFTER": [40.41, 39.07, 39.98, 38.86, 38.61, 39.01, 39.29, 38.36, 38.7, 39.03]} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_p2p-Gnutella30_10000.json b/pytorch/output_cpu/epyc_7313p_10_10_p2p-Gnutella30_10000.json deleted file mode 100644 index b283dff..0000000 --- a/pytorch/output_cpu/epyc_7313p_10_10_p2p-Gnutella30_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 10000, "MATRIX_FILE": "p2p-Gnutella30", "MATRIX_SHAPE": [36682, 36682], "MATRIX_SIZE": 1345569124, "MATRIX_NNZ": 88328, "MATRIX_DENSITY": 6.564359899804003e-05, "TIME_S": 0.5678565502166748, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [40.93, 39.31, 38.47, 39.16, 38.63, 39.42, 38.44, 39.46, 38.84, 39.32], "POWER": [80.0], "JOULES": 45.428524017333984, "POWER_AFTER": [42.33, 38.87, 38.51, 39.86, 38.5, 39.57, 38.64, 39.3, 38.59, 39.52]} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_ri2010_10000.json b/pytorch/output_cpu/epyc_7313p_10_10_ri2010_10000.json deleted file mode 100644 index ad9b1c9..0000000 --- a/pytorch/output_cpu/epyc_7313p_10_10_ri2010_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 10000, "MATRIX_FILE": "ri2010", "MATRIX_SHAPE": [25181, 25181], "MATRIX_SIZE": 634082761, "MATRIX_NNZ": 125750, "MATRIX_DENSITY": 0.00019831796057928155, "TIME_S": 0.610253095626831, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [39.97, 38.69, 38.83, 38.7, 38.39, 38.32, 38.45, 38.47, 38.36, 38.28], "POWER": [80.37], "JOULES": 49.04604129552841, "POWER_AFTER": [40.01, 38.51, 38.35, 38.59, 39.52, 38.74, 38.53, 38.37, 38.81, 39.15]} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_rma10_10000.json b/pytorch/output_cpu/epyc_7313p_10_10_rma10_10000.json deleted file mode 100644 index 2264e68..0000000 --- a/pytorch/output_cpu/epyc_7313p_10_10_rma10_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 10000, "MATRIX_FILE": "rma10", "MATRIX_SHAPE": [46835, 46835], "MATRIX_SIZE": 2193517225, "MATRIX_NNZ": 2374001, "MATRIX_DENSITY": 0.0010822805369125833, "TIME_S": 3.9620909690856934, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [40.62, 38.96, 38.47, 39.5, 38.54, 39.4, 38.42, 39.25, 39.24, 38.47], "POWER": [112.26], "JOULES": 444.78433218955996, "POWER_AFTER": [41.23, 39.08, 39.0, 39.24, 38.82, 39.26, 38.75, 40.54, 45.02, 39.59]} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_soc-sign-Slashdot090216_10000.json b/pytorch/output_cpu/epyc_7313p_10_10_soc-sign-Slashdot090216_10000.json deleted file mode 100644 index acd2056..0000000 --- a/pytorch/output_cpu/epyc_7313p_10_10_soc-sign-Slashdot090216_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 10000, "MATRIX_FILE": "soc-sign-Slashdot090216", "MATRIX_SHAPE": [81871, 81871], "MATRIX_SIZE": 6702860641, "MATRIX_NNZ": 545671, "MATRIX_DENSITY": 8.140867447881048e-05, "TIME_S": 1.620380163192749, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [41.09, 39.77, 38.98, 38.92, 39.2, 38.85, 38.8, 38.42, 38.67, 38.84], "POWER": [95.89], "JOULES": 155.3782538485527, "POWER_AFTER": [41.04, 38.44, 38.51, 39.43, 38.9, 38.66, 38.49, 38.69, 40.35, 38.43]} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_soc-sign-Slashdot090221_10000.json b/pytorch/output_cpu/epyc_7313p_10_10_soc-sign-Slashdot090221_10000.json deleted file mode 100644 index 3694b0b..0000000 --- a/pytorch/output_cpu/epyc_7313p_10_10_soc-sign-Slashdot090221_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 10000, "MATRIX_FILE": "soc-sign-Slashdot090221", "MATRIX_SHAPE": [82144, 82144], "MATRIX_SIZE": 6747636736, "MATRIX_NNZ": 549202, "MATRIX_DENSITY": 8.13917555860553e-05, "TIME_S": 1.6988587379455566, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [39.74, 38.23, 38.66, 38.34, 38.69, 38.27, 38.61, 38.22, 38.36, 38.7], "POWER": [95.19], "JOULES": 161.71436326503752, "POWER_AFTER": [39.47, 38.44, 38.74, 38.72, 38.97, 38.52, 38.32, 38.61, 38.31, 38.32]} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_soc-sign-epinions_10000.json b/pytorch/output_cpu/epyc_7313p_10_10_soc-sign-epinions_10000.json deleted file mode 100644 index 997180b..0000000 --- a/pytorch/output_cpu/epyc_7313p_10_10_soc-sign-epinions_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 10000, "MATRIX_FILE": "soc-sign-epinions", "MATRIX_SHAPE": [131828, 131828], "MATRIX_SIZE": 17378621584, "MATRIX_NNZ": 841372, "MATRIX_DENSITY": 4.841419648464106e-05, "TIME_S": 2.818403720855713, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [40.15, 38.33, 39.01, 38.43, 38.82, 38.58, 39.15, 39.79, 38.39, 38.52], "POWER": [101.75], "JOULES": 286.7725785970688, "POWER_AFTER": [40.36, 38.72, 38.87, 38.41, 38.45, 38.49, 39.37, 38.38, 38.82, 38.92]} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_sx-mathoverflow_10000.json b/pytorch/output_cpu/epyc_7313p_10_10_sx-mathoverflow_10000.json deleted file mode 100644 index 6ff876d..0000000 --- a/pytorch/output_cpu/epyc_7313p_10_10_sx-mathoverflow_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 10000, "MATRIX_FILE": "sx-mathoverflow", "MATRIX_SHAPE": [24818, 24818], "MATRIX_SIZE": 615933124, "MATRIX_NNZ": 239978, "MATRIX_DENSITY": 0.00038961697406616504, "TIME_S": 0.9492478370666504, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [39.95, 38.76, 38.89, 38.34, 38.41, 38.56, 38.77, 38.39, 43.82, 38.34], "POWER": [85.3], "JOULES": 80.97084050178528, "POWER_AFTER": [40.98, 38.86, 38.38, 38.62, 38.93, 38.52, 39.17, 38.59, 38.99, 38.46]} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_tn2010_10000.json b/pytorch/output_cpu/epyc_7313p_10_10_tn2010_10000.json deleted file mode 100644 index 567a51c..0000000 --- a/pytorch/output_cpu/epyc_7313p_10_10_tn2010_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 10000, "MATRIX_FILE": "tn2010", "MATRIX_SHAPE": [240116, 240116], "MATRIX_SIZE": 57655693456, "MATRIX_NNZ": 1193966, "MATRIX_DENSITY": 2.070855328296721e-05, "TIME_S": 1.7836709022521973, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [40.32, 39.68, 39.86, 39.48, 38.52, 38.29, 40.37, 38.34, 39.45, 38.27], "POWER": [106.23], "JOULES": 189.47935994625092, "POWER_AFTER": [41.93, 39.25, 38.74, 39.37, 39.14, 39.52, 38.5, 38.58, 38.4, 39.3]} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_ut2010_10000.json b/pytorch/output_cpu/epyc_7313p_10_10_ut2010_10000.json deleted file mode 100644 index 5a910a7..0000000 --- a/pytorch/output_cpu/epyc_7313p_10_10_ut2010_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 10000, "MATRIX_FILE": "ut2010", "MATRIX_SHAPE": [115406, 115406], "MATRIX_SIZE": 13318544836, "MATRIX_NNZ": 572066, "MATRIX_DENSITY": 4.295259032005559e-05, "TIME_S": 0.7757325172424316, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [39.73, 39.13, 38.38, 39.43, 38.36, 39.7, 43.3, 38.81, 38.54, 39.24], "POWER": [90.41], "JOULES": 70.13397688388824, "POWER_AFTER": [40.58, 38.4, 39.23, 38.86, 38.48, 38.28, 39.25, 38.5, 40.62, 44.86]} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_va2010_10000.json b/pytorch/output_cpu/epyc_7313p_10_10_va2010_10000.json deleted file mode 100644 index 7c1a790..0000000 --- a/pytorch/output_cpu/epyc_7313p_10_10_va2010_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 10000, "MATRIX_FILE": "va2010", "MATRIX_SHAPE": [285762, 285762], "MATRIX_SIZE": 81659920644, "MATRIX_NNZ": 1402128, "MATRIX_DENSITY": 1.717033263003816e-05, "TIME_S": 2.389526844024658, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [42.21, 38.46, 39.09, 39.12, 38.6, 38.8, 38.74, 38.74, 38.51, 38.8], "POWER": [112.07], "JOULES": 267.7942734098434, "POWER_AFTER": [41.44, 38.92, 38.62, 39.01, 38.95, 38.72, 39.78, 38.59, 38.49, 38.56]} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_vt2010_10000.json b/pytorch/output_cpu/epyc_7313p_10_10_vt2010_10000.json deleted file mode 100644 index c68bb46..0000000 --- a/pytorch/output_cpu/epyc_7313p_10_10_vt2010_10000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 10000, "MATRIX_FILE": "vt2010", "MATRIX_SHAPE": [32580, 32580], "MATRIX_SIZE": 1061456400, "MATRIX_NNZ": 155598, "MATRIX_DENSITY": 0.00014658915806621921, "TIME_S": 0.8104038238525391, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [41.86, 43.27, 40.2, 39.3, 38.73, 39.46, 39.11, 39.26, 38.8, 38.85], "POWER": [81.98], "JOULES": 66.43690547943116, "POWER_AFTER": [41.6, 38.57, 39.45, 38.33, 39.52, 44.29, 38.68, 38.83, 39.46, 38.86]} diff --git a/pytorch/output_test/altra_1_1_ASIC_680k_1000.json b/pytorch/output_test/altra_1_1_ASIC_680k_1000.json deleted file mode 100644 index 6a7de8a..0000000 --- a/pytorch/output_test/altra_1_1_ASIC_680k_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 1000, "BASELINE_TIME_S": 1, "BASELINE_DELAY_S": 1, "POWER_BEFORE": [21.12], "MATRIX_FILE": "ASIC_680k", "MATRIX_SHAPE": [682862, 682862], "MATRIX_SIZE": 466300511044, "MATRIX_NNZ": 3871773, "MATRIX_DENSITY": 8.303171256088674e-06, "TIME_S": 38.89516997337341, "POWER": [88.68, 87.04, 72.4, 72.4, 60.68, 47.72, 46.48, 52.44, 68.08, 80.16, 86.8, 88.68, 85.76, 86.04, 84.96, 83.44, 83.44, 83.0, 82.0, 81.84, 82.36, 82.6, 85.52, 87.52, 86.68, 85.76, 85.04, 84.72, 86.24, 86.24, 87.44, 89.2, 88.92, 89.76, 88.92, 91.56, 90.0, 89.8, 89.08, 89.36, 89.52, 89.4, 89.4], "JOULES": 3207.1881956195834, "POWER_AFTER": [92.84]} diff --git a/pytorch/output_test/altra_1_1_ASIC_680k_1000.output b/pytorch/output_test/altra_1_1_ASIC_680k_1000.output deleted file mode 100644 index 1d92614..0000000 --- a/pytorch/output_test/altra_1_1_ASIC_680k_1000.output +++ /dev/null @@ -1,9 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3438785 queued and waiting for resources -srun: job 3438785 has been allocated resources diff --git a/pytorch/output_test/altra_1_1_Oregon-2_1000.json b/pytorch/output_test/altra_1_1_Oregon-2_1000.json deleted file mode 100644 index e5fe567..0000000 --- a/pytorch/output_test/altra_1_1_Oregon-2_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 1000, "BASELINE_TIME_S": 1, "BASELINE_DELAY_S": 1, "POWER_BEFORE": [22.36], "MATRIX_FILE": "Oregon-2", "MATRIX_SHAPE": [11806, 11806], "MATRIX_SIZE": 139381636, "MATRIX_NNZ": 65460, "MATRIX_DENSITY": 0.0004696458003979807, "TIME_S": 1.5269460678100586, "POWER": [32.68, 30.8, 30.8, 30.44, 30.48, 26.92], "JOULES": 44.66538814544678, "POWER_AFTER": [29.32]} diff --git a/pytorch/output_test/altra_1_1_Oregon-2_1000.output b/pytorch/output_test/altra_1_1_Oregon-2_1000.output deleted file mode 100644 index 2311ce7..0000000 --- a/pytorch/output_test/altra_1_1_Oregon-2_1000.output +++ /dev/null @@ -1,9 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3438787 queued and waiting for resources -srun: job 3438787 has been allocated resources diff --git a/pytorch/output_test/altra_1_1_as-caida_1000.json b/pytorch/output_test/altra_1_1_as-caida_1000.json deleted file mode 100644 index 742e5fa..0000000 --- a/pytorch/output_test/altra_1_1_as-caida_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 1000, "BASELINE_TIME_S": 1, "BASELINE_DELAY_S": 1, "POWER_BEFORE": [20.12], "MATRIX_FILE": "as-caida", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 106762, "MATRIX_DENSITY": 0.00010842726485909405, "TIME_S": 2.7389426231384277, "POWER": [28.4, 27.64, 27.16, 28.52, 28.52, 25.56, 27.8], "JOULES": 74.62260492324829, "POWER_AFTER": [27.84]} diff --git a/pytorch/output_test/altra_1_1_as-caida_1000.output b/pytorch/output_test/altra_1_1_as-caida_1000.output deleted file mode 100644 index b8e2fd7..0000000 --- a/pytorch/output_test/altra_1_1_as-caida_1000.output +++ /dev/null @@ -1,9 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3438495 queued and waiting for resources -srun: job 3438495 has been allocated resources diff --git a/pytorch/output_test/altra_1_1_dc2_1000.json b/pytorch/output_test/altra_1_1_dc2_1000.json deleted file mode 100644 index 41d784c..0000000 --- a/pytorch/output_test/altra_1_1_dc2_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 1000, "BASELINE_TIME_S": 1, "BASELINE_DELAY_S": 1, "POWER_BEFORE": [18.96], "MATRIX_FILE": "dc2", "MATRIX_SHAPE": [116835, 116835], "MATRIX_SIZE": 13650417225, "MATRIX_NNZ": 766396, "MATRIX_DENSITY": 5.614451099680581e-05, "TIME_S": 14.711330890655518, "POWER": [84.72, 84.68, 73.76, 58.68, 58.68, 45.88, 47.68, 53.28, 67.6, 79.68, 86.44, 84.28, 81.12, 80.52, 81.0, 81.88, 81.88, 85.08, 87.32], "JOULES": 1077.1134133720398, "POWER_AFTER": [89.84]} diff --git a/pytorch/output_test/altra_1_1_dc2_1000.output b/pytorch/output_test/altra_1_1_dc2_1000.output deleted file mode 100644 index a0b0842..0000000 --- a/pytorch/output_test/altra_1_1_dc2_1000.output +++ /dev/null @@ -1,9 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3438493 queued and waiting for resources -srun: job 3438493 has been allocated resources diff --git a/pytorch/output_test/altra_1_1_de2010_1000.json b/pytorch/output_test/altra_1_1_de2010_1000.json deleted file mode 100644 index 49e2828..0000000 --- a/pytorch/output_test/altra_1_1_de2010_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 1000, "BASELINE_TIME_S": 1, "BASELINE_DELAY_S": 1, "POWER_BEFORE": [16.04], "MATRIX_FILE": "de2010", "MATRIX_SHAPE": [24115, 24115], "MATRIX_SIZE": 581533225, "MATRIX_NNZ": 116056, "MATRIX_DENSITY": 0.0001995689928120616, "TIME_S": 2.6967580318450928, "POWER": [25.48, 25.64, 25.68, 27.4, 22.08, 25.52, 25.28], "JOULES": 65.21404304504394, "POWER_AFTER": [24.72]} diff --git a/pytorch/output_test/altra_1_1_de2010_1000.output b/pytorch/output_test/altra_1_1_de2010_1000.output deleted file mode 100644 index 1cb2ec9..0000000 --- a/pytorch/output_test/altra_1_1_de2010_1000.output +++ /dev/null @@ -1,9 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3438490 queued and waiting for resources -srun: job 3438490 has been allocated resources diff --git a/pytorch/output_test/altra_1_1_email-Enron_1000.json b/pytorch/output_test/altra_1_1_email-Enron_1000.json deleted file mode 100644 index 8071df9..0000000 --- a/pytorch/output_test/altra_1_1_email-Enron_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 1000, "BASELINE_TIME_S": 1, "BASELINE_DELAY_S": 1, "POWER_BEFORE": [20.8], "MATRIX_FILE": "email-Enron", "MATRIX_SHAPE": [36692, 36692], "MATRIX_SIZE": 1346302864, "MATRIX_NNZ": 367662, "MATRIX_DENSITY": 0.0002730901120626302, "TIME_S": 25.368478298187256, "POWER": [81.32, 79.32, 70.6, 70.6, 56.96, 45.92, 47.56, 54.48, 66.76, 78.76, 85.92, 85.36, 84.48, 84.0, 82.68, 82.68, 81.92, 81.0, 80.36, 80.68, 81.2, 81.6, 81.96, 81.84, 82.32, 81.84, 81.24, 80.52, 80.52], "JOULES": 1932.3098725700374, "POWER_AFTER": [79.24]} diff --git a/pytorch/output_test/altra_1_1_email-Enron_1000.output b/pytorch/output_test/altra_1_1_email-Enron_1000.output deleted file mode 100644 index c1dc4e1..0000000 --- a/pytorch/output_test/altra_1_1_email-Enron_1000.output +++ /dev/null @@ -1,9 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3438478 queued and waiting for resources -srun: job 3438478 has been allocated resources diff --git a/pytorch/output_test/altra_1_1_fl2010_1000.json b/pytorch/output_test/altra_1_1_fl2010_1000.json deleted file mode 100644 index f9f0768..0000000 --- a/pytorch/output_test/altra_1_1_fl2010_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 1000, "BASELINE_TIME_S": 1, "BASELINE_DELAY_S": 1, "POWER_BEFORE": [20.8], "MATRIX_FILE": "fl2010", "MATRIX_SHAPE": [484481, 484481], "MATRIX_SIZE": 234721839361, "MATRIX_NNZ": 2346294, "MATRIX_DENSITY": 9.99606174861054e-06, "TIME_S": 16.032954692840576, "POWER": [91.2, 90.4, 87.2, 72.2, 59.4, 53.24, 53.04, 63.84, 79.04, 92.48, 95.56, 94.04, 94.04, 93.56, 95.72, 96.2, 99.88, 97.32, 93.28, 92.64, 89.12], "JOULES": 1356.2169222259522, "POWER_AFTER": [87.04]} diff --git a/pytorch/output_test/altra_1_1_fl2010_1000.output b/pytorch/output_test/altra_1_1_fl2010_1000.output deleted file mode 100644 index dd22362..0000000 --- a/pytorch/output_test/altra_1_1_fl2010_1000.output +++ /dev/null @@ -1,9 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3438642 queued and waiting for resources -srun: job 3438642 has been allocated resources diff --git a/pytorch/output_test/altra_1_1_ga2010_1000.json b/pytorch/output_test/altra_1_1_ga2010_1000.json deleted file mode 100644 index 618418a..0000000 --- a/pytorch/output_test/altra_1_1_ga2010_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 1000, "BASELINE_TIME_S": 1, "BASELINE_DELAY_S": 1, "POWER_BEFORE": [20.96], "MATRIX_FILE": "ga2010", "MATRIX_SHAPE": [291086, 291086], "MATRIX_SIZE": 84731059396, "MATRIX_NNZ": 1418056, "MATRIX_DENSITY": 1.6735964475229304e-05, "TIME_S": 15.640851259231567, "POWER": [84.32, 84.68, 70.76, 59.4, 59.4, 45.64, 43.08, 49.32, 60.56, 76.04, 84.84, 85.52, 86.44, 87.04, 84.72, 84.2, 84.2, 83.56, 82.8, 84.16], "JOULES": 1151.2940419769286, "POWER_AFTER": [84.72]} diff --git a/pytorch/output_test/altra_1_1_ga2010_1000.output b/pytorch/output_test/altra_1_1_ga2010_1000.output deleted file mode 100644 index 02c5aae..0000000 --- a/pytorch/output_test/altra_1_1_ga2010_1000.output +++ /dev/null @@ -1,9 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3438480 queued and waiting for resources -srun: job 3438480 has been allocated resources diff --git a/pytorch/output_test/altra_1_1_mac_econ_fwd500_1000.json b/pytorch/output_test/altra_1_1_mac_econ_fwd500_1000.json deleted file mode 100644 index 50734d9..0000000 --- a/pytorch/output_test/altra_1_1_mac_econ_fwd500_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 1000, "BASELINE_TIME_S": 1, "BASELINE_DELAY_S": 1, "POWER_BEFORE": [18.2], "MATRIX_FILE": "mac_econ_fwd500", "MATRIX_SHAPE": [206500, 206500], "MATRIX_SIZE": 42642250000, "MATRIX_NNZ": 1273389, "MATRIX_DENSITY": 2.9862143765866013e-05, "TIME_S": 14.637847661972046, "POWER": [82.52, 82.52, 75.04, 62.96, 51.36, 50.12, 52.44, 66.8, 66.8, 79.88, 89.84, 87.12, 85.96, 83.6, 82.2, 82.2, 81.8, 82.24, 82.0], "JOULES": 1094.663508281708, "POWER_AFTER": [81.8]} diff --git a/pytorch/output_test/altra_1_1_mac_econ_fwd500_1000.output b/pytorch/output_test/altra_1_1_mac_econ_fwd500_1000.output deleted file mode 100644 index 1b60785..0000000 --- a/pytorch/output_test/altra_1_1_mac_econ_fwd500_1000.output +++ /dev/null @@ -1,9 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3438629 queued and waiting for resources -srun: job 3438629 has been allocated resources diff --git a/pytorch/output_test/altra_1_1_mc2depi_1000.json b/pytorch/output_test/altra_1_1_mc2depi_1000.json deleted file mode 100644 index 51e2631..0000000 --- a/pytorch/output_test/altra_1_1_mc2depi_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 1000, "BASELINE_TIME_S": 1, "BASELINE_DELAY_S": 1, "POWER_BEFORE": [22.52], "MATRIX_FILE": "mc2depi", "MATRIX_SHAPE": [525825, 525825], "MATRIX_SIZE": 276491930625, "MATRIX_NNZ": 2100225, "MATRIX_DENSITY": 7.595972132902821e-06, "TIME_S": 13.071983337402344, "POWER": [85.64, 80.8, 80.8, 70.16, 56.48, 49.32, 47.12, 57.88, 72.04, 82.2, 87.44, 87.4, 88.44, 90.52, 90.52, 89.68, 90.96], "JOULES": 975.7476043701172, "POWER_AFTER": [89.12]} diff --git a/pytorch/output_test/altra_1_1_mc2depi_1000.output b/pytorch/output_test/altra_1_1_mc2depi_1000.output deleted file mode 100644 index 0e9b1d3..0000000 --- a/pytorch/output_test/altra_1_1_mc2depi_1000.output +++ /dev/null @@ -1,9 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3438488 queued and waiting for resources -srun: job 3438488 has been allocated resources diff --git a/pytorch/output_test/altra_1_1_p2p-Gnutella04_1000.json b/pytorch/output_test/altra_1_1_p2p-Gnutella04_1000.json deleted file mode 100644 index 6ca9dfb..0000000 --- a/pytorch/output_test/altra_1_1_p2p-Gnutella04_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 1000, "BASELINE_TIME_S": 1, "BASELINE_DELAY_S": 1, "POWER_BEFORE": [20.12], "MATRIX_FILE": "p2p-Gnutella04", "MATRIX_SHAPE": [10879, 10879], "MATRIX_SIZE": 118352641, "MATRIX_NNZ": 39994, "MATRIX_DENSITY": 0.0003379223282393842, "TIME_S": 1.0336177349090576, "POWER": [25.32, 30.88, 31.2, 33.68, 33.0], "JOULES": 34.7893852519989, "POWER_AFTER": [32.44]} diff --git a/pytorch/output_test/altra_1_1_p2p-Gnutella04_1000.output b/pytorch/output_test/altra_1_1_p2p-Gnutella04_1000.output deleted file mode 100644 index 4859919..0000000 --- a/pytorch/output_test/altra_1_1_p2p-Gnutella04_1000.output +++ /dev/null @@ -1,9 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3438627 queued and waiting for resources -srun: job 3438627 has been allocated resources diff --git a/pytorch/output_test/altra_1_1_p2p-Gnutella24_1000.json b/pytorch/output_test/altra_1_1_p2p-Gnutella24_1000.json deleted file mode 100644 index 2184e9d..0000000 --- a/pytorch/output_test/altra_1_1_p2p-Gnutella24_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 1000, "BASELINE_TIME_S": 1, "BASELINE_DELAY_S": 1, "POWER_BEFORE": [17.92], "MATRIX_FILE": "p2p-Gnutella24", "MATRIX_SHAPE": [26518, 26518], "MATRIX_SIZE": 703204324, "MATRIX_NNZ": 65369, "MATRIX_DENSITY": 9.295875717624285e-05, "TIME_S": 1.6974623203277588, "POWER": [26.32, 26.48, 27.24, 27.2, 25.12, 27.44], "JOULES": 44.2583660697937, "POWER_AFTER": [26.96]} diff --git a/pytorch/output_test/altra_1_1_p2p-Gnutella24_1000.output b/pytorch/output_test/altra_1_1_p2p-Gnutella24_1000.output deleted file mode 100644 index b113ec6..0000000 --- a/pytorch/output_test/altra_1_1_p2p-Gnutella24_1000.output +++ /dev/null @@ -1,9 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3438786 queued and waiting for resources -srun: job 3438786 has been allocated resources diff --git a/pytorch/output_test/altra_1_1_p2p-Gnutella25_1000.json b/pytorch/output_test/altra_1_1_p2p-Gnutella25_1000.json deleted file mode 100644 index d154b91..0000000 --- a/pytorch/output_test/altra_1_1_p2p-Gnutella25_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 1000, "BASELINE_TIME_S": 1, "BASELINE_DELAY_S": 1, "POWER_BEFORE": [18.88], "MATRIX_FILE": "p2p-Gnutella25", "MATRIX_SHAPE": [22687, 22687], "MATRIX_SIZE": 514699969, "MATRIX_NNZ": 54705, "MATRIX_DENSITY": 0.00010628522108964806, "TIME_S": 1.4318811893463135, "POWER": [26.16, 28.4, 27.68, 30.2, 28.92], "JOULES": 42.69000399589539, "POWER_AFTER": [27.88]} diff --git a/pytorch/output_test/altra_1_1_p2p-Gnutella25_1000.output b/pytorch/output_test/altra_1_1_p2p-Gnutella25_1000.output deleted file mode 100644 index 2eec2d5..0000000 --- a/pytorch/output_test/altra_1_1_p2p-Gnutella25_1000.output +++ /dev/null @@ -1,9 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3438623 queued and waiting for resources -srun: job 3438623 has been allocated resources diff --git a/pytorch/output_test/altra_1_1_p2p-Gnutella30_1000.json b/pytorch/output_test/altra_1_1_p2p-Gnutella30_1000.json deleted file mode 100644 index e06f6c1..0000000 --- a/pytorch/output_test/altra_1_1_p2p-Gnutella30_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 1000, "BASELINE_TIME_S": 1, "BASELINE_DELAY_S": 1, "POWER_BEFORE": [22.92], "MATRIX_FILE": "p2p-Gnutella30", "MATRIX_SHAPE": [36682, 36682], "MATRIX_SIZE": 1345569124, "MATRIX_NNZ": 88328, "MATRIX_DENSITY": 6.564359899804003e-05, "TIME_S": 3.1562492847442627, "POWER": [68.48, 82.16, 77.4, 65.64, 54.88, 50.28, 53.56], "JOULES": 179.1687116909027, "POWER_AFTER": [63.16]} diff --git a/pytorch/output_test/altra_1_1_p2p-Gnutella30_1000.output b/pytorch/output_test/altra_1_1_p2p-Gnutella30_1000.output deleted file mode 100644 index 00e2d2b..0000000 --- a/pytorch/output_test/altra_1_1_p2p-Gnutella30_1000.output +++ /dev/null @@ -1,9 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3438639 queued and waiting for resources -srun: job 3438639 has been allocated resources diff --git a/pytorch/output_test/altra_1_1_ri2010_1000.json b/pytorch/output_test/altra_1_1_ri2010_1000.json deleted file mode 100644 index fddc478..0000000 --- a/pytorch/output_test/altra_1_1_ri2010_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 1000, "BASELINE_TIME_S": 1, "BASELINE_DELAY_S": 1, "POWER_BEFORE": [18.76], "MATRIX_FILE": "ri2010", "MATRIX_SHAPE": [25181, 25181], "MATRIX_SIZE": 634082761, "MATRIX_NNZ": 125750, "MATRIX_DENSITY": 0.00019831796057928155, "TIME_S": 3.022062301635742, "POWER": [27.72, 27.72, 28.28, 25.68, 23.96, 28.76, 28.68], "JOULES": 79.0327468109131, "POWER_AFTER": [27.96]} diff --git a/pytorch/output_test/altra_1_1_ri2010_1000.output b/pytorch/output_test/altra_1_1_ri2010_1000.output deleted file mode 100644 index 853f806..0000000 --- a/pytorch/output_test/altra_1_1_ri2010_1000.output +++ /dev/null @@ -1,9 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3438486 queued and waiting for resources -srun: job 3438486 has been allocated resources diff --git a/pytorch/output_test/altra_1_1_rma10_1000.json b/pytorch/output_test/altra_1_1_rma10_1000.json deleted file mode 100644 index dfdcbb4..0000000 --- a/pytorch/output_test/altra_1_1_rma10_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 1000, "BASELINE_TIME_S": 1, "BASELINE_DELAY_S": 1, "POWER_BEFORE": [21.12], "MATRIX_FILE": "rma10", "MATRIX_SHAPE": [46835, 46835], "MATRIX_SIZE": 2193517225, "MATRIX_NNZ": 2374001, "MATRIX_DENSITY": 0.0010822805369125833, "TIME_S": 61.10852575302124, "POWER": [81.32, 83.24, 72.48, 60.04, 60.04, 48.64, 44.88, 49.6, 63.44, 75.2, 84.4, 86.48, 85.8, 85.56, 86.0, 85.96, 85.56, 85.56, 85.0, 83.88, 83.36, 82.84, 81.96, 81.52, 81.32, 80.96, 80.48, 80.8, 81.0, 81.0, 81.36, 81.6, 82.48, 83.04, 83.92, 83.76, 83.76, 83.72, 83.12, 82.64, 82.64, 82.4, 82.4, 82.84, 82.88, 83.04, 83.36, 83.24, 83.4, 82.76, 82.56, 83.16, 83.8, 84.36, 84.36, 84.96, 83.72, 82.64, 82.68, 82.08, 83.04, 84.2, 85.52, 85.12, 84.4], "JOULES": 4910.999573554994, "POWER_AFTER": [83.6]} diff --git a/pytorch/output_test/altra_1_1_rma10_1000.output b/pytorch/output_test/altra_1_1_rma10_1000.output deleted file mode 100644 index f9e3a1e..0000000 --- a/pytorch/output_test/altra_1_1_rma10_1000.output +++ /dev/null @@ -1,9 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3438624 queued and waiting for resources -srun: job 3438624 has been allocated resources diff --git a/pytorch/output_test/altra_1_1_soc-sign-Slashdot090216_1000.json b/pytorch/output_test/altra_1_1_soc-sign-Slashdot090216_1000.json deleted file mode 100644 index e9d1035..0000000 --- a/pytorch/output_test/altra_1_1_soc-sign-Slashdot090216_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 1000, "BASELINE_TIME_S": 1, "BASELINE_DELAY_S": 1, "POWER_BEFORE": [19.84], "MATRIX_FILE": "soc-sign-Slashdot090216", "MATRIX_SHAPE": [81871, 81871], "MATRIX_SIZE": 6702860641, "MATRIX_NNZ": 545671, "MATRIX_DENSITY": 8.140867447881048e-05, "TIME_S": 24.203877449035645, "POWER": [81.28, 79.56, 79.56, 70.4, 57.16, 46.28, 40.0, 47.52, 59.32, 71.08, 85.24, 83.36, 83.24, 82.88, 82.6, 82.6, 82.6, 83.04, 82.84, 82.8, 83.68, 83.08, 82.84, 82.56, 82.4, 81.44, 82.08, 81.76, 81.76], "JOULES": 1829.069020233154, "POWER_AFTER": [82.04]} diff --git a/pytorch/output_test/altra_1_1_soc-sign-Slashdot090216_1000.output b/pytorch/output_test/altra_1_1_soc-sign-Slashdot090216_1000.output deleted file mode 100644 index ed24ef1..0000000 --- a/pytorch/output_test/altra_1_1_soc-sign-Slashdot090216_1000.output +++ /dev/null @@ -1,9 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3438788 queued and waiting for resources -srun: job 3438788 has been allocated resources diff --git a/pytorch/output_test/altra_1_1_soc-sign-Slashdot090221_1000.json b/pytorch/output_test/altra_1_1_soc-sign-Slashdot090221_1000.json deleted file mode 100644 index 3cf231a..0000000 --- a/pytorch/output_test/altra_1_1_soc-sign-Slashdot090221_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 1000, "BASELINE_TIME_S": 1, "BASELINE_DELAY_S": 1, "POWER_BEFORE": [20.64], "MATRIX_FILE": "soc-sign-Slashdot090221", "MATRIX_SHAPE": [82144, 82144], "MATRIX_SIZE": 6747636736, "MATRIX_NNZ": 549202, "MATRIX_DENSITY": 8.13917555860553e-05, "TIME_S": 18.71693205833435, "POWER": [80.68, 80.36, 77.12, 66.12, 55.68, 48.96, 51.56, 62.24, 75.0, 90.36, 90.64, 90.64, 88.92, 87.52, 86.76, 86.08, 85.48, 85.08, 84.52, 83.8, 83.84, 83.6, 82.96], "JOULES": 1480.1566835594176, "POWER_AFTER": [82.52]} diff --git a/pytorch/output_test/altra_1_1_soc-sign-Slashdot090221_1000.output b/pytorch/output_test/altra_1_1_soc-sign-Slashdot090221_1000.output deleted file mode 100644 index 1c83cee..0000000 --- a/pytorch/output_test/altra_1_1_soc-sign-Slashdot090221_1000.output +++ /dev/null @@ -1,9 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3438477 queued and waiting for resources -srun: job 3438477 has been allocated resources diff --git a/pytorch/output_test/altra_1_1_soc-sign-epinions_1000.json b/pytorch/output_test/altra_1_1_soc-sign-epinions_1000.json deleted file mode 100644 index 0229d0b..0000000 --- a/pytorch/output_test/altra_1_1_soc-sign-epinions_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 1000, "BASELINE_TIME_S": 1, "BASELINE_DELAY_S": 1, "POWER_BEFORE": [22.52], "MATRIX_FILE": "soc-sign-epinions", "MATRIX_SHAPE": [131828, 131828], "MATRIX_SIZE": 17378621584, "MATRIX_NNZ": 841372, "MATRIX_DENSITY": 4.841419648464106e-05, "TIME_S": 35.66161274909973, "POWER": [84.68, 84.68, 76.2, 64.48, 50.48, 50.48, 49.08, 51.0, 64.36, 75.28, 87.4, 86.08, 85.84, 84.88, 84.4, 82.84, 83.2, 83.2, 81.48, 82.04, 82.48, 82.84, 82.96, 83.04, 81.88, 80.92, 81.6, 81.68, 82.28, 82.28, 83.32, 84.04, 83.36, 83.76, 83.56, 83.12, 83.2, 83.84, 83.48, 83.96], "JOULES": 2811.229006414413, "POWER_AFTER": [83.96]} diff --git a/pytorch/output_test/altra_1_1_soc-sign-epinions_1000.output b/pytorch/output_test/altra_1_1_soc-sign-epinions_1000.output deleted file mode 100644 index 55154e1..0000000 --- a/pytorch/output_test/altra_1_1_soc-sign-epinions_1000.output +++ /dev/null @@ -1,9 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3438634 queued and waiting for resources -srun: job 3438634 has been allocated resources diff --git a/pytorch/output_test/altra_1_1_sx-mathoverflow_1000.json b/pytorch/output_test/altra_1_1_sx-mathoverflow_1000.json deleted file mode 100644 index 8904c29..0000000 --- a/pytorch/output_test/altra_1_1_sx-mathoverflow_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 1000, "BASELINE_TIME_S": 1, "BASELINE_DELAY_S": 1, "POWER_BEFORE": [19.96], "MATRIX_FILE": "sx-mathoverflow", "MATRIX_SHAPE": [24818, 24818], "MATRIX_SIZE": 615933124, "MATRIX_NNZ": 239978, "MATRIX_DENSITY": 0.00038961697406616504, "TIME_S": 5.487327575683594, "POWER": [30.32, 24.48, 25.2, 25.24, 28.44, 28.44, 31.04, 30.6, 30.12], "JOULES": 158.43830657958983, "POWER_AFTER": [24.16]} diff --git a/pytorch/output_test/altra_1_1_sx-mathoverflow_1000.output b/pytorch/output_test/altra_1_1_sx-mathoverflow_1000.output deleted file mode 100644 index be3be8e..0000000 --- a/pytorch/output_test/altra_1_1_sx-mathoverflow_1000.output +++ /dev/null @@ -1,9 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3438492 queued and waiting for resources -srun: job 3438492 has been allocated resources diff --git a/pytorch/output_test/altra_1_1_tn2010_1000.json b/pytorch/output_test/altra_1_1_tn2010_1000.json deleted file mode 100644 index 030080e..0000000 --- a/pytorch/output_test/altra_1_1_tn2010_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 1000, "BASELINE_TIME_S": 1, "BASELINE_DELAY_S": 1, "POWER_BEFORE": [18.72], "MATRIX_FILE": "tn2010", "MATRIX_SHAPE": [240116, 240116], "MATRIX_SIZE": 57655693456, "MATRIX_NNZ": 1193966, "MATRIX_DENSITY": 2.070855328296721e-05, "TIME_S": 14.758731126785278, "POWER": [79.48, 80.44, 69.4, 55.16, 43.84, 43.84, 42.32, 47.56, 62.32, 75.6, 84.64, 85.68, 86.92, 87.08, 88.16, 89.0, 91.12, 90.4, 90.4], "JOULES": 1087.0692938613893, "POWER_AFTER": [88.36]} diff --git a/pytorch/output_test/altra_1_1_tn2010_1000.output b/pytorch/output_test/altra_1_1_tn2010_1000.output deleted file mode 100644 index ad48df5..0000000 --- a/pytorch/output_test/altra_1_1_tn2010_1000.output +++ /dev/null @@ -1,9 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3438483 queued and waiting for resources -srun: job 3438483 has been allocated resources diff --git a/pytorch/output_test/altra_1_1_ut2010_1000.json b/pytorch/output_test/altra_1_1_ut2010_1000.json deleted file mode 100644 index 85d1844..0000000 --- a/pytorch/output_test/altra_1_1_ut2010_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 1000, "BASELINE_TIME_S": 1, "BASELINE_DELAY_S": 1, "POWER_BEFORE": [18.72], "MATRIX_FILE": "ut2010", "MATRIX_SHAPE": [115406, 115406], "MATRIX_SIZE": 13318544836, "MATRIX_NNZ": 572066, "MATRIX_DENSITY": 4.295259032005559e-05, "TIME_S": 8.555217266082764, "POWER": [85.92, 85.56, 78.12, 65.56, 52.08, 41.52, 42.52, 54.36, 65.72, 81.28, 86.72, 87.32, 87.32], "JOULES": 560.0015716743469, "POWER_AFTER": [88.56]} diff --git a/pytorch/output_test/altra_1_1_ut2010_1000.output b/pytorch/output_test/altra_1_1_ut2010_1000.output deleted file mode 100644 index 0b8e6d0..0000000 --- a/pytorch/output_test/altra_1_1_ut2010_1000.output +++ /dev/null @@ -1,9 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3438638 queued and waiting for resources -srun: job 3438638 has been allocated resources diff --git a/pytorch/output_test/altra_1_1_va2010_1000.json b/pytorch/output_test/altra_1_1_va2010_1000.json deleted file mode 100644 index d37430c..0000000 --- a/pytorch/output_test/altra_1_1_va2010_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 1000, "BASELINE_TIME_S": 1, "BASELINE_DELAY_S": 1, "POWER_BEFORE": [22.12], "MATRIX_FILE": "va2010", "MATRIX_SHAPE": [285762, 285762], "MATRIX_SIZE": 81659920644, "MATRIX_NNZ": 1402128, "MATRIX_DENSITY": 1.717033263003816e-05, "TIME_S": 13.728005409240723, "POWER": [95.56, 96.48, 88.08, 76.2, 60.44, 54.28, 54.48, 67.72, 83.28, 83.28, 96.92, 94.84, 93.96, 94.76, 93.24, 94.48, 96.56, 95.28], "JOULES": 1137.604355392456, "POWER_AFTER": [97.44]} diff --git a/pytorch/output_test/altra_1_1_va2010_1000.output b/pytorch/output_test/altra_1_1_va2010_1000.output deleted file mode 100644 index 66c3958..0000000 --- a/pytorch/output_test/altra_1_1_va2010_1000.output +++ /dev/null @@ -1,9 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3438632 queued and waiting for resources -srun: job 3438632 has been allocated resources diff --git a/pytorch/output_test/altra_1_1_vt2010_1000.json b/pytorch/output_test/altra_1_1_vt2010_1000.json deleted file mode 100644 index 37e3723..0000000 --- a/pytorch/output_test/altra_1_1_vt2010_1000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 1000, "BASELINE_TIME_S": 1, "BASELINE_DELAY_S": 1, "POWER_BEFORE": [25.84], "MATRIX_FILE": "vt2010", "MATRIX_SHAPE": [32580, 32580], "MATRIX_SIZE": 1061456400, "MATRIX_NNZ": 155598, "MATRIX_DENSITY": 0.00014658915806621921, "TIME_S": 4.005021333694458, "POWER": [30.28, 30.88, 28.92, 27.56, 27.68, 27.68, 30.32, 29.28], "JOULES": 113.38702465057371, "POWER_AFTER": [29.0]} diff --git a/pytorch/output_test/altra_1_1_vt2010_1000.output b/pytorch/output_test/altra_1_1_vt2010_1000.output deleted file mode 100644 index 146b4f7..0000000 --- a/pytorch/output_test/altra_1_1_vt2010_1000.output +++ /dev/null @@ -1,9 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3438640 queued and waiting for resources -srun: job 3438640 has been allocated resources diff --git a/pytorch/output_test/altra_20_10_10_ASIC_680k.json b/pytorch/output_test/altra_20_10_10_ASIC_680k.json new file mode 100644 index 0000000..b514a37 --- /dev/null +++ b/pytorch/output_test/altra_20_10_10_ASIC_680k.json @@ -0,0 +1 @@ +{"CPU": "ALTRA", "ITERATIONS": 22291, "MATRIX_FILE": "ASIC_680k", "MATRIX_SHAPE": [682862, 682862], "MATRIX_SIZE": 466300511044, "MATRIX_NNZ": 3871773, "MATRIX_DENSITY": 8.303171256088674e-06, "TIME_S": 27.8618381023407, "TIME_S_1KI": 1.2499142300632855, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2431.3417924404143, "W": 85.83483990377374, "J_1KI": 109.0728003427578, "W_1KI": 3.8506500338151604, "W_D": 70.11083990377374, "J_D": 1985.9466779718396, "W_D_1KI": 3.145253236901608, "J_D_1KI": 0.14109969211348114} diff --git a/pytorch/output_cpu/altra_10_10_ASIC_680k_10000.output b/pytorch/output_test/altra_20_10_10_ASIC_680k.output similarity index 82% rename from pytorch/output_cpu/altra_10_10_ASIC_680k_10000.output rename to pytorch/output_test/altra_20_10_10_ASIC_680k.output index 928b345..fe3ff12 100644 --- a/pytorch/output_cpu/altra_10_10_ASIC_680k_10000.output +++ b/pytorch/output_test/altra_20_10_10_ASIC_680k.output @@ -5,10 +5,10 @@ srun: # All submission nodes and all other compute nodes have x86_64 architectur srun: # CPUs. Programs, environments, or other software that was built on x86_64 # srun: # nodes may need to be rebuilt to properly execute on these nodes. # srun: ################################################################################ -srun: job 3471013 queued and waiting for resources -srun: job 3471013 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) +srun: job 3476562 queued and waiting for resources +srun: job 3476562 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 3, 4, ..., 3871767, 3871770, 3871773]), col_indices=tensor([ 0, 11698, 11699, ..., 169456, 645874, @@ -16,11 +16,11 @@ tensor(crow_indices=tensor([ 0, 3, 4, ..., 3871767, values=tensor([ 3.8333e-04, -3.3333e-04, -5.0000e-05, ..., 0.0000e+00, 0.0000e+00, 7.9289e-02]), size=(682862, 682862), nnz=3871773, layout=torch.sparse_csr) -tensor([0.6902, 0.5218, 0.8924, ..., 0.0864, 0.5539, 0.5194]) +tensor([0.0745, 0.1612, 0.7336, ..., 0.2369, 0.4612, 0.1481]) Matrix: ASIC_680k Shape: torch.Size([682862, 682862]) Size: 466300511044 NNZ: 3871773 Density: 8.303171256088674e-06 -Time: 11.77456283569336 seconds +Time: 27.8618381023407 seconds diff --git a/pytorch/output_test/altra_20_10_10_Oregon-2.json b/pytorch/output_test/altra_20_10_10_Oregon-2.json new file mode 100644 index 0000000..884adfe --- /dev/null +++ b/pytorch/output_test/altra_20_10_10_Oregon-2.json @@ -0,0 +1 @@ +{"CPU": "ALTRA", "ITERATIONS": 294245, "MATRIX_FILE": "Oregon-2", "MATRIX_SHAPE": [11806, 11806], "MATRIX_SIZE": 139381636, "MATRIX_NNZ": 65460, "MATRIX_DENSITY": 0.0004696458003979807, "TIME_S": 23.451528072357178, "TIME_S_1KI": 0.07970068504938803, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1816.6004677104947, "W": 88.63538209958323, "J_1KI": 6.173768348520773, "W_1KI": 0.30122986660634243, "W_D": 69.05538209958323, "J_D": 1415.304322590828, "W_D_1KI": 0.23468667980622687, "J_D_1KI": 0.0007975893551503912} diff --git a/pytorch/output_test2/altra_10_10_Oregon-2_100000.output b/pytorch/output_test/altra_20_10_10_Oregon-2.output similarity index 80% rename from pytorch/output_test2/altra_10_10_Oregon-2_100000.output rename to pytorch/output_test/altra_20_10_10_Oregon-2.output index d2559a2..d40c5ab 100644 --- a/pytorch/output_test2/altra_10_10_Oregon-2_100000.output +++ b/pytorch/output_test/altra_20_10_10_Oregon-2.output @@ -5,19 +5,19 @@ srun: # All submission nodes and all other compute nodes have x86_64 architectur srun: # CPUs. Programs, environments, or other software that was built on x86_64 # srun: # nodes may need to be rebuilt to properly execute on these nodes. # srun: ################################################################################ -srun: job 3471857 queued and waiting for resources -srun: job 3471857 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) +srun: job 3476564 queued and waiting for resources +srun: job 3476564 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 583, 584, ..., 65459, 65460, 65460]), col_indices=tensor([ 2, 23, 27, ..., 3324, 958, 841]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(11806, 11806), nnz=65460, layout=torch.sparse_csr) -tensor([0.8374, 0.0143, 0.3251, ..., 0.2693, 0.4062, 0.8940]) +tensor([0.0871, 0.9010, 0.0267, ..., 0.4888, 0.6851, 0.5038]) Matrix: Oregon-2 Shape: torch.Size([11806, 11806]) Size: 139381636 NNZ: 65460 Density: 0.0004696458003979807 -Time: 8.373449563980103 seconds +Time: 23.451528072357178 seconds diff --git a/pytorch/output_test/altra_20_10_10_as-caida.json b/pytorch/output_test/altra_20_10_10_as-caida.json new file mode 100644 index 0000000..683382f --- /dev/null +++ b/pytorch/output_test/altra_20_10_10_as-caida.json @@ -0,0 +1 @@ +{"CPU": "ALTRA", "ITERATIONS": 243841, "MATRIX_FILE": "as-caida", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 106762, "MATRIX_DENSITY": 0.00010842726485909405, "TIME_S": 22.4771625995636, "TIME_S_1KI": 0.0921795866960995, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2017.4365046882629, "W": 86.41223761794208, "J_1KI": 8.27357378245768, "W_1KI": 0.35437944241510694, "W_D": 66.99623761794209, "J_D": 1564.137894968033, "W_D_1KI": 0.27475378471193146, "J_D_1KI": 0.0011267743517781319} diff --git a/pytorch/output_test2/altra_10_10_as-caida_100000.output b/pytorch/output_test/altra_20_10_10_as-caida.output similarity index 80% rename from pytorch/output_test2/altra_10_10_as-caida_100000.output rename to pytorch/output_test/altra_20_10_10_as-caida.output index b3a7860..7a29fd3 100644 --- a/pytorch/output_test2/altra_10_10_as-caida_100000.output +++ b/pytorch/output_test/altra_20_10_10_as-caida.output @@ -5,20 +5,20 @@ srun: # All submission nodes and all other compute nodes have x86_64 architectur srun: # CPUs. Programs, environments, or other software that was built on x86_64 # srun: # nodes may need to be rebuilt to properly execute on these nodes. # srun: ################################################################################ -srun: job 3471796 queued and waiting for resources -srun: job 3471796 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) +srun: job 3476493 queued and waiting for resources +srun: job 3476493 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 0, 0, ..., 106761, 106761, 106762]), col_indices=tensor([ 106, 329, 1040, ..., 155, 160, 12170]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379), nnz=106762, layout=torch.sparse_csr) -tensor([0.3242, 0.9198, 0.8266, ..., 0.4648, 0.8946, 0.6351]) +tensor([0.4153, 0.7846, 0.8319, ..., 0.7108, 0.4540, 0.9956]) Matrix: as-caida Shape: torch.Size([31379, 31379]) Size: 984641641 NNZ: 106762 Density: 0.00010842726485909405 -Time: 7.69922399520874 seconds +Time: 22.4771625995636 seconds diff --git a/pytorch/output_test/altra_20_10_10_dc2.json b/pytorch/output_test/altra_20_10_10_dc2.json new file mode 100644 index 0000000..00a83f0 --- /dev/null +++ b/pytorch/output_test/altra_20_10_10_dc2.json @@ -0,0 +1 @@ +{"CPU": "ALTRA", "ITERATIONS": 65670, "MATRIX_FILE": "dc2", "MATRIX_SHAPE": [116835, 116835], "MATRIX_SIZE": 13650417225, "MATRIX_NNZ": 766396, "MATRIX_DENSITY": 5.614451099680581e-05, "TIME_S": 27.732829570770264, "TIME_S_1KI": 0.4223059170210182, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2552.051627426147, "W": 84.46197815348627, "J_1KI": 38.86175768884037, "W_1KI": 1.2861577303713456, "W_D": 65.13197815348627, "J_D": 1967.9881347560884, "W_D_1KI": 0.9918071897896494, "J_D_1KI": 0.01510289614420054} diff --git a/pytorch/output_test2/altra_10_10_dc2_100000.output b/pytorch/output_test/altra_20_10_10_dc2.output similarity index 82% rename from pytorch/output_test2/altra_10_10_dc2_100000.output rename to pytorch/output_test/altra_20_10_10_dc2.output index 79c34b5..89a17d5 100644 --- a/pytorch/output_test2/altra_10_10_dc2_100000.output +++ b/pytorch/output_test/altra_20_10_10_dc2.output @@ -5,10 +5,10 @@ srun: # All submission nodes and all other compute nodes have x86_64 architectur srun: # CPUs. Programs, environments, or other software that was built on x86_64 # srun: # nodes may need to be rebuilt to properly execute on these nodes. # srun: ################################################################################ -srun: job 3471782 queued and waiting for resources -srun: job 3471782 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) +srun: job 3476488 queued and waiting for resources +srun: job 3476488 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 1, 2, ..., 766390, 766394, 766396]), col_indices=tensor([ 0, 1, 2, ..., 116833, 89, @@ -16,11 +16,11 @@ tensor(crow_indices=tensor([ 0, 1, 2, ..., 766390, 766394, values=tensor([-1.0000e+00, -1.0000e+00, -1.0000e+00, ..., 1.0331e+01, -1.0000e-03, 1.0000e-03]), size=(116835, 116835), nnz=766396, layout=torch.sparse_csr) -tensor([0.1528, 0.7657, 0.8355, ..., 0.4682, 0.1999, 0.0103]) +tensor([0.7501, 0.8783, 0.5650, ..., 0.7882, 0.9151, 0.3811]) Matrix: dc2 Shape: torch.Size([116835, 116835]) Size: 13650417225 NNZ: 766396 Density: 5.614451099680581e-05 -Time: 37.14217662811279 seconds +Time: 27.732829570770264 seconds diff --git a/pytorch/output_test/altra_20_10_10_de2010.json b/pytorch/output_test/altra_20_10_10_de2010.json new file mode 100644 index 0000000..6db94a7 --- /dev/null +++ b/pytorch/output_test/altra_20_10_10_de2010.json @@ -0,0 +1 @@ +{"CPU": "ALTRA", "ITERATIONS": 233556, "MATRIX_FILE": "de2010", "MATRIX_SHAPE": [24115, 24115], "MATRIX_SIZE": 581533225, "MATRIX_NNZ": 116056, "MATRIX_DENSITY": 0.0001995689928120616, "TIME_S": 20.548720121383667, "TIME_S_1KI": 0.08798198342745923, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1612.0595497703553, "W": 85.71045805717254, "J_1KI": 6.9022399329084045, "W_1KI": 0.36698033044397294, "W_D": 66.07445805717254, "J_D": 1242.741708787918, "W_D_1KI": 0.28290627539935836, "J_D_1KI": 0.0012112995401503638} diff --git a/pytorch/output_cpu/altra_10_10_de2010_10000.output b/pytorch/output_test/altra_20_10_10_de2010.output similarity index 81% rename from pytorch/output_cpu/altra_10_10_de2010_10000.output rename to pytorch/output_test/altra_20_10_10_de2010.output index 97dc072..382f015 100644 --- a/pytorch/output_cpu/altra_10_10_de2010_10000.output +++ b/pytorch/output_test/altra_20_10_10_de2010.output @@ -5,21 +5,21 @@ srun: # All submission nodes and all other compute nodes have x86_64 architectur srun: # CPUs. Programs, environments, or other software that was built on x86_64 # srun: # nodes may need to be rebuilt to properly execute on these nodes. # srun: ################################################################################ -srun: job 3470980 queued and waiting for resources -srun: job 3470980 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) +srun: job 3476490 queued and waiting for resources +srun: job 3476490 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 13, 21, ..., 116047, 116051, 116056]), col_indices=tensor([ 250, 251, 757, ..., 23334, 23553, 24050]), values=tensor([ 14900., 33341., 20255., ..., 164227., 52413., 16949.]), size=(24115, 24115), nnz=116056, layout=torch.sparse_csr) -tensor([0.3562, 0.7994, 0.9047, ..., 0.2891, 0.3611, 0.5704]) +tensor([0.6419, 0.2202, 0.7400, ..., 0.5620, 0.4260, 0.4035]) Matrix: de2010 Shape: torch.Size([24115, 24115]) Size: 581533225 NNZ: 116056 Density: 0.0001995689928120616 -Time: 1.1378686428070068 seconds +Time: 20.548720121383667 seconds diff --git a/pytorch/output_test/altra_20_10_10_email-Enron.json b/pytorch/output_test/altra_20_10_10_email-Enron.json new file mode 100644 index 0000000..db4c4c6 --- /dev/null +++ b/pytorch/output_test/altra_20_10_10_email-Enron.json @@ -0,0 +1 @@ +{"CPU": "ALTRA", "ITERATIONS": 190727, "MATRIX_FILE": "email-Enron", "MATRIX_SHAPE": [36692, 36692], "MATRIX_SIZE": 1346302864, "MATRIX_NNZ": 367662, "MATRIX_DENSITY": 0.0002730901120626302, "TIME_S": 24.423046112060547, "TIME_S_1KI": 0.1280523791181141, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2348.8964970970155, "W": 91.678164901281, "J_1KI": 12.315490188054211, "W_1KI": 0.4806774337208733, "W_D": 72.27316490128099, "J_D": 1851.7188258898261, "W_D_1KI": 0.3789351528691847, "J_D_1KI": 0.0019867934422980738} diff --git a/pytorch/output_test2/altra_10_10_email-Enron_100000.output b/pytorch/output_test/altra_20_10_10_email-Enron.output similarity index 80% rename from pytorch/output_test2/altra_10_10_email-Enron_100000.output rename to pytorch/output_test/altra_20_10_10_email-Enron.output index 49f40f7..fc29c73 100644 --- a/pytorch/output_test2/altra_10_10_email-Enron_100000.output +++ b/pytorch/output_test/altra_20_10_10_email-Enron.output @@ -5,20 +5,20 @@ srun: # All submission nodes and all other compute nodes have x86_64 architectur srun: # CPUs. Programs, environments, or other software that was built on x86_64 # srun: # nodes may need to be rebuilt to properly execute on these nodes. # srun: ################################################################################ -srun: job 3471786 queued and waiting for resources -srun: job 3471786 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) +srun: job 3476483 queued and waiting for resources +srun: job 3476483 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 1, 71, ..., 367660, 367661, 367662]), col_indices=tensor([ 1, 0, 2, ..., 36690, 36689, 8203]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(36692, 36692), nnz=367662, layout=torch.sparse_csr) -tensor([0.6039, 0.3557, 0.6656, ..., 0.1586, 0.2866, 0.7610]) +tensor([0.8879, 0.6844, 0.4775, ..., 0.9641, 0.8907, 0.4040]) Matrix: email-Enron Shape: torch.Size([36692, 36692]) Size: 1346302864 NNZ: 367662 Density: 0.0002730901120626302 -Time: 12.88691234588623 seconds +Time: 24.423046112060547 seconds diff --git a/pytorch/output_test/altra_20_10_10_fl2010.json b/pytorch/output_test/altra_20_10_10_fl2010.json new file mode 100644 index 0000000..645c7e5 --- /dev/null +++ b/pytorch/output_test/altra_20_10_10_fl2010.json @@ -0,0 +1 @@ +{"CPU": "ALTRA", "ITERATIONS": 59576, "MATRIX_FILE": "fl2010", "MATRIX_SHAPE": [484481, 484481], "MATRIX_SIZE": 234721839361, "MATRIX_NNZ": 2346294, "MATRIX_DENSITY": 9.99606174861054e-06, "TIME_S": 20.8360857963562, "TIME_S_1KI": 0.3497395897065295, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1789.3761782836914, "W": 99.90947352525058, "J_1KI": 30.035184945006232, "W_1KI": 1.6770087539487475, "W_D": 80.62447352525058, "J_D": 1443.982309408188, "W_D_1KI": 1.3533045777704205, "J_D_1KI": 0.022715599868578296} diff --git a/pytorch/output_cpu/altra_10_10_fl2010_10000.output b/pytorch/output_test/altra_20_10_10_fl2010.output similarity index 81% rename from pytorch/output_cpu/altra_10_10_fl2010_10000.output rename to pytorch/output_test/altra_20_10_10_fl2010.output index 7b4ecc2..0970e39 100644 --- a/pytorch/output_cpu/altra_10_10_fl2010_10000.output +++ b/pytorch/output_test/altra_20_10_10_fl2010.output @@ -5,21 +5,21 @@ srun: # All submission nodes and all other compute nodes have x86_64 architectur srun: # CPUs. Programs, environments, or other software that was built on x86_64 # srun: # nodes may need to be rebuilt to properly execute on these nodes. # srun: ################################################################################ -srun: job 3471002 queued and waiting for resources -srun: job 3471002 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) +srun: job 3476532 queued and waiting for resources +srun: job 3476532 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 2, 5, ..., 2346288, 2346292, 2346294]), col_indices=tensor([ 1513, 5311, 947, ..., 484460, 482463, 484022]), values=tensor([28364., 12497., 11567., ..., 8532., 22622., 35914.]), size=(484481, 484481), nnz=2346294, layout=torch.sparse_csr) -tensor([0.5561, 0.7849, 0.5628, ..., 0.5545, 0.2543, 0.1741]) +tensor([0.2840, 0.7196, 0.5870, ..., 0.4602, 0.2499, 0.1696]) Matrix: fl2010 Shape: torch.Size([484481, 484481]) Size: 234721839361 NNZ: 2346294 Density: 9.99606174861054e-06 -Time: 2.924255609512329 seconds +Time: 20.8360857963562 seconds diff --git a/pytorch/output_test/altra_20_10_10_ga2010.json b/pytorch/output_test/altra_20_10_10_ga2010.json new file mode 100644 index 0000000..898b8c6 --- /dev/null +++ b/pytorch/output_test/altra_20_10_10_ga2010.json @@ -0,0 +1 @@ +{"CPU": "ALTRA", "ITERATIONS": 104029, "MATRIX_FILE": "ga2010", "MATRIX_SHAPE": [291086, 291086], "MATRIX_SIZE": 84731059396, "MATRIX_NNZ": 1418056, "MATRIX_DENSITY": 1.6735964475229304e-05, "TIME_S": 21.23989486694336, "TIME_S_1KI": 0.2041728255288752, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2888.099227333069, "W": 101.0000104688918, "J_1KI": 27.76244342763142, "W_1KI": 0.9708832197646022, "W_D": 81.67701046889181, "J_D": 2335.5572908453946, "W_D_1KI": 0.7851369374779321, "J_D_1KI": 0.0075472890970588215} diff --git a/pytorch/output_test2/altra_10_10_ga2010_100000.output b/pytorch/output_test/altra_20_10_10_ga2010.output similarity index 81% rename from pytorch/output_test2/altra_10_10_ga2010_100000.output rename to pytorch/output_test/altra_20_10_10_ga2010.output index 8557254..e759e47 100644 --- a/pytorch/output_test2/altra_10_10_ga2010_100000.output +++ b/pytorch/output_test/altra_20_10_10_ga2010.output @@ -5,21 +5,21 @@ srun: # All submission nodes and all other compute nodes have x86_64 architectur srun: # CPUs. Programs, environments, or other software that was built on x86_64 # srun: # nodes may need to be rebuilt to properly execute on these nodes. # srun: ################################################################################ -srun: job 3471788 queued and waiting for resources -srun: job 3471788 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) +srun: job 3476491 queued and waiting for resources +srun: job 3476491 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 3, 10, ..., 1418047, 1418054, 1418056]), col_indices=tensor([ 1566, 1871, 1997, ..., 291064, 289820, 290176]), values=tensor([18760., 17851., 18847., ..., 65219., 56729., 77629.]), size=(291086, 291086), nnz=1418056, layout=torch.sparse_csr) -tensor([0.6229, 0.6308, 0.8573, ..., 0.9191, 0.9418, 0.6011]) +tensor([0.3095, 0.1993, 0.5353, ..., 0.0168, 0.6101, 0.3556]) Matrix: ga2010 Shape: torch.Size([291086, 291086]) Size: 84731059396 NNZ: 1418056 Density: 1.6735964475229304e-05 -Time: 17.813313722610474 seconds +Time: 21.23989486694336 seconds diff --git a/pytorch/output_test/altra_20_10_10_mac_econ_fwd500.json b/pytorch/output_test/altra_20_10_10_mac_econ_fwd500.json new file mode 100644 index 0000000..245a924 --- /dev/null +++ b/pytorch/output_test/altra_20_10_10_mac_econ_fwd500.json @@ -0,0 +1 @@ +{"CPU": "ALTRA", "ITERATIONS": 172723, "MATRIX_FILE": "mac_econ_fwd500", "MATRIX_SHAPE": [206500, 206500], "MATRIX_SIZE": 42642250000, "MATRIX_NNZ": 1273389, "MATRIX_DENSITY": 2.9862143765866013e-05, "TIME_S": 22.76954960823059, "TIME_S_1KI": 0.13182696924110043, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2236.046720199585, "W": 97.7930267311125, "J_1KI": 12.94585388280417, "W_1KI": 0.56618416036725, "W_D": 78.1710267311125, "J_D": 1787.3878514604567, "W_D_1KI": 0.4525802975348535, "J_D_1KI": 0.0026202665396898705} diff --git a/pytorch/output_cpu/altra_10_10_mac_econ_fwd500_10000.output b/pytorch/output_test/altra_20_10_10_mac_econ_fwd500.output similarity index 82% rename from pytorch/output_cpu/altra_10_10_mac_econ_fwd500_10000.output rename to pytorch/output_test/altra_20_10_10_mac_econ_fwd500.output index aad4fb3..a205229 100644 --- a/pytorch/output_cpu/altra_10_10_mac_econ_fwd500_10000.output +++ b/pytorch/output_test/altra_20_10_10_mac_econ_fwd500.output @@ -5,10 +5,10 @@ srun: # All submission nodes and all other compute nodes have x86_64 architectur srun: # CPUs. Programs, environments, or other software that was built on x86_64 # srun: # nodes may need to be rebuilt to properly execute on these nodes. # srun: ################################################################################ -srun: job 3471003 queued and waiting for resources -srun: job 3471003 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) +srun: job 3476533 queued and waiting for resources +srun: job 3476533 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 3, 8, ..., 1273376, 1273379, 1273389]), col_indices=tensor([ 3, 30, 44, ..., 206363, 206408, @@ -16,11 +16,11 @@ tensor(crow_indices=tensor([ 0, 3, 8, ..., 1273376, values=tensor([-3.7877e-03, -1.5420e-01, 9.5305e-04, ..., 1.2290e-01, 2.2235e-01, -1.0000e+00]), size=(206500, 206500), nnz=1273389, layout=torch.sparse_csr) -tensor([0.8982, 0.5128, 0.1053, ..., 0.5733, 0.7437, 0.9673]) +tensor([0.1042, 0.1637, 0.7950, ..., 0.9705, 0.0203, 0.3514]) Matrix: mac_econ_fwd500 Shape: torch.Size([206500, 206500]) Size: 42642250000 NNZ: 1273389 Density: 2.9862143765866013e-05 -Time: 1.6093401908874512 seconds +Time: 22.76954960823059 seconds diff --git a/pytorch/output_test/altra_20_10_10_mc2depi.json b/pytorch/output_test/altra_20_10_10_mc2depi.json new file mode 100644 index 0000000..0e4ccbd --- /dev/null +++ b/pytorch/output_test/altra_20_10_10_mc2depi.json @@ -0,0 +1 @@ +{"CPU": "ALTRA", "ITERATIONS": 103897, "MATRIX_FILE": "mc2depi", "MATRIX_SHAPE": [525825, 525825], "MATRIX_SIZE": 276491930625, "MATRIX_NNZ": 2100225, "MATRIX_DENSITY": 7.595972132902821e-06, "TIME_S": 20.842941761016846, "TIME_S_1KI": 0.20061158417487363, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1868.1001363945004, "W": 104.62435208458886, "J_1KI": 17.98030873263425, "W_1KI": 1.0070007034331008, "W_D": 85.70235208458887, "J_D": 1530.2419793157574, "W_D_1KI": 0.8248780242412088, "J_D_1KI": 0.007939382506147518} diff --git a/pytorch/output_test2/altra_10_10_mc2depi_100000.output b/pytorch/output_test/altra_20_10_10_mc2depi.output similarity index 81% rename from pytorch/output_test2/altra_10_10_mc2depi_100000.output rename to pytorch/output_test/altra_20_10_10_mc2depi.output index 486689e..fb772a2 100644 --- a/pytorch/output_test2/altra_10_10_mc2depi_100000.output +++ b/pytorch/output_test/altra_20_10_10_mc2depi.output @@ -5,21 +5,21 @@ srun: # All submission nodes and all other compute nodes have x86_64 architectur srun: # CPUs. Programs, environments, or other software that was built on x86_64 # srun: # nodes may need to be rebuilt to properly execute on these nodes. # srun: ################################################################################ -srun: job 3471790 queued and waiting for resources -srun: job 3471790 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) +srun: job 3476480 queued and waiting for resources +srun: job 3476480 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 2, 5, ..., 2100220, 2100223, 2100225]), col_indices=tensor([ 0, 1, 1, ..., 525824, 525821, 525824]), values=tensor([-2025., 2025., -2026., ..., 2025., 1024., -1024.]), size=(525825, 525825), nnz=2100225, layout=torch.sparse_csr) -tensor([0.4809, 0.5361, 0.4713, ..., 0.3506, 0.4153, 0.4817]) +tensor([0.2212, 0.6291, 0.2605, ..., 0.9527, 0.3124, 0.8273]) Matrix: mc2depi Shape: torch.Size([525825, 525825]) Size: 276491930625 NNZ: 2100225 Density: 7.595972132902821e-06 -Time: 19.404656887054443 seconds +Time: 20.842941761016846 seconds diff --git a/pytorch/output_test/altra_20_10_10_p2p-Gnutella04.json b/pytorch/output_test/altra_20_10_10_p2p-Gnutella04.json new file mode 100644 index 0000000..b1b87b9 --- /dev/null +++ b/pytorch/output_test/altra_20_10_10_p2p-Gnutella04.json @@ -0,0 +1 @@ +{"CPU": "ALTRA", "ITERATIONS": 311083, "MATRIX_FILE": "p2p-Gnutella04", "MATRIX_SHAPE": [10879, 10879], "MATRIX_SIZE": 118352641, "MATRIX_NNZ": 39994, "MATRIX_DENSITY": 0.0003379223282393842, "TIME_S": 24.138607263565063, "TIME_S_1KI": 0.0775953917879314, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2205.7783544921876, "W": 87.34642113118167, "J_1KI": 7.0906425439261795, "W_1KI": 0.2807817242703126, "W_D": 67.74142113118168, "J_D": 1710.6889841461186, "W_D_1KI": 0.21775995837503714, "J_D_1KI": 0.0007000059738881171} diff --git a/pytorch/output_cpu/altra_10_10_p2p-Gnutella04_10000.output b/pytorch/output_test/altra_20_10_10_p2p-Gnutella04.output similarity index 80% rename from pytorch/output_cpu/altra_10_10_p2p-Gnutella04_10000.output rename to pytorch/output_test/altra_20_10_10_p2p-Gnutella04.output index e250b7b..c0ef61c 100644 --- a/pytorch/output_cpu/altra_10_10_p2p-Gnutella04_10000.output +++ b/pytorch/output_test/altra_20_10_10_p2p-Gnutella04.output @@ -5,19 +5,19 @@ srun: # All submission nodes and all other compute nodes have x86_64 architectur srun: # CPUs. Programs, environments, or other software that was built on x86_64 # srun: # nodes may need to be rebuilt to properly execute on these nodes. # srun: ################################################################################ -srun: job 3471000 queued and waiting for resources -srun: job 3471000 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) +srun: job 3476535 queued and waiting for resources +srun: job 3476535 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 10, 20, ..., 39994, 39994, 39994]), col_indices=tensor([ 1, 2, 3, ..., 9711, 10875, 10876]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(10879, 10879), nnz=39994, layout=torch.sparse_csr) -tensor([0.2688, 0.1431, 0.7891, ..., 0.0735, 0.7672, 0.4174]) +tensor([0.7445, 0.8973, 0.7626, ..., 0.6605, 0.5600, 0.7373]) Matrix: p2p-Gnutella04 Shape: torch.Size([10879, 10879]) Size: 118352641 NNZ: 39994 Density: 0.0003379223282393842 -Time: 0.9692902565002441 seconds +Time: 24.138607263565063 seconds diff --git a/pytorch/output_test/altra_20_10_10_p2p-Gnutella24.json b/pytorch/output_test/altra_20_10_10_p2p-Gnutella24.json new file mode 100644 index 0000000..1e0975b --- /dev/null +++ b/pytorch/output_test/altra_20_10_10_p2p-Gnutella24.json @@ -0,0 +1 @@ +{"CPU": "ALTRA", "ITERATIONS": 260700, "MATRIX_FILE": "p2p-Gnutella24", "MATRIX_SHAPE": [26518, 26518], "MATRIX_SIZE": 703204324, "MATRIX_NNZ": 65369, "MATRIX_DENSITY": 9.295875717624285e-05, "TIME_S": 23.59526252746582, "TIME_S_1KI": 0.09050733612376609, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2132.9754725265498, "W": 89.43798037251499, "J_1KI": 8.181724098682585, "W_1KI": 0.3430685860088799, "W_D": 69.91098037251498, "J_D": 1667.2828005928986, "W_D_1KI": 0.2681663995877061, "J_D_1KI": 0.0010286398142988343} diff --git a/pytorch/output_test2/altra_10_10_p2p-Gnutella24_100000.output b/pytorch/output_test/altra_20_10_10_p2p-Gnutella24.output similarity index 80% rename from pytorch/output_test2/altra_10_10_p2p-Gnutella24_100000.output rename to pytorch/output_test/altra_20_10_10_p2p-Gnutella24.output index 6d6442d..9d848fc 100644 --- a/pytorch/output_test2/altra_10_10_p2p-Gnutella24_100000.output +++ b/pytorch/output_test/altra_20_10_10_p2p-Gnutella24.output @@ -5,19 +5,19 @@ srun: # All submission nodes and all other compute nodes have x86_64 architectur srun: # CPUs. Programs, environments, or other software that was built on x86_64 # srun: # nodes may need to be rebuilt to properly execute on these nodes. # srun: ################################################################################ -srun: job 3471854 queued and waiting for resources -srun: job 3471854 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) +srun: job 3476561 queued and waiting for resources +srun: job 3476561 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 9, 9, ..., 65369, 65369, 65369]), col_indices=tensor([ 1, 2, 3, ..., 15065, 9401, 26517]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(26518, 26518), nnz=65369, layout=torch.sparse_csr) -tensor([0.4761, 0.4887, 0.0195, ..., 0.5651, 0.2234, 0.2511]) +tensor([0.5210, 0.8215, 0.1854, ..., 0.0439, 0.4101, 0.7995]) Matrix: p2p-Gnutella24 Shape: torch.Size([26518, 26518]) Size: 703204324 NNZ: 65369 Density: 9.295875717624285e-05 -Time: 8.68448281288147 seconds +Time: 23.59526252746582 seconds diff --git a/pytorch/output_test/altra_20_10_10_p2p-Gnutella25.json b/pytorch/output_test/altra_20_10_10_p2p-Gnutella25.json new file mode 100644 index 0000000..7147733 --- /dev/null +++ b/pytorch/output_test/altra_20_10_10_p2p-Gnutella25.json @@ -0,0 +1 @@ +{"CPU": "ALTRA", "ITERATIONS": 264070, "MATRIX_FILE": "p2p-Gnutella25", "MATRIX_SHAPE": [22687, 22687], "MATRIX_SIZE": 514699969, "MATRIX_NNZ": 54705, "MATRIX_DENSITY": 0.00010628522108964806, "TIME_S": 20.885182857513428, "TIME_S_1KI": 0.07908957040751857, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1880.1110801506043, "W": 84.75840980808144, "J_1KI": 7.119745068166033, "W_1KI": 0.3209694770632084, "W_D": 65.18440980808144, "J_D": 1445.920604347706, "W_D_1KI": 0.24684519183580655, "J_D_1KI": 0.0009347718098830104} diff --git a/pytorch/output_test2/altra_10_10_p2p-Gnutella25_100000.output b/pytorch/output_test/altra_20_10_10_p2p-Gnutella25.output similarity index 80% rename from pytorch/output_test2/altra_10_10_p2p-Gnutella25_100000.output rename to pytorch/output_test/altra_20_10_10_p2p-Gnutella25.output index 860e3b3..ebbbfbf 100644 --- a/pytorch/output_test2/altra_10_10_p2p-Gnutella25_100000.output +++ b/pytorch/output_test/altra_20_10_10_p2p-Gnutella25.output @@ -5,19 +5,19 @@ srun: # All submission nodes and all other compute nodes have x86_64 architectur srun: # CPUs. Programs, environments, or other software that was built on x86_64 # srun: # nodes may need to be rebuilt to properly execute on these nodes. # srun: ################################################################################ -srun: job 3471827 queued and waiting for resources -srun: job 3471827 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) +srun: job 3476534 queued and waiting for resources +srun: job 3476534 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 9, 9, ..., 54704, 54704, 54705]), col_indices=tensor([ 1, 2, 3, ..., 17949, 22685, 144]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(22687, 22687), nnz=54705, layout=torch.sparse_csr) -tensor([0.4250, 0.5008, 0.7599, ..., 0.4696, 0.2842, 0.9247]) +tensor([0.0526, 0.5914, 0.0989, ..., 0.9852, 0.3692, 0.4908]) Matrix: p2p-Gnutella25 Shape: torch.Size([22687, 22687]) Size: 514699969 NNZ: 54705 Density: 0.00010628522108964806 -Time: 8.185347080230713 seconds +Time: 20.885182857513428 seconds diff --git a/pytorch/output_test/altra_20_10_10_p2p-Gnutella30.json b/pytorch/output_test/altra_20_10_10_p2p-Gnutella30.json new file mode 100644 index 0000000..a47c6f2 --- /dev/null +++ b/pytorch/output_test/altra_20_10_10_p2p-Gnutella30.json @@ -0,0 +1 @@ +{"CPU": "ALTRA", "ITERATIONS": 252975, "MATRIX_FILE": "p2p-Gnutella30", "MATRIX_SHAPE": [36682, 36682], "MATRIX_SIZE": 1345569124, "MATRIX_NNZ": 88328, "MATRIX_DENSITY": 6.564359899804003e-05, "TIME_S": 22.134840488433838, "TIME_S_1KI": 0.0874981341572639, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1734.7168189907075, "W": 86.01309722549402, "J_1KI": 6.857265812790621, "W_1KI": 0.3400063137681353, "W_D": 67.17509722549403, "J_D": 1354.7910112910274, "W_D_1KI": 0.26554045745822324, "J_D_1KI": 0.0010496707479324963} diff --git a/pytorch/output_cpu/altra_10_10_p2p-Gnutella30_10000.output b/pytorch/output_test/altra_20_10_10_p2p-Gnutella30.output similarity index 80% rename from pytorch/output_cpu/altra_10_10_p2p-Gnutella30_10000.output rename to pytorch/output_test/altra_20_10_10_p2p-Gnutella30.output index d00ff11..87c129b 100644 --- a/pytorch/output_cpu/altra_10_10_p2p-Gnutella30_10000.output +++ b/pytorch/output_test/altra_20_10_10_p2p-Gnutella30.output @@ -5,19 +5,19 @@ srun: # All submission nodes and all other compute nodes have x86_64 architectur srun: # CPUs. Programs, environments, or other software that was built on x86_64 # srun: # nodes may need to be rebuilt to properly execute on these nodes. # srun: ################################################################################ -srun: job 3471007 queued and waiting for resources -srun: job 3471007 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) +srun: job 3476530 queued and waiting for resources +srun: job 3476530 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 10, 10, ..., 88328, 88328, 88328]), col_indices=tensor([ 1, 2, 3, ..., 36675, 36676, 36677]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(36682, 36682), nnz=88328, layout=torch.sparse_csr) -tensor([0.4265, 0.5292, 0.2746, ..., 0.3064, 0.8544, 0.6969]) +tensor([0.0757, 0.9426, 0.1312, ..., 0.6759, 0.1000, 0.4339]) Matrix: p2p-Gnutella30 Shape: torch.Size([36682, 36682]) Size: 1345569124 NNZ: 88328 Density: 6.564359899804003e-05 -Time: 1.022092580795288 seconds +Time: 22.134840488433838 seconds diff --git a/pytorch/output_test/altra_20_10_10_ri2010.json b/pytorch/output_test/altra_20_10_10_ri2010.json new file mode 100644 index 0000000..4c94fda --- /dev/null +++ b/pytorch/output_test/altra_20_10_10_ri2010.json @@ -0,0 +1 @@ +{"CPU": "ALTRA", "ITERATIONS": 287242, "MATRIX_FILE": "ri2010", "MATRIX_SHAPE": [25181, 25181], "MATRIX_SIZE": 634082761, "MATRIX_NNZ": 125750, "MATRIX_DENSITY": 0.00019831796057928155, "TIME_S": 30.075697422027588, "TIME_S_1KI": 0.10470508289883648, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2279.409811325073, "W": 89.12233152384779, "J_1KI": 7.9355032040059355, "W_1KI": 0.3102691511820966, "W_D": 69.78633152384779, "J_D": 1784.868573925018, "W_D_1KI": 0.24295309016037972, "J_D_1KI": 0.0008458132521023378} diff --git a/pytorch/output_cpu/altra_10_10_ri2010_10000.output b/pytorch/output_test/altra_20_10_10_ri2010.output similarity index 81% rename from pytorch/output_cpu/altra_10_10_ri2010_10000.output rename to pytorch/output_test/altra_20_10_10_ri2010.output index 7d61a84..39ec60b 100644 --- a/pytorch/output_cpu/altra_10_10_ri2010_10000.output +++ b/pytorch/output_test/altra_20_10_10_ri2010.output @@ -5,20 +5,20 @@ srun: # All submission nodes and all other compute nodes have x86_64 architectur srun: # CPUs. Programs, environments, or other software that was built on x86_64 # srun: # nodes may need to be rebuilt to properly execute on these nodes. # srun: ################################################################################ -srun: job 3470987 queued and waiting for resources -srun: job 3470987 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) +srun: job 3476489 queued and waiting for resources +srun: job 3476489 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 3, 8, ..., 125742, 125747, 125750]), col_indices=tensor([ 25, 56, 662, ..., 21738, 22279, 23882]), values=tensor([17171., 37318., 5284., ..., 25993., 24918., 803.]), size=(25181, 25181), nnz=125750, layout=torch.sparse_csr) -tensor([0.8235, 0.3045, 0.3176, ..., 0.8277, 0.2909, 0.5754]) +tensor([0.3523, 0.7020, 0.0637, ..., 0.0891, 0.3374, 0.6105]) Matrix: ri2010 Shape: torch.Size([25181, 25181]) Size: 634082761 NNZ: 125750 Density: 0.00019831796057928155 -Time: 0.7675364017486572 seconds +Time: 30.075697422027588 seconds diff --git a/pytorch/output_test/altra_20_10_10_rma10.json b/pytorch/output_test/altra_20_10_10_rma10.json new file mode 100644 index 0000000..aab0c9a --- /dev/null +++ b/pytorch/output_test/altra_20_10_10_rma10.json @@ -0,0 +1 @@ +{"CPU": "ALTRA", "ITERATIONS": 137395, "MATRIX_FILE": "rma10", "MATRIX_SHAPE": [46835, 46835], "MATRIX_SIZE": 2193517225, "MATRIX_NNZ": 2374001, "MATRIX_DENSITY": 0.0010822805369125833, "TIME_S": 21.550864696502686, "TIME_S_1KI": 0.1568533403435546, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2575.452181930543, "W": 94.43274399658898, "J_1KI": 18.744875591764934, "W_1KI": 0.68730844642519, "W_D": 74.66774399658898, "J_D": 2036.4038579964645, "W_D_1KI": 0.5434531387356817, "J_D_1KI": 0.003955406956116902} diff --git a/pytorch/output_cpu/altra_10_10_rma10_10000.output b/pytorch/output_test/altra_20_10_10_rma10.output similarity index 81% rename from pytorch/output_cpu/altra_10_10_rma10_10000.output rename to pytorch/output_test/altra_20_10_10_rma10.output index 83cecf3..a822e0d 100644 --- a/pytorch/output_cpu/altra_10_10_rma10_10000.output +++ b/pytorch/output_test/altra_20_10_10_rma10.output @@ -5,21 +5,21 @@ srun: # All submission nodes and all other compute nodes have x86_64 architectur srun: # CPUs. Programs, environments, or other software that was built on x86_64 # srun: # nodes may need to be rebuilt to properly execute on these nodes. # srun: ################################################################################ -srun: job 3471006 queued and waiting for resources -srun: job 3471006 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) +srun: job 3476538 queued and waiting for resources +srun: job 3476538 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 17, 34, ..., 2373939, 2373970, 2374001]), col_indices=tensor([ 0, 1, 2, ..., 46831, 46833, 46834]), values=tensor([ 1.2636e+05, -1.6615e+07, -8.2015e+04, ..., 8.3378e+01, 2.5138e+00, 1.2184e+03]), size=(46835, 46835), nnz=2374001, layout=torch.sparse_csr) -tensor([0.3759, 0.1778, 0.4707, ..., 0.4812, 0.6721, 0.5216]) +tensor([0.5629, 0.1373, 0.0510, ..., 0.8391, 0.7383, 0.0356]) Matrix: rma10 Shape: torch.Size([46835, 46835]) Size: 2193517225 NNZ: 2374001 Density: 0.0010822805369125833 -Time: 2.688584089279175 seconds +Time: 21.550864696502686 seconds diff --git a/pytorch/output_test/altra_20_10_10_soc-sign-Slashdot090216.json b/pytorch/output_test/altra_20_10_10_soc-sign-Slashdot090216.json new file mode 100644 index 0000000..b941fac --- /dev/null +++ b/pytorch/output_test/altra_20_10_10_soc-sign-Slashdot090216.json @@ -0,0 +1 @@ +{"CPU": "ALTRA", "ITERATIONS": 157730, "MATRIX_FILE": "soc-sign-Slashdot090216", "MATRIX_SHAPE": [81871, 81871], "MATRIX_SIZE": 6702860641, "MATRIX_NNZ": 545671, "MATRIX_DENSITY": 8.140867447881048e-05, "TIME_S": 23.577486515045166, "TIME_S_1KI": 0.14948003876906846, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1623.1555816650387, "W": 82.75784946345307, "J_1KI": 10.290722003835914, "W_1KI": 0.524680463218494, "W_D": 63.524849463453066, "J_D": 1245.9327380971904, "W_D_1KI": 0.40274424309549905, "J_D_1KI": 0.002553377563529443} diff --git a/pytorch/output_cpu/altra_10_10_soc-sign-Slashdot090216_10000.output b/pytorch/output_test/altra_20_10_10_soc-sign-Slashdot090216.output similarity index 81% rename from pytorch/output_cpu/altra_10_10_soc-sign-Slashdot090216_10000.output rename to pytorch/output_test/altra_20_10_10_soc-sign-Slashdot090216.output index c4b74ca..e74f3f1 100644 --- a/pytorch/output_cpu/altra_10_10_soc-sign-Slashdot090216_10000.output +++ b/pytorch/output_test/altra_20_10_10_soc-sign-Slashdot090216.output @@ -5,20 +5,20 @@ srun: # All submission nodes and all other compute nodes have x86_64 architectur srun: # CPUs. Programs, environments, or other software that was built on x86_64 # srun: # nodes may need to be rebuilt to properly execute on these nodes. # srun: ################################################################################ -srun: job 3471015 queued and waiting for resources -srun: job 3471015 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) +srun: job 3476563 queued and waiting for resources +srun: job 3476563 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 29, 124, ..., 545669, 545669, 545671]), col_indices=tensor([ 1, 2, 3, ..., 81869, 81699, 81863]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(81871, 81871), nnz=545671, layout=torch.sparse_csr) -tensor([0.7292, 0.5775, 0.7105, ..., 0.2374, 0.7415, 0.8438]) +tensor([0.7400, 0.6636, 0.1172, ..., 0.9381, 0.6960, 0.9014]) Matrix: soc-sign-Slashdot090216 Shape: torch.Size([81871, 81871]) Size: 6702860641 NNZ: 545671 Density: 8.140867447881048e-05 -Time: 1.4809374809265137 seconds +Time: 23.577486515045166 seconds diff --git a/pytorch/output_test/altra_20_10_10_soc-sign-Slashdot090221.json b/pytorch/output_test/altra_20_10_10_soc-sign-Slashdot090221.json new file mode 100644 index 0000000..a3e3a24 --- /dev/null +++ b/pytorch/output_test/altra_20_10_10_soc-sign-Slashdot090221.json @@ -0,0 +1 @@ +{"CPU": "ALTRA", "ITERATIONS": 162425, "MATRIX_FILE": "soc-sign-Slashdot090221", "MATRIX_SHAPE": [82144, 82144], "MATRIX_SIZE": 6747636736, "MATRIX_NNZ": 549202, "MATRIX_DENSITY": 8.13917555860553e-05, "TIME_S": 20.395315170288086, "TIME_S_1KI": 0.12556758608765944, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1580.4874530792238, "W": 87.7410552563459, "J_1KI": 9.730567665563944, "W_1KI": 0.5401942758586789, "W_D": 68.2940552563459, "J_D": 1230.1869077959063, "W_D_1KI": 0.4204651701175675, "J_D_1KI": 0.0025886727419890255} diff --git a/pytorch/output_cpu/altra_10_10_soc-sign-Slashdot090221_10000.output b/pytorch/output_test/altra_20_10_10_soc-sign-Slashdot090221.output similarity index 81% rename from pytorch/output_cpu/altra_10_10_soc-sign-Slashdot090221_10000.output rename to pytorch/output_test/altra_20_10_10_soc-sign-Slashdot090221.output index 5745a5e..e1a1dc7 100644 --- a/pytorch/output_cpu/altra_10_10_soc-sign-Slashdot090221_10000.output +++ b/pytorch/output_test/altra_20_10_10_soc-sign-Slashdot090221.output @@ -5,20 +5,20 @@ srun: # All submission nodes and all other compute nodes have x86_64 architectur srun: # CPUs. Programs, environments, or other software that was built on x86_64 # srun: # nodes may need to be rebuilt to properly execute on these nodes. # srun: ################################################################################ -srun: job 3470983 queued and waiting for resources -srun: job 3470983 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) +srun: job 3476485 queued and waiting for resources +srun: job 3476485 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 29, 124, ..., 549200, 549200, 549202]), col_indices=tensor([ 1, 2, 3, ..., 82142, 81974, 82136]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(82144, 82144), nnz=549202, layout=torch.sparse_csr) -tensor([0.2718, 0.1909, 0.9904, ..., 0.8130, 0.5743, 0.4283]) +tensor([0.3021, 0.6870, 0.1347, ..., 0.7910, 0.1946, 0.3705]) Matrix: soc-sign-Slashdot090221 Shape: torch.Size([82144, 82144]) Size: 6747636736 NNZ: 549202 Density: 8.13917555860553e-05 -Time: 1.608903408050537 seconds +Time: 20.395315170288086 seconds diff --git a/pytorch/output_test/altra_20_10_10_soc-sign-epinions.json b/pytorch/output_test/altra_20_10_10_soc-sign-epinions.json new file mode 100644 index 0000000..ab596d8 --- /dev/null +++ b/pytorch/output_test/altra_20_10_10_soc-sign-epinions.json @@ -0,0 +1 @@ +{"CPU": "ALTRA", "ITERATIONS": 94360, "MATRIX_FILE": "soc-sign-epinions", "MATRIX_SHAPE": [131828, 131828], "MATRIX_SIZE": 17378621584, "MATRIX_NNZ": 841372, "MATRIX_DENSITY": 4.841419648464106e-05, "TIME_S": 33.08808898925781, "TIME_S_1KI": 0.3506580011578827, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2145.436586656571, "W": 86.98470867178541, "J_1KI": 22.736716687755095, "W_1KI": 0.9218387947412612, "W_D": 71.23970867178541, "J_D": 1757.0936287653449, "W_D_1KI": 0.7549778367081963, "J_D_1KI": 0.008001036845148328} diff --git a/pytorch/output_cpu/altra_10_10_soc-sign-epinions_10000.output b/pytorch/output_test/altra_20_10_10_soc-sign-epinions.output similarity index 81% rename from pytorch/output_cpu/altra_10_10_soc-sign-epinions_10000.output rename to pytorch/output_test/altra_20_10_10_soc-sign-epinions.output index 35af03c..31d4efb 100644 --- a/pytorch/output_cpu/altra_10_10_soc-sign-epinions_10000.output +++ b/pytorch/output_test/altra_20_10_10_soc-sign-epinions.output @@ -5,21 +5,21 @@ srun: # All submission nodes and all other compute nodes have x86_64 architectur srun: # CPUs. Programs, environments, or other software that was built on x86_64 # srun: # nodes may need to be rebuilt to properly execute on these nodes. # srun: ################################################################################ -srun: job 3470998 queued and waiting for resources -srun: job 3470998 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) +srun: job 3476529 queued and waiting for resources +srun: job 3476529 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 1, 2, ..., 841371, 841371, 841372]), col_indices=tensor([ 1, 128552, 3, ..., 131824, 131826, 7714]), values=tensor([-1., -1., 1., ..., 1., 1., 1.]), size=(131828, 131828), nnz=841372, layout=torch.sparse_csr) -tensor([0.6727, 0.2484, 0.1189, ..., 0.2578, 0.7441, 0.8799]) +tensor([0.5788, 0.3846, 0.0341, ..., 0.5887, 0.9861, 0.3240]) Matrix: soc-sign-epinions Shape: torch.Size([131828, 131828]) Size: 17378621584 NNZ: 841372 Density: 4.841419648464106e-05 -Time: 4.555854320526123 seconds +Time: 33.08808898925781 seconds diff --git a/pytorch/output_test/altra_20_10_10_sx-mathoverflow.json b/pytorch/output_test/altra_20_10_10_sx-mathoverflow.json new file mode 100644 index 0000000..58ae56c --- /dev/null +++ b/pytorch/output_test/altra_20_10_10_sx-mathoverflow.json @@ -0,0 +1 @@ +{"CPU": "ALTRA", "ITERATIONS": 244076, "MATRIX_FILE": "sx-mathoverflow", "MATRIX_SHAPE": [24818, 24818], "MATRIX_SIZE": 615933124, "MATRIX_NNZ": 239978, "MATRIX_DENSITY": 0.00038961697406616504, "TIME_S": 23.018490076065063, "TIME_S_1KI": 0.09430869924148652, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2292.358201313019, "W": 90.97612167842344, "J_1KI": 9.391985288651972, "W_1KI": 0.37273685933243517, "W_D": 72.03912167842344, "J_D": 1815.1957716844083, "W_D_1KI": 0.2951503698783307, "J_D_1KI": 0.0012092560099244937} diff --git a/pytorch/output_test2/altra_10_10_sx-mathoverflow_100000.output b/pytorch/output_test/altra_20_10_10_sx-mathoverflow.output similarity index 81% rename from pytorch/output_test2/altra_10_10_sx-mathoverflow_100000.output rename to pytorch/output_test/altra_20_10_10_sx-mathoverflow.output index deb95fa..96370b6 100644 --- a/pytorch/output_test2/altra_10_10_sx-mathoverflow_100000.output +++ b/pytorch/output_test/altra_20_10_10_sx-mathoverflow.output @@ -5,20 +5,20 @@ srun: # All submission nodes and all other compute nodes have x86_64 architectur srun: # CPUs. Programs, environments, or other software that was built on x86_64 # srun: # nodes may need to be rebuilt to properly execute on these nodes. # srun: ################################################################################ -srun: job 3471784 queued and waiting for resources -srun: job 3471784 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) +srun: job 3476481 queued and waiting for resources +srun: job 3476481 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 317, 416, ..., 239976, 239977, 239978]), col_indices=tensor([ 0, 1, 2, ..., 1483, 2179, 24817]), values=tensor([151., 17., 6., ..., 1., 1., 1.]), size=(24818, 24818), nnz=239978, layout=torch.sparse_csr) -tensor([0.0721, 0.7772, 0.5440, ..., 0.2599, 0.9247, 0.3684]) +tensor([0.2494, 0.7432, 0.1406, ..., 0.4775, 0.7260, 0.0607]) Matrix: sx-mathoverflow Shape: torch.Size([24818, 24818]) Size: 615933124 NNZ: 239978 Density: 0.00038961697406616504 -Time: 9.512531042098999 seconds +Time: 23.018490076065063 seconds diff --git a/pytorch/output_test/altra_20_10_10_tn2010.json b/pytorch/output_test/altra_20_10_10_tn2010.json new file mode 100644 index 0000000..474f6ef --- /dev/null +++ b/pytorch/output_test/altra_20_10_10_tn2010.json @@ -0,0 +1 @@ +{"CPU": "ALTRA", "ITERATIONS": 117743, "MATRIX_FILE": "tn2010", "MATRIX_SHAPE": [240116, 240116], "MATRIX_SIZE": 57655693456, "MATRIX_NNZ": 1193966, "MATRIX_DENSITY": 2.070855328296721e-05, "TIME_S": 22.556398630142212, "TIME_S_1KI": 0.19157316044386682, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1764.753028202057, "W": 96.00486890378265, "J_1KI": 14.98817788065581, "W_1KI": 0.8153764461902843, "W_D": 76.65086890378265, "J_D": 1408.989508103371, "W_D_1KI": 0.651001493963825, "J_D_1KI": 0.0055290037960967955} diff --git a/pytorch/output_test2/altra_10_10_tn2010_100000.output b/pytorch/output_test/altra_20_10_10_tn2010.output similarity index 81% rename from pytorch/output_test2/altra_10_10_tn2010_100000.output rename to pytorch/output_test/altra_20_10_10_tn2010.output index a98c8ff..b882f8f 100644 --- a/pytorch/output_test2/altra_10_10_tn2010_100000.output +++ b/pytorch/output_test/altra_20_10_10_tn2010.output @@ -5,10 +5,10 @@ srun: # All submission nodes and all other compute nodes have x86_64 architectur srun: # CPUs. Programs, environments, or other software that was built on x86_64 # srun: # nodes may need to be rebuilt to properly execute on these nodes. # srun: ################################################################################ -srun: job 3471794 queued and waiting for resources -srun: job 3471794 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) +srun: job 3476495 queued and waiting for resources +srun: job 3476495 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 4, 20, ..., 1193961, 1193963, 1193966]), col_indices=tensor([ 1152, 1272, 1961, ..., 238254, 239142, @@ -16,11 +16,11 @@ tensor(crow_indices=tensor([ 0, 4, 20, ..., 1193961, values=tensor([ 5728., 2871., 418449., ..., 10058., 33324., 34928.]), size=(240116, 240116), nnz=1193966, layout=torch.sparse_csr) -tensor([0.7187, 0.4492, 0.0121, ..., 0.1002, 0.2839, 0.4108]) +tensor([0.5589, 0.4305, 0.8268, ..., 0.1583, 0.3546, 0.6697]) Matrix: tn2010 Shape: torch.Size([240116, 240116]) Size: 57655693456 NNZ: 1193966 Density: 2.070855328296721e-05 -Time: 16.210495948791504 seconds +Time: 22.556398630142212 seconds diff --git a/pytorch/output_test/altra_20_10_10_ut2010.json b/pytorch/output_test/altra_20_10_10_ut2010.json new file mode 100644 index 0000000..586b635 --- /dev/null +++ b/pytorch/output_test/altra_20_10_10_ut2010.json @@ -0,0 +1 @@ +{"CPU": "ALTRA", "ITERATIONS": 194381, "MATRIX_FILE": "ut2010", "MATRIX_SHAPE": [115406, 115406], "MATRIX_SIZE": 13318544836, "MATRIX_NNZ": 572066, "MATRIX_DENSITY": 4.295259032005559e-05, "TIME_S": 21.924265384674072, "TIME_S_1KI": 0.11279016665555827, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2239.0512651824956, "W": 97.93349071317404, "J_1KI": 11.518879238107097, "W_1KI": 0.5038223422720021, "W_D": 78.12249071317405, "J_D": 1786.1128036663538, "W_D_1KI": 0.40190394489777315, "J_D_1KI": 0.0020676092051063284} diff --git a/pytorch/output_test2/altra_10_10_ut2010_100000.output b/pytorch/output_test/altra_20_10_10_ut2010.output similarity index 81% rename from pytorch/output_test2/altra_10_10_ut2010_100000.output rename to pytorch/output_test/altra_20_10_10_ut2010.output index cb8b45e..8143448 100644 --- a/pytorch/output_test2/altra_10_10_ut2010_100000.output +++ b/pytorch/output_test/altra_20_10_10_ut2010.output @@ -5,10 +5,10 @@ srun: # All submission nodes and all other compute nodes have x86_64 architectur srun: # CPUs. Programs, environments, or other software that was built on x86_64 # srun: # nodes may need to be rebuilt to properly execute on these nodes. # srun: ################################################################################ -srun: job 3471824 queued and waiting for resources -srun: job 3471824 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) +srun: job 3476537 queued and waiting for resources +srun: job 3476537 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 3, 9, ..., 572056, 572061, 572066]), col_indices=tensor([ 453, 1291, 1979, ..., 113521, 114509, @@ -16,11 +16,11 @@ tensor(crow_indices=tensor([ 0, 3, 9, ..., 572056, 572061, values=tensor([160642., 31335., 282373., ..., 88393., 99485., 18651.]), size=(115406, 115406), nnz=572066, layout=torch.sparse_csr) -tensor([0.7125, 0.8600, 0.2723, ..., 0.9659, 0.9794, 0.8036]) +tensor([0.2802, 0.8038, 0.7377, ..., 0.2628, 0.5642, 0.2541]) Matrix: ut2010 Shape: torch.Size([115406, 115406]) Size: 13318544836 NNZ: 572066 Density: 4.295259032005559e-05 -Time: 14.674797296524048 seconds +Time: 21.924265384674072 seconds diff --git a/pytorch/output_test/altra_20_10_10_va2010.json b/pytorch/output_test/altra_20_10_10_va2010.json new file mode 100644 index 0000000..1cbd4f5 --- /dev/null +++ b/pytorch/output_test/altra_20_10_10_va2010.json @@ -0,0 +1 @@ +{"CPU": "ALTRA", "ITERATIONS": 102911, "MATRIX_FILE": "va2010", "MATRIX_SHAPE": [285762, 285762], "MATRIX_SIZE": 81659920644, "MATRIX_NNZ": 1402128, "MATRIX_DENSITY": 1.717033263003816e-05, "TIME_S": 20.771064281463623, "TIME_S_1KI": 0.20183521957286998, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1694.2755664825436, "W": 96.0772535065061, "J_1KI": 16.463503089879058, "W_1KI": 0.9335955680783016, "W_D": 76.6152535065061, "J_D": 1351.0726763973232, "W_D_1KI": 0.7444807018346541, "J_D_1KI": 0.007234218905993082} diff --git a/pytorch/output_test2/altra_10_10_va2010_100000.output b/pytorch/output_test/altra_20_10_10_va2010.output similarity index 81% rename from pytorch/output_test2/altra_10_10_va2010_100000.output rename to pytorch/output_test/altra_20_10_10_va2010.output index 9c2d785..eef7ba2 100644 --- a/pytorch/output_test2/altra_10_10_va2010_100000.output +++ b/pytorch/output_test/altra_20_10_10_va2010.output @@ -5,10 +5,10 @@ srun: # All submission nodes and all other compute nodes have x86_64 architectur srun: # CPUs. Programs, environments, or other software that was built on x86_64 # srun: # nodes may need to be rebuilt to properly execute on these nodes. # srun: ################################################################################ -srun: job 3471831 queued and waiting for resources -srun: job 3471831 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) +srun: job 3476531 queued and waiting for resources +srun: job 3476531 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 2, 8, ..., 1402119, 1402123, 1402128]), col_indices=tensor([ 2006, 2464, 1166, ..., 285581, 285634, @@ -16,11 +16,11 @@ tensor(crow_indices=tensor([ 0, 2, 8, ..., 1402119, values=tensor([125334., 3558., 1192., ..., 10148., 1763., 9832.]), size=(285762, 285762), nnz=1402128, layout=torch.sparse_csr) -tensor([0.4623, 0.7205, 0.5451, ..., 0.0101, 0.1478, 0.8275]) +tensor([0.9522, 0.7820, 0.1660, ..., 0.6823, 0.9171, 0.7749]) Matrix: va2010 Shape: torch.Size([285762, 285762]) Size: 81659920644 NNZ: 1402128 Density: 1.717033263003816e-05 -Time: 21.11183762550354 seconds +Time: 20.771064281463623 seconds diff --git a/pytorch/output_test/altra_20_10_10_vt2010.json b/pytorch/output_test/altra_20_10_10_vt2010.json new file mode 100644 index 0000000..bff7650 --- /dev/null +++ b/pytorch/output_test/altra_20_10_10_vt2010.json @@ -0,0 +1 @@ +{"CPU": "ALTRA", "ITERATIONS": 284551, "MATRIX_FILE": "vt2010", "MATRIX_SHAPE": [32580, 32580], "MATRIX_SIZE": 1061456400, "MATRIX_NNZ": 155598, "MATRIX_DENSITY": 0.00014658915806621921, "TIME_S": 24.583266496658325, "TIME_S_1KI": 0.08639318258118343, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1941.750954065323, "W": 90.30821048926845, "J_1KI": 6.823911896515293, "W_1KI": 0.3173709123822037, "W_D": 70.52321048926845, "J_D": 1516.3461938774587, "W_D_1KI": 0.24784031856949526, "J_D_1KI": 0.000870987339947831} diff --git a/pytorch/output_test2/altra_10_10_vt2010_100000.output b/pytorch/output_test/altra_20_10_10_vt2010.output similarity index 79% rename from pytorch/output_test2/altra_10_10_vt2010_100000.output rename to pytorch/output_test/altra_20_10_10_vt2010.output index 708f767..07f86c1 100644 --- a/pytorch/output_test2/altra_10_10_vt2010_100000.output +++ b/pytorch/output_test/altra_20_10_10_vt2010.output @@ -5,20 +5,21 @@ srun: # All submission nodes and all other compute nodes have x86_64 architectur srun: # CPUs. Programs, environments, or other software that was built on x86_64 # srun: # nodes may need to be rebuilt to properly execute on these nodes. # srun: ################################################################################ -srun: job 3471830 queued and waiting for resources -srun: job 3471830 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) +srun: job 3476539 queued and waiting for resources +srun: job 3476539 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 4, 7, ..., 155588, 155592, 155598]), col_indices=tensor([ 131, 561, 996, ..., 32237, 32238, 32570]), values=tensor([79040., 7820., 15136., ..., 2828., 17986., 2482.]), size=(32580, 32580), nnz=155598, layout=torch.sparse_csr) -tensor([0.7636, 0.2831, 0.7866, ..., 0.4359, 0.2796, 0.7453]) +tensor([2.3985e-01, 1.8117e-01, 4.3245e-01, ..., 3.0947e-04, 7.7641e-02, + 4.5171e-01]) Matrix: vt2010 Shape: torch.Size([32580, 32580]) Size: 1061456400 NNZ: 155598 Density: 0.00014658915806621921 -Time: 10.058021783828735 seconds +Time: 24.583266496658325 seconds diff --git a/pytorch/output_test/epyc_7313p_20_10_10_ASIC_680k.json b/pytorch/output_test/epyc_7313p_20_10_10_ASIC_680k.json new file mode 100644 index 0000000..9108b9e --- /dev/null +++ b/pytorch/output_test/epyc_7313p_20_10_10_ASIC_680k.json @@ -0,0 +1 @@ +{"CPU": "EPYC_7313P", "ITERATIONS": 29801, "MATRIX_FILE": "ASIC_680k", "MATRIX_SHAPE": [682862, 682862], "MATRIX_SIZE": 466300511044, "MATRIX_NNZ": 3871773, "MATRIX_DENSITY": 8.303171256088674e-06, "TIME_S": 25.471954822540283, "TIME_S_1KI": 0.854734902269732, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 3209.2452868652344, "W": 130.56, "J_1KI": 107.68918113033907, "W_1KI": 4.381061038220194, "W_D": 95.022, "J_D": 2335.6993386068348, "W_D_1KI": 3.188550719774504, "J_D_1KI": 0.10699475587310842} diff --git a/pytorch/output_test2/epyc_7313p_10_10_ASIC_680k_100000.output b/pytorch/output_test/epyc_7313p_20_10_10_ASIC_680k.output similarity index 73% rename from pytorch/output_test2/epyc_7313p_10_10_ASIC_680k_100000.output rename to pytorch/output_test/epyc_7313p_20_10_10_ASIC_680k.output index 99df8d4..fce0709 100644 --- a/pytorch/output_test2/epyc_7313p_10_10_ASIC_680k_100000.output +++ b/pytorch/output_test/epyc_7313p_20_10_10_ASIC_680k.output @@ -1,8 +1,8 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3471836 queued and waiting for resources -srun: job 3471836 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476569 queued and waiting for resources +srun: job 3476569 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 3, 4, ..., 3871767, 3871770, 3871773]), col_indices=tensor([ 0, 11698, 11699, ..., 169456, 645874, @@ -10,11 +10,11 @@ tensor(crow_indices=tensor([ 0, 3, 4, ..., 3871767, values=tensor([ 3.8333e-04, -3.3333e-04, -5.0000e-05, ..., 0.0000e+00, 0.0000e+00, 7.9289e-02]), size=(682862, 682862), nnz=3871773, layout=torch.sparse_csr) -tensor([0.9586, 0.4554, 0.3276, ..., 0.2106, 0.5062, 0.3303]) +tensor([0.1829, 0.4130, 0.6505, ..., 0.6400, 0.4894, 0.9473]) Matrix: ASIC_680k Shape: torch.Size([682862, 682862]) Size: 466300511044 NNZ: 3871773 Density: 8.303171256088674e-06 -Time: 77.6055359840393 seconds +Time: 25.471954822540283 seconds diff --git a/pytorch/output_test/epyc_7313p_20_10_10_Oregon-2.json b/pytorch/output_test/epyc_7313p_20_10_10_Oregon-2.json new file mode 100644 index 0000000..6e7e05a --- /dev/null +++ b/pytorch/output_test/epyc_7313p_20_10_10_Oregon-2.json @@ -0,0 +1 @@ +{"CPU": "EPYC_7313P", "ITERATIONS": 433594, "MATRIX_FILE": "Oregon-2", "MATRIX_SHAPE": [11806, 11806], "MATRIX_SIZE": 139381636, "MATRIX_NNZ": 65460, "MATRIX_DENSITY": 0.0004696458003979807, "TIME_S": 20.345709323883057, "TIME_S_1KI": 0.04692341066500702, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2256.82851102829, "W": 104.56, "J_1KI": 5.204934826192913, "W_1KI": 0.24114724834753248, "W_D": 68.82475000000001, "J_D": 1485.5170052065255, "W_D_1KI": 0.15873086343445714, "J_D_1KI": 0.00036608178027015397} diff --git a/pytorch/output_test2/epyc_7313p_10_10_Oregon-2_100000.output b/pytorch/output_test/epyc_7313p_20_10_10_Oregon-2.output similarity index 69% rename from pytorch/output_test2/epyc_7313p_10_10_Oregon-2_100000.output rename to pytorch/output_test/epyc_7313p_20_10_10_Oregon-2.output index d802d44..2b734ba 100644 --- a/pytorch/output_test2/epyc_7313p_10_10_Oregon-2_100000.output +++ b/pytorch/output_test/epyc_7313p_20_10_10_Oregon-2.output @@ -1,17 +1,17 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3471837 queued and waiting for resources -srun: job 3471837 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476570 queued and waiting for resources +srun: job 3476570 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 583, 584, ..., 65459, 65460, 65460]), col_indices=tensor([ 2, 23, 27, ..., 3324, 958, 841]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(11806, 11806), nnz=65460, layout=torch.sparse_csr) -tensor([0.4829, 0.4475, 0.1256, ..., 0.6137, 0.5875, 0.8973]) +tensor([0.8173, 0.1585, 0.3512, ..., 0.2844, 0.9695, 0.3188]) Matrix: Oregon-2 Shape: torch.Size([11806, 11806]) Size: 139381636 NNZ: 65460 Density: 0.0004696458003979807 -Time: 4.933578252792358 seconds +Time: 20.345709323883057 seconds diff --git a/pytorch/output_test/epyc_7313p_20_10_10_as-caida.json b/pytorch/output_test/epyc_7313p_20_10_10_as-caida.json new file mode 100644 index 0000000..f3598d2 --- /dev/null +++ b/pytorch/output_test/epyc_7313p_20_10_10_as-caida.json @@ -0,0 +1 @@ +{"CPU": "EPYC_7313P", "ITERATIONS": 314835, "MATRIX_FILE": "as-caida", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 106762, "MATRIX_DENSITY": 0.00010842726485909405, "TIME_S": 20.991143941879272, "TIME_S_1KI": 0.06667347639836509, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2225.5722650814055, "W": 108.27999999999999, "J_1KI": 7.069011593632873, "W_1KI": 0.34392618355646604, "W_D": 72.26499999999999, "J_D": 1485.3248959743974, "W_D_1KI": 0.22953292994743274, "J_D_1KI": 0.0007290578555352255} diff --git a/pytorch/output_test2/epyc_7313p_10_10_as-caida_100000.output b/pytorch/output_test/epyc_7313p_20_10_10_as-caida.output similarity index 70% rename from pytorch/output_test2/epyc_7313p_10_10_as-caida_100000.output rename to pytorch/output_test/epyc_7313p_20_10_10_as-caida.output index 81d5172..b4310b4 100644 --- a/pytorch/output_test2/epyc_7313p_10_10_as-caida_100000.output +++ b/pytorch/output_test/epyc_7313p_20_10_10_as-caida.output @@ -1,18 +1,18 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3471778 queued and waiting for resources -srun: job 3471778 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476478 queued and waiting for resources +srun: job 3476478 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 0, 0, ..., 106761, 106761, 106762]), col_indices=tensor([ 106, 329, 1040, ..., 155, 160, 12170]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379), nnz=106762, layout=torch.sparse_csr) -tensor([0.1183, 0.9529, 0.6144, ..., 0.4979, 0.4476, 0.7005]) +tensor([0.4004, 0.5887, 0.1490, ..., 0.4370, 0.7302, 0.0714]) Matrix: as-caida Shape: torch.Size([31379, 31379]) Size: 984641641 NNZ: 106762 Density: 0.00010842726485909405 -Time: 7.285882234573364 seconds +Time: 20.991143941879272 seconds diff --git a/pytorch/output_test/epyc_7313p_20_10_10_dc2.json b/pytorch/output_test/epyc_7313p_20_10_10_dc2.json new file mode 100644 index 0000000..186f713 --- /dev/null +++ b/pytorch/output_test/epyc_7313p_20_10_10_dc2.json @@ -0,0 +1 @@ +{"CPU": "EPYC_7313P", "ITERATIONS": 104532, "MATRIX_FILE": "dc2", "MATRIX_SHAPE": [116835, 116835], "MATRIX_SIZE": 13650417225, "MATRIX_NNZ": 766396, "MATRIX_DENSITY": 5.614451099680581e-05, "TIME_S": 21.358728170394897, "TIME_S_1KI": 0.2043271741705401, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2847.6675070524216, "W": 122.99, "J_1KI": 27.242064698393044, "W_1KI": 1.1765775073661655, "W_D": 87.15674999999999, "J_D": 2017.9969509333368, "W_D_1KI": 0.8337805648031223, "J_D_1KI": 0.007976318876546151} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_dc2_10000.output b/pytorch/output_test/epyc_7313p_20_10_10_dc2.output similarity index 73% rename from pytorch/output_cpu/epyc_7313p_10_10_dc2_10000.output rename to pytorch/output_test/epyc_7313p_20_10_10_dc2.output index f90c1ff..a779d4b 100644 --- a/pytorch/output_cpu/epyc_7313p_10_10_dc2_10000.output +++ b/pytorch/output_test/epyc_7313p_20_10_10_dc2.output @@ -1,8 +1,8 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3470885 queued and waiting for resources -srun: job 3470885 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476482 queued and waiting for resources +srun: job 3476482 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 1, 2, ..., 766390, 766394, 766396]), col_indices=tensor([ 0, 1, 2, ..., 116833, 89, @@ -10,11 +10,11 @@ tensor(crow_indices=tensor([ 0, 1, 2, ..., 766390, 766394, values=tensor([-1.0000e+00, -1.0000e+00, -1.0000e+00, ..., 1.0331e+01, -1.0000e-03, 1.0000e-03]), size=(116835, 116835), nnz=766396, layout=torch.sparse_csr) -tensor([0.9489, 0.1111, 0.7586, ..., 0.1064, 0.9062, 0.5747]) +tensor([0.5345, 0.2310, 0.4381, ..., 0.2090, 0.0292, 0.2443]) Matrix: dc2 Shape: torch.Size([116835, 116835]) Size: 13650417225 NNZ: 766396 Density: 5.614451099680581e-05 -Time: 2.0699713230133057 seconds +Time: 21.358728170394897 seconds diff --git a/pytorch/output_test/epyc_7313p_20_10_10_de2010.json b/pytorch/output_test/epyc_7313p_20_10_10_de2010.json new file mode 100644 index 0000000..88bff54 --- /dev/null +++ b/pytorch/output_test/epyc_7313p_20_10_10_de2010.json @@ -0,0 +1 @@ +{"CPU": "EPYC_7313P", "ITERATIONS": 372946, "MATRIX_FILE": "de2010", "MATRIX_SHAPE": [24115, 24115], "MATRIX_SIZE": 581533225, "MATRIX_NNZ": 116056, "MATRIX_DENSITY": 0.0001995689928120616, "TIME_S": 21.302247047424316, "TIME_S_1KI": 0.057118851113631235, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2632.722003650665, "W": 111.6, "J_1KI": 7.059257918440378, "W_1KI": 0.29923903192419277, "W_D": 75.222, "J_D": 1774.539556976795, "W_D_1KI": 0.20169676038890347, "J_D_1KI": 0.0005408202806543131} diff --git a/pytorch/output_test2/epyc_7313p_10_10_de2010_100000.output b/pytorch/output_test/epyc_7313p_20_10_10_de2010.output similarity index 71% rename from pytorch/output_test2/epyc_7313p_10_10_de2010_100000.output rename to pytorch/output_test/epyc_7313p_20_10_10_de2010.output index e843cc2..ddf66d9 100644 --- a/pytorch/output_test2/epyc_7313p_10_10_de2010_100000.output +++ b/pytorch/output_test/epyc_7313p_20_10_10_de2010.output @@ -1,19 +1,19 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3471785 queued and waiting for resources -srun: job 3471785 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476492 queued and waiting for resources +srun: job 3476492 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 13, 21, ..., 116047, 116051, 116056]), col_indices=tensor([ 250, 251, 757, ..., 23334, 23553, 24050]), values=tensor([ 14900., 33341., 20255., ..., 164227., 52413., 16949.]), size=(24115, 24115), nnz=116056, layout=torch.sparse_csr) -tensor([0.8359, 0.4165, 0.5742, ..., 0.6583, 0.2127, 0.8459]) +tensor([0.2357, 0.3255, 0.0913, ..., 0.1879, 0.8832, 0.4995]) Matrix: de2010 Shape: torch.Size([24115, 24115]) Size: 581533225 NNZ: 116056 Density: 0.0001995689928120616 -Time: 5.716832399368286 seconds +Time: 21.302247047424316 seconds diff --git a/pytorch/output_test/epyc_7313p_20_10_10_email-Enron.json b/pytorch/output_test/epyc_7313p_20_10_10_email-Enron.json new file mode 100644 index 0000000..33b8648 --- /dev/null +++ b/pytorch/output_test/epyc_7313p_20_10_10_email-Enron.json @@ -0,0 +1 @@ +{"CPU": "EPYC_7313P", "ITERATIONS": 174488, "MATRIX_FILE": "email-Enron", "MATRIX_SHAPE": [36692, 36692], "MATRIX_SIZE": 1346302864, "MATRIX_NNZ": 367662, "MATRIX_DENSITY": 0.0002730901120626302, "TIME_S": 23.066561698913574, "TIME_S_1KI": 0.13219569081491894, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2611.392163631916, "W": 116.92999999999999, "J_1KI": 14.966027254779217, "W_1KI": 0.670132043464307, "W_D": 81.1995, "J_D": 1813.4245958336592, "W_D_1KI": 0.4653586493053964, "J_D_1KI": 0.0026669951475482346} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_email-Enron_10000.output b/pytorch/output_test/epyc_7313p_20_10_10_email-Enron.output similarity index 70% rename from pytorch/output_cpu/epyc_7313p_10_10_email-Enron_10000.output rename to pytorch/output_test/epyc_7313p_20_10_10_email-Enron.output index fed87ab..87c8b35 100644 --- a/pytorch/output_cpu/epyc_7313p_10_10_email-Enron_10000.output +++ b/pytorch/output_test/epyc_7313p_20_10_10_email-Enron.output @@ -1,18 +1,18 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3470880 queued and waiting for resources -srun: job 3470880 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476487 queued and waiting for resources +srun: job 3476487 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 1, 71, ..., 367660, 367661, 367662]), col_indices=tensor([ 1, 0, 2, ..., 36690, 36689, 8203]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(36692, 36692), nnz=367662, layout=torch.sparse_csr) -tensor([0.0951, 0.9410, 0.9223, ..., 0.8038, 0.1166, 0.7786]) +tensor([0.3916, 0.9787, 0.6931, ..., 0.1482, 0.3272, 0.6963]) Matrix: email-Enron Shape: torch.Size([36692, 36692]) Size: 1346302864 NNZ: 367662 Density: 0.0002730901120626302 -Time: 1.2558205127716064 seconds +Time: 23.066561698913574 seconds diff --git a/pytorch/output_test/epyc_7313p_20_10_10_fl2010.json b/pytorch/output_test/epyc_7313p_20_10_10_fl2010.json new file mode 100644 index 0000000..8051792 --- /dev/null +++ b/pytorch/output_test/epyc_7313p_20_10_10_fl2010.json @@ -0,0 +1 @@ +{"CPU": "EPYC_7313P", "ITERATIONS": 54234, "MATRIX_FILE": "fl2010", "MATRIX_SHAPE": [484481, 484481], "MATRIX_SIZE": 234721839361, "MATRIX_NNZ": 2346294, "MATRIX_DENSITY": 9.99606174861054e-06, "TIME_S": 20.59074831008911, "TIME_S_1KI": 0.3796649391542042, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 3192.624046869278, "W": 154.09, "J_1KI": 58.867574710869164, "W_1KI": 2.8412066231515283, "W_D": 117.84075000000001, "J_D": 2441.567993712187, "W_D_1KI": 2.1728205553711697, "J_D_1KI": 0.04006380785800733} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_fl2010_10000.output b/pytorch/output_test/epyc_7313p_20_10_10_fl2010.output similarity index 72% rename from pytorch/output_cpu/epyc_7313p_10_10_fl2010_10000.output rename to pytorch/output_test/epyc_7313p_20_10_10_fl2010.output index 739be80..baa1efe 100644 --- a/pytorch/output_cpu/epyc_7313p_10_10_fl2010_10000.output +++ b/pytorch/output_test/epyc_7313p_20_10_10_fl2010.output @@ -1,19 +1,19 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3470889 queued and waiting for resources -srun: job 3470889 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476521 queued and waiting for resources +srun: job 3476521 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 2, 5, ..., 2346288, 2346292, 2346294]), col_indices=tensor([ 1513, 5311, 947, ..., 484460, 482463, 484022]), values=tensor([28364., 12497., 11567., ..., 8532., 22622., 35914.]), size=(484481, 484481), nnz=2346294, layout=torch.sparse_csr) -tensor([0.2616, 0.5904, 0.5539, ..., 0.1315, 0.7299, 0.8588]) +tensor([0.7969, 0.6577, 0.2505, ..., 0.6065, 0.4065, 0.5973]) Matrix: fl2010 Shape: torch.Size([484481, 484481]) Size: 234721839361 NNZ: 2346294 Density: 9.99606174861054e-06 -Time: 3.8482837677001953 seconds +Time: 20.59074831008911 seconds diff --git a/pytorch/output_test/epyc_7313p_20_10_10_ga2010.json b/pytorch/output_test/epyc_7313p_20_10_10_ga2010.json new file mode 100644 index 0000000..4b4df1a --- /dev/null +++ b/pytorch/output_test/epyc_7313p_20_10_10_ga2010.json @@ -0,0 +1 @@ +{"CPU": "EPYC_7313P", "ITERATIONS": 87760, "MATRIX_FILE": "ga2010", "MATRIX_SHAPE": [291086, 291086], "MATRIX_SIZE": 84731059396, "MATRIX_NNZ": 1418056, "MATRIX_DENSITY": 1.6735964475229304e-05, "TIME_S": 20.59249520301819, "TIME_S_1KI": 0.23464556977003406, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 3028.7060955476763, "W": 144.74, "J_1KI": 34.51123627561162, "W_1KI": 1.649270738377393, "W_D": 109.16125000000001, "J_D": 2284.21544336468, "W_D_1KI": 1.2438610984503193, "J_D_1KI": 0.014173440046152227} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_ga2010_10000.output b/pytorch/output_test/epyc_7313p_20_10_10_ga2010.output similarity index 72% rename from pytorch/output_cpu/epyc_7313p_10_10_ga2010_10000.output rename to pytorch/output_test/epyc_7313p_20_10_10_ga2010.output index ceb943b..07e87d6 100644 --- a/pytorch/output_cpu/epyc_7313p_10_10_ga2010_10000.output +++ b/pytorch/output_test/epyc_7313p_20_10_10_ga2010.output @@ -1,19 +1,19 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3470878 queued and waiting for resources -srun: job 3470878 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476477 queued and waiting for resources +srun: job 3476477 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 3, 10, ..., 1418047, 1418054, 1418056]), col_indices=tensor([ 1566, 1871, 1997, ..., 291064, 289820, 290176]), values=tensor([18760., 17851., 18847., ..., 65219., 56729., 77629.]), size=(291086, 291086), nnz=1418056, layout=torch.sparse_csr) -tensor([0.2127, 0.1840, 0.5883, ..., 0.9651, 0.0622, 0.2931]) +tensor([0.5219, 0.0268, 0.4810, ..., 0.6615, 0.9263, 0.8136]) Matrix: ga2010 Shape: torch.Size([291086, 291086]) Size: 84731059396 NNZ: 1418056 Density: 1.6735964475229304e-05 -Time: 2.374833583831787 seconds +Time: 20.59249520301819 seconds diff --git a/pytorch/output_test/epyc_7313p_20_10_10_mac_econ_fwd500.json b/pytorch/output_test/epyc_7313p_20_10_10_mac_econ_fwd500.json new file mode 100644 index 0000000..a2ff469 --- /dev/null +++ b/pytorch/output_test/epyc_7313p_20_10_10_mac_econ_fwd500.json @@ -0,0 +1 @@ +{"CPU": "EPYC_7313P", "ITERATIONS": 187096, "MATRIX_FILE": "mac_econ_fwd500", "MATRIX_SHAPE": [206500, 206500], "MATRIX_SIZE": 42642250000, "MATRIX_NNZ": 1273389, "MATRIX_DENSITY": 2.9862143765866013e-05, "TIME_S": 20.061123847961426, "TIME_S_1KI": 0.10722369183713937, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 3326.9761991357805, "W": 159.93, "J_1KI": 17.782187749261237, "W_1KI": 0.8548018129730193, "W_D": 123.86825, "J_D": 2576.794344892144, "W_D_1KI": 0.6620571792021209, "J_D_1KI": 0.0035385961175125116} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_mac_econ_fwd500_10000.output b/pytorch/output_test/epyc_7313p_20_10_10_mac_econ_fwd500.output similarity index 73% rename from pytorch/output_cpu/epyc_7313p_10_10_mac_econ_fwd500_10000.output rename to pytorch/output_test/epyc_7313p_20_10_10_mac_econ_fwd500.output index 74c0f20..201cc47 100644 --- a/pytorch/output_cpu/epyc_7313p_10_10_mac_econ_fwd500_10000.output +++ b/pytorch/output_test/epyc_7313p_20_10_10_mac_econ_fwd500.output @@ -1,8 +1,8 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3470890 queued and waiting for resources -srun: job 3470890 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476514 queued and waiting for resources +srun: job 3476514 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 3, 8, ..., 1273376, 1273379, 1273389]), col_indices=tensor([ 3, 30, 44, ..., 206363, 206408, @@ -10,11 +10,11 @@ tensor(crow_indices=tensor([ 0, 3, 8, ..., 1273376, values=tensor([-3.7877e-03, -1.5420e-01, 9.5305e-04, ..., 1.2290e-01, 2.2235e-01, -1.0000e+00]), size=(206500, 206500), nnz=1273389, layout=torch.sparse_csr) -tensor([0.2225, 0.9184, 0.8891, ..., 0.8781, 0.9920, 0.8523]) +tensor([0.8302, 0.9512, 0.0935, ..., 0.4727, 0.5988, 0.2894]) Matrix: mac_econ_fwd500 Shape: torch.Size([206500, 206500]) Size: 42642250000 NNZ: 1273389 Density: 2.9862143765866013e-05 -Time: 1.166548252105713 seconds +Time: 20.061123847961426 seconds diff --git a/pytorch/output_test/epyc_7313p_20_10_10_mc2depi.json b/pytorch/output_test/epyc_7313p_20_10_10_mc2depi.json new file mode 100644 index 0000000..1ea2ff9 --- /dev/null +++ b/pytorch/output_test/epyc_7313p_20_10_10_mc2depi.json @@ -0,0 +1 @@ +{"CPU": "EPYC_7313P", "ITERATIONS": 151961, "MATRIX_FILE": "mc2depi", "MATRIX_SHAPE": [525825, 525825], "MATRIX_SIZE": 276491930625, "MATRIX_NNZ": 2100225, "MATRIX_DENSITY": 7.595972132902821e-06, "TIME_S": 24.043957471847534, "TIME_S_1KI": 0.1582245278186346, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 3613.2343026351928, "W": 167.19, "J_1KI": 23.777379081706442, "W_1KI": 1.100216502918512, "W_D": 131.1055, "J_D": 2833.392486776352, "W_D_1KI": 0.8627575496344457, "J_D_1KI": 0.005677493235991114} diff --git a/pytorch/output_test2/epyc_7313p_10_10_mc2depi_100000.output b/pytorch/output_test/epyc_7313p_20_10_10_mc2depi.output similarity index 72% rename from pytorch/output_test2/epyc_7313p_10_10_mc2depi_100000.output rename to pytorch/output_test/epyc_7313p_20_10_10_mc2depi.output index 7c54e85..2bf412e 100644 --- a/pytorch/output_test2/epyc_7313p_10_10_mc2depi_100000.output +++ b/pytorch/output_test/epyc_7313p_20_10_10_mc2depi.output @@ -1,19 +1,19 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3471789 queued and waiting for resources -srun: job 3471789 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476479 queued and waiting for resources +srun: job 3476479 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 2, 5, ..., 2100220, 2100223, 2100225]), col_indices=tensor([ 0, 1, 1, ..., 525824, 525821, 525824]), values=tensor([-2025., 2025., -2026., ..., 2025., 1024., -1024.]), size=(525825, 525825), nnz=2100225, layout=torch.sparse_csr) -tensor([0.0548, 0.4624, 0.2352, ..., 0.4021, 0.8916, 0.8349]) +tensor([0.2605, 0.5203, 0.4377, ..., 0.8693, 0.5968, 0.8774]) Matrix: mc2depi Shape: torch.Size([525825, 525825]) Size: 276491930625 NNZ: 2100225 Density: 7.595972132902821e-06 -Time: 14.108525037765503 seconds +Time: 24.043957471847534 seconds diff --git a/pytorch/output_test/epyc_7313p_20_10_10_p2p-Gnutella04.json b/pytorch/output_test/epyc_7313p_20_10_10_p2p-Gnutella04.json new file mode 100644 index 0000000..a6348f1 --- /dev/null +++ b/pytorch/output_test/epyc_7313p_20_10_10_p2p-Gnutella04.json @@ -0,0 +1 @@ +{"CPU": "EPYC_7313P", "ITERATIONS": 532918, "MATRIX_FILE": "p2p-Gnutella04", "MATRIX_SHAPE": [10879, 10879], "MATRIX_SIZE": 118352641, "MATRIX_NNZ": 39994, "MATRIX_DENSITY": 0.0003379223282393842, "TIME_S": 20.967505931854248, "TIME_S_1KI": 0.03934471331772289, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2097.7220935153964, "W": 103.48000000000002, "J_1KI": 3.936294314538815, "W_1KI": 0.19417621472721885, "W_D": 67.41700000000002, "J_D": 1366.661484137297, "W_D_1KI": 0.12650539107329833, "J_D_1KI": 0.0002373824698608385} diff --git a/pytorch/output_test2/epyc_7313p_10_10_p2p-Gnutella04_100000.output b/pytorch/output_test/epyc_7313p_20_10_10_p2p-Gnutella04.output similarity index 69% rename from pytorch/output_test2/epyc_7313p_10_10_p2p-Gnutella04_100000.output rename to pytorch/output_test/epyc_7313p_20_10_10_p2p-Gnutella04.output index c6aa4a4..35ed1cd 100644 --- a/pytorch/output_test2/epyc_7313p_10_10_p2p-Gnutella04_100000.output +++ b/pytorch/output_test/epyc_7313p_20_10_10_p2p-Gnutella04.output @@ -1,17 +1,17 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3471809 queued and waiting for resources -srun: job 3471809 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476519 queued and waiting for resources +srun: job 3476519 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 10, 20, ..., 39994, 39994, 39994]), col_indices=tensor([ 1, 2, 3, ..., 9711, 10875, 10876]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(10879, 10879), nnz=39994, layout=torch.sparse_csr) -tensor([0.0521, 0.7363, 0.1682, ..., 0.5599, 0.1291, 0.8935]) +tensor([0.9793, 0.7652, 0.9605, ..., 0.7359, 0.2674, 0.3728]) Matrix: p2p-Gnutella04 Shape: torch.Size([10879, 10879]) Size: 118352641 NNZ: 39994 Density: 0.0003379223282393842 -Time: 3.682297468185425 seconds +Time: 20.967505931854248 seconds diff --git a/pytorch/output_test/epyc_7313p_20_10_10_p2p-Gnutella24.json b/pytorch/output_test/epyc_7313p_20_10_10_p2p-Gnutella24.json new file mode 100644 index 0000000..4e69947 --- /dev/null +++ b/pytorch/output_test/epyc_7313p_20_10_10_p2p-Gnutella24.json @@ -0,0 +1 @@ +{"CPU": "EPYC_7313P", "ITERATIONS": 348260, "MATRIX_FILE": "p2p-Gnutella24", "MATRIX_SHAPE": [26518, 26518], "MATRIX_SIZE": 703204324, "MATRIX_NNZ": 65369, "MATRIX_DENSITY": 9.295875717624285e-05, "TIME_S": 20.19377064704895, "TIME_S_1KI": 0.05798475462886622, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2049.2844516277314, "W": 105.72, "J_1KI": 5.884352069223372, "W_1KI": 0.30356630103945326, "W_D": 70.467, "J_D": 1365.9376414382457, "W_D_1KI": 0.20234020559352206, "J_D_1KI": 0.0005810032894777524} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_p2p-Gnutella24_10000.output b/pytorch/output_test/epyc_7313p_20_10_10_p2p-Gnutella24.output similarity index 69% rename from pytorch/output_cpu/epyc_7313p_10_10_p2p-Gnutella24_10000.output rename to pytorch/output_test/epyc_7313p_20_10_10_p2p-Gnutella24.output index 2557a07..cdc5f00 100644 --- a/pytorch/output_cpu/epyc_7313p_10_10_p2p-Gnutella24_10000.output +++ b/pytorch/output_test/epyc_7313p_20_10_10_p2p-Gnutella24.output @@ -1,17 +1,17 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3470896 queued and waiting for resources -srun: job 3470896 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476568 queued and waiting for resources +srun: job 3476568 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 9, 9, ..., 65369, 65369, 65369]), col_indices=tensor([ 1, 2, 3, ..., 15065, 9401, 26517]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(26518, 26518), nnz=65369, layout=torch.sparse_csr) -tensor([0.6457, 0.2082, 0.9929, ..., 0.5035, 0.5783, 0.4428]) +tensor([0.2462, 0.3933, 0.7513, ..., 0.2091, 0.2884, 0.4014]) Matrix: p2p-Gnutella24 Shape: torch.Size([26518, 26518]) Size: 703204324 NNZ: 65369 Density: 9.295875717624285e-05 -Time: 0.5904901027679443 seconds +Time: 20.19377064704895 seconds diff --git a/pytorch/output_test/epyc_7313p_20_10_10_p2p-Gnutella25.json b/pytorch/output_test/epyc_7313p_20_10_10_p2p-Gnutella25.json new file mode 100644 index 0000000..7b6b10b --- /dev/null +++ b/pytorch/output_test/epyc_7313p_20_10_10_p2p-Gnutella25.json @@ -0,0 +1 @@ +{"CPU": "EPYC_7313P", "ITERATIONS": 412691, "MATRIX_FILE": "p2p-Gnutella25", "MATRIX_SHAPE": [22687, 22687], "MATRIX_SIZE": 514699969, "MATRIX_NNZ": 54705, "MATRIX_DENSITY": 0.00010628522108964806, "TIME_S": 30.748587369918823, "TIME_S_1KI": 0.07450753074314395, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2195.6448175048827, "W": 104.96, "J_1KI": 5.320311849555437, "W_1KI": 0.25433072201719925, "W_D": 69.17124999999999, "J_D": 1446.9845329919456, "W_D_1KI": 0.16761027015369848, "J_D_1KI": 0.00040613987257705756} diff --git a/pytorch/output_test2/epyc_7313p_10_10_p2p-Gnutella25_100000.output b/pytorch/output_test/epyc_7313p_20_10_10_p2p-Gnutella25.output similarity index 69% rename from pytorch/output_test2/epyc_7313p_10_10_p2p-Gnutella25_100000.output rename to pytorch/output_test/epyc_7313p_20_10_10_p2p-Gnutella25.output index ce9222e..8d0fda9 100644 --- a/pytorch/output_test2/epyc_7313p_10_10_p2p-Gnutella25_100000.output +++ b/pytorch/output_test/epyc_7313p_20_10_10_p2p-Gnutella25.output @@ -1,17 +1,17 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3471810 queued and waiting for resources -srun: job 3471810 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476515 queued and waiting for resources +srun: job 3476515 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 9, 9, ..., 54704, 54704, 54705]), col_indices=tensor([ 1, 2, 3, ..., 17949, 22685, 144]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(22687, 22687), nnz=54705, layout=torch.sparse_csr) -tensor([0.2183, 0.6576, 0.4780, ..., 0.0534, 0.0208, 0.5648]) +tensor([0.4580, 0.2646, 0.5068, ..., 0.8695, 0.6231, 0.1460]) Matrix: p2p-Gnutella25 Shape: torch.Size([22687, 22687]) Size: 514699969 NNZ: 54705 Density: 0.00010628522108964806 -Time: 5.152007102966309 seconds +Time: 30.748587369918823 seconds diff --git a/pytorch/output_test/epyc_7313p_20_10_10_p2p-Gnutella30.json b/pytorch/output_test/epyc_7313p_20_10_10_p2p-Gnutella30.json new file mode 100644 index 0000000..c457c6c --- /dev/null +++ b/pytorch/output_test/epyc_7313p_20_10_10_p2p-Gnutella30.json @@ -0,0 +1 @@ +{"CPU": "EPYC_7313P", "ITERATIONS": 378073, "MATRIX_FILE": "p2p-Gnutella30", "MATRIX_SHAPE": [36682, 36682], "MATRIX_SIZE": 1345569124, "MATRIX_NNZ": 88328, "MATRIX_DENSITY": 6.564359899804003e-05, "TIME_S": 21.37001347541809, "TIME_S_1KI": 0.05652351126744859, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2379.664047117233, "W": 109.47999999999999, "J_1KI": 6.294191987042802, "W_1KI": 0.28957370666511495, "W_D": 73.58274999999999, "J_D": 1599.399202256262, "W_D_1KI": 0.19462577332948927, "J_D_1KI": 0.0005147835823491476} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_p2p-Gnutella30_10000.output b/pytorch/output_test/epyc_7313p_20_10_10_p2p-Gnutella30.output similarity index 69% rename from pytorch/output_cpu/epyc_7313p_10_10_p2p-Gnutella30_10000.output rename to pytorch/output_test/epyc_7313p_20_10_10_p2p-Gnutella30.output index c28b6a9..5cd2745 100644 --- a/pytorch/output_cpu/epyc_7313p_10_10_p2p-Gnutella30_10000.output +++ b/pytorch/output_test/epyc_7313p_20_10_10_p2p-Gnutella30.output @@ -1,17 +1,17 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3470892 queued and waiting for resources -srun: job 3470892 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476517 queued and waiting for resources +srun: job 3476517 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 10, 10, ..., 88328, 88328, 88328]), col_indices=tensor([ 1, 2, 3, ..., 36675, 36676, 36677]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(36682, 36682), nnz=88328, layout=torch.sparse_csr) -tensor([0.8398, 0.4628, 0.1152, ..., 0.1515, 0.0358, 0.8190]) +tensor([0.8310, 0.1313, 0.7903, ..., 0.2085, 0.0604, 0.3812]) Matrix: p2p-Gnutella30 Shape: torch.Size([36682, 36682]) Size: 1345569124 NNZ: 88328 Density: 6.564359899804003e-05 -Time: 0.5678565502166748 seconds +Time: 21.37001347541809 seconds diff --git a/pytorch/output_test/epyc_7313p_20_10_10_ri2010.json b/pytorch/output_test/epyc_7313p_20_10_10_ri2010.json new file mode 100644 index 0000000..623e5b0 --- /dev/null +++ b/pytorch/output_test/epyc_7313p_20_10_10_ri2010.json @@ -0,0 +1 @@ +{"CPU": "EPYC_7313P", "ITERATIONS": 354560, "MATRIX_FILE": "ri2010", "MATRIX_SHAPE": [25181, 25181], "MATRIX_SIZE": 634082761, "MATRIX_NNZ": 125750, "MATRIX_DENSITY": 0.00019831796057928155, "TIME_S": 20.56230664253235, "TIME_S_1KI": 0.057993870268875085, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2649.2086783885957, "W": 112.60000000000001, "J_1KI": 7.471820505382998, "W_1KI": 0.31757671480144406, "W_D": 76.87800000000001, "J_D": 1808.7554598326687, "W_D_1KI": 0.21682648916967512, "J_D_1KI": 0.0006115368038404647} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_ri2010_10000.output b/pytorch/output_test/epyc_7313p_20_10_10_ri2010.output similarity index 71% rename from pytorch/output_cpu/epyc_7313p_10_10_ri2010_10000.output rename to pytorch/output_test/epyc_7313p_20_10_10_ri2010.output index 533b0f1..6b8b411 100644 --- a/pytorch/output_cpu/epyc_7313p_10_10_ri2010_10000.output +++ b/pytorch/output_test/epyc_7313p_20_10_10_ri2010.output @@ -1,18 +1,18 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3470884 queued and waiting for resources -srun: job 3470884 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476484 queued and waiting for resources +srun: job 3476484 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 3, 8, ..., 125742, 125747, 125750]), col_indices=tensor([ 25, 56, 662, ..., 21738, 22279, 23882]), values=tensor([17171., 37318., 5284., ..., 25993., 24918., 803.]), size=(25181, 25181), nnz=125750, layout=torch.sparse_csr) -tensor([0.8353, 0.9273, 0.0726, ..., 0.3513, 0.9132, 0.6466]) +tensor([0.6635, 0.8632, 0.1964, ..., 0.1037, 0.6639, 0.7727]) Matrix: ri2010 Shape: torch.Size([25181, 25181]) Size: 634082761 NNZ: 125750 Density: 0.00019831796057928155 -Time: 0.610253095626831 seconds +Time: 20.56230664253235 seconds diff --git a/pytorch/output_test/epyc_7313p_20_10_10_rma10.json b/pytorch/output_test/epyc_7313p_20_10_10_rma10.json new file mode 100644 index 0000000..428682a --- /dev/null +++ b/pytorch/output_test/epyc_7313p_20_10_10_rma10.json @@ -0,0 +1 @@ +{"CPU": "EPYC_7313P", "ITERATIONS": 58657, "MATRIX_FILE": "rma10", "MATRIX_SHAPE": [46835, 46835], "MATRIX_SIZE": 2193517225, "MATRIX_NNZ": 2374001, "MATRIX_DENSITY": 0.0010822805369125833, "TIME_S": 23.640302419662476, "TIME_S_1KI": 0.40302610804614075, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 3140.030371599197, "W": 137.22, "J_1KI": 53.532065594885474, "W_1KI": 2.339362735905348, "W_D": 101.1375, "J_D": 2314.3479209125044, "W_D_1KI": 1.724218763318956, "J_D_1KI": 0.029394936040352492} diff --git a/pytorch/output_test2/epyc_7313p_10_10_rma10_100000.output b/pytorch/output_test/epyc_7313p_20_10_10_rma10.output similarity index 72% rename from pytorch/output_test2/epyc_7313p_10_10_rma10_100000.output rename to pytorch/output_test/epyc_7313p_20_10_10_rma10.output index 7f77177..ae22e1b 100644 --- a/pytorch/output_test2/epyc_7313p_10_10_rma10_100000.output +++ b/pytorch/output_test/epyc_7313p_20_10_10_rma10.output @@ -1,19 +1,19 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3471808 queued and waiting for resources -srun: job 3471808 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476518 queued and waiting for resources +srun: job 3476518 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 17, 34, ..., 2373939, 2373970, 2374001]), col_indices=tensor([ 0, 1, 2, ..., 46831, 46833, 46834]), values=tensor([ 1.2636e+05, -1.6615e+07, -8.2015e+04, ..., 8.3378e+01, 2.5138e+00, 1.2184e+03]), size=(46835, 46835), nnz=2374001, layout=torch.sparse_csr) -tensor([0.0908, 0.0974, 0.3859, ..., 0.1088, 0.9810, 0.2978]) +tensor([0.3635, 0.9483, 0.2648, ..., 0.0300, 0.4359, 0.9173]) Matrix: rma10 Shape: torch.Size([46835, 46835]) Size: 2193517225 NNZ: 2374001 Density: 0.0010822805369125833 -Time: 38.7850341796875 seconds +Time: 23.640302419662476 seconds diff --git a/pytorch/output_test/epyc_7313p_20_10_10_soc-sign-Slashdot090216.json b/pytorch/output_test/epyc_7313p_20_10_10_soc-sign-Slashdot090216.json new file mode 100644 index 0000000..aeb2ee7 --- /dev/null +++ b/pytorch/output_test/epyc_7313p_20_10_10_soc-sign-Slashdot090216.json @@ -0,0 +1 @@ +{"CPU": "EPYC_7313P", "ITERATIONS": 119682, "MATRIX_FILE": "soc-sign-Slashdot090216", "MATRIX_SHAPE": [81871, 81871], "MATRIX_SIZE": 6702860641, "MATRIX_NNZ": 545671, "MATRIX_DENSITY": 8.140867447881048e-05, "TIME_S": 20.252686262130737, "TIME_S_1KI": 0.1692208206925915, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2435.2421276807786, "W": 123.69, "J_1KI": 20.347605552052762, "W_1KI": 1.0334887451747128, "W_D": 88.36225, "J_D": 1739.6998439377546, "W_D_1KI": 0.7383086011263181, "J_D_1KI": 0.006168919312230061} diff --git a/pytorch/output_test2/epyc_7313p_10_10_soc-sign-Slashdot090216_100000.output b/pytorch/output_test/epyc_7313p_20_10_10_soc-sign-Slashdot090216.output similarity index 70% rename from pytorch/output_test2/epyc_7313p_10_10_soc-sign-Slashdot090216_100000.output rename to pytorch/output_test/epyc_7313p_20_10_10_soc-sign-Slashdot090216.output index 808d59b..1cc1309 100644 --- a/pytorch/output_test2/epyc_7313p_10_10_soc-sign-Slashdot090216_100000.output +++ b/pytorch/output_test/epyc_7313p_20_10_10_soc-sign-Slashdot090216.output @@ -1,18 +1,18 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3471835 queued and waiting for resources -srun: job 3471835 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476571 queued and waiting for resources +srun: job 3476571 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 29, 124, ..., 545669, 545669, 545671]), col_indices=tensor([ 1, 2, 3, ..., 81869, 81699, 81863]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(81871, 81871), nnz=545671, layout=torch.sparse_csr) -tensor([0.7693, 0.4110, 0.2737, ..., 0.8913, 0.4051, 0.1845]) +tensor([0.5432, 0.6789, 0.4167, ..., 0.6057, 0.5803, 0.8106]) Matrix: soc-sign-Slashdot090216 Shape: torch.Size([81871, 81871]) Size: 6702860641 NNZ: 545671 Density: 8.140867447881048e-05 -Time: 16.565748691558838 seconds +Time: 20.252686262130737 seconds diff --git a/pytorch/output_test/epyc_7313p_20_10_10_soc-sign-Slashdot090221.json b/pytorch/output_test/epyc_7313p_20_10_10_soc-sign-Slashdot090221.json new file mode 100644 index 0000000..24dad59 --- /dev/null +++ b/pytorch/output_test/epyc_7313p_20_10_10_soc-sign-Slashdot090221.json @@ -0,0 +1 @@ +{"CPU": "EPYC_7313P", "ITERATIONS": 126009, "MATRIX_FILE": "soc-sign-Slashdot090221", "MATRIX_SHAPE": [82144, 82144], "MATRIX_SIZE": 6747636736, "MATRIX_NNZ": 549202, "MATRIX_DENSITY": 8.13917555860553e-05, "TIME_S": 21.559077739715576, "TIME_S_1KI": 0.17109157075856152, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2628.299125556946, "W": 123.24000000000001, "J_1KI": 20.858027010427396, "W_1KI": 0.9780253791395854, "W_D": 88.09300000000002, "J_D": 1878.7305653009419, "W_D_1KI": 0.6991008578752312, "J_D_1KI": 0.005548023219573453} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_soc-sign-Slashdot090221_10000.output b/pytorch/output_test/epyc_7313p_20_10_10_soc-sign-Slashdot090221.output similarity index 70% rename from pytorch/output_cpu/epyc_7313p_10_10_soc-sign-Slashdot090221_10000.output rename to pytorch/output_test/epyc_7313p_20_10_10_soc-sign-Slashdot090221.output index 5482408..6accf09 100644 --- a/pytorch/output_cpu/epyc_7313p_10_10_soc-sign-Slashdot090221_10000.output +++ b/pytorch/output_test/epyc_7313p_20_10_10_soc-sign-Slashdot090221.output @@ -1,18 +1,18 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3470876 queued and waiting for resources -srun: job 3470876 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476476 queued and waiting for resources +srun: job 3476476 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 29, 124, ..., 549200, 549200, 549202]), col_indices=tensor([ 1, 2, 3, ..., 82142, 81974, 82136]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(82144, 82144), nnz=549202, layout=torch.sparse_csr) -tensor([0.5845, 0.4829, 0.3749, ..., 0.6026, 0.8058, 0.2362]) +tensor([0.1139, 0.6400, 0.0975, ..., 0.4554, 0.7678, 0.8954]) Matrix: soc-sign-Slashdot090221 Shape: torch.Size([82144, 82144]) Size: 6747636736 NNZ: 549202 Density: 8.13917555860553e-05 -Time: 1.6988587379455566 seconds +Time: 21.559077739715576 seconds diff --git a/pytorch/output_test/epyc_7313p_20_10_10_soc-sign-epinions.json b/pytorch/output_test/epyc_7313p_20_10_10_soc-sign-epinions.json new file mode 100644 index 0000000..ca58890 --- /dev/null +++ b/pytorch/output_test/epyc_7313p_20_10_10_soc-sign-epinions.json @@ -0,0 +1 @@ +{"CPU": "EPYC_7313P", "ITERATIONS": 76654, "MATRIX_FILE": "soc-sign-epinions", "MATRIX_SHAPE": [131828, 131828], "MATRIX_SIZE": 17378621584, "MATRIX_NNZ": 841372, "MATRIX_DENSITY": 4.841419648464106e-05, "TIME_S": 21.83398413658142, "TIME_S_1KI": 0.2848381576510217, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2661.028549144268, "W": 124.21, "J_1KI": 34.71480352159402, "W_1KI": 1.6203981527382785, "W_D": 88.33224999999999, "J_D": 1892.3970619124768, "W_D_1KI": 1.1523501708978003, "J_D_1KI": 0.015033138138881212} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_soc-sign-epinions_10000.output b/pytorch/output_test/epyc_7313p_20_10_10_soc-sign-epinions.output similarity index 71% rename from pytorch/output_cpu/epyc_7313p_10_10_soc-sign-epinions_10000.output rename to pytorch/output_test/epyc_7313p_20_10_10_soc-sign-epinions.output index ece723e..c4818e4 100644 --- a/pytorch/output_cpu/epyc_7313p_10_10_soc-sign-epinions_10000.output +++ b/pytorch/output_test/epyc_7313p_20_10_10_soc-sign-epinions.output @@ -1,19 +1,19 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3470886 queued and waiting for resources -srun: job 3470886 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476516 queued and waiting for resources +srun: job 3476516 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 1, 2, ..., 841371, 841371, 841372]), col_indices=tensor([ 1, 128552, 3, ..., 131824, 131826, 7714]), values=tensor([-1., -1., 1., ..., 1., 1., 1.]), size=(131828, 131828), nnz=841372, layout=torch.sparse_csr) -tensor([0.2911, 0.5946, 0.7956, ..., 0.3632, 0.5862, 0.9286]) +tensor([0.7115, 0.6150, 0.0225, ..., 0.1717, 0.5804, 0.3011]) Matrix: soc-sign-epinions Shape: torch.Size([131828, 131828]) Size: 17378621584 NNZ: 841372 Density: 4.841419648464106e-05 -Time: 2.818403720855713 seconds +Time: 21.83398413658142 seconds diff --git a/pytorch/output_test/epyc_7313p_20_10_10_sx-mathoverflow.json b/pytorch/output_test/epyc_7313p_20_10_10_sx-mathoverflow.json new file mode 100644 index 0000000..ac59cc8 --- /dev/null +++ b/pytorch/output_test/epyc_7313p_20_10_10_sx-mathoverflow.json @@ -0,0 +1 @@ +{"CPU": "EPYC_7313P", "ITERATIONS": 229626, "MATRIX_FILE": "sx-mathoverflow", "MATRIX_SHAPE": [24818, 24818], "MATRIX_SIZE": 615933124, "MATRIX_NNZ": 239978, "MATRIX_DENSITY": 0.00038961697406616504, "TIME_S": 21.080405950546265, "TIME_S_1KI": 0.09180321893229106, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2484.714141454697, "W": 116.73, "J_1KI": 10.820700362566507, "W_1KI": 0.5083483577643647, "W_D": 80.68350000000001, "J_D": 1717.4285396389964, "W_D_1KI": 0.3513691829322464, "J_D_1KI": 0.0015301803059420379} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_sx-mathoverflow_10000.output b/pytorch/output_test/epyc_7313p_20_10_10_sx-mathoverflow.output similarity index 70% rename from pytorch/output_cpu/epyc_7313p_10_10_sx-mathoverflow_10000.output rename to pytorch/output_test/epyc_7313p_20_10_10_sx-mathoverflow.output index f0fdd59..7a41df6 100644 --- a/pytorch/output_cpu/epyc_7313p_10_10_sx-mathoverflow_10000.output +++ b/pytorch/output_test/epyc_7313p_20_10_10_sx-mathoverflow.output @@ -1,18 +1,18 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3470882 queued and waiting for resources -srun: job 3470882 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476494 queued and waiting for resources +srun: job 3476494 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 317, 416, ..., 239976, 239977, 239978]), col_indices=tensor([ 0, 1, 2, ..., 1483, 2179, 24817]), values=tensor([151., 17., 6., ..., 1., 1., 1.]), size=(24818, 24818), nnz=239978, layout=torch.sparse_csr) -tensor([0.5148, 0.6679, 0.4736, ..., 0.5604, 0.1954, 0.1745]) +tensor([0.3978, 0.7775, 0.1403, ..., 0.8086, 0.5011, 0.4547]) Matrix: sx-mathoverflow Shape: torch.Size([24818, 24818]) Size: 615933124 NNZ: 239978 Density: 0.00038961697406616504 -Time: 0.9492478370666504 seconds +Time: 21.080405950546265 seconds diff --git a/pytorch/output_test/epyc_7313p_20_10_10_tn2010.json b/pytorch/output_test/epyc_7313p_20_10_10_tn2010.json new file mode 100644 index 0000000..1f2c706 --- /dev/null +++ b/pytorch/output_test/epyc_7313p_20_10_10_tn2010.json @@ -0,0 +1 @@ +{"CPU": "EPYC_7313P", "ITERATIONS": 121676, "MATRIX_FILE": "tn2010", "MATRIX_SHAPE": [240116, 240116], "MATRIX_SIZE": 57655693456, "MATRIX_NNZ": 1193966, "MATRIX_DENSITY": 2.070855328296721e-05, "TIME_S": 21.533467769622803, "TIME_S_1KI": 0.17697383025101748, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 3271.5562079071997, "W": 148.67, "J_1KI": 26.887440480515465, "W_1KI": 1.2218514744074425, "W_D": 112.09899999999999, "J_D": 2466.7934307539463, "W_D_1KI": 0.9212909694598769, "J_D_1KI": 0.007571673702783432} diff --git a/pytorch/output_test2/epyc_7313p_10_10_tn2010_100000.output b/pytorch/output_test/epyc_7313p_20_10_10_tn2010.output similarity index 71% rename from pytorch/output_test2/epyc_7313p_10_10_tn2010_100000.output rename to pytorch/output_test/epyc_7313p_20_10_10_tn2010.output index 0147812..8780eba 100644 --- a/pytorch/output_test2/epyc_7313p_10_10_tn2010_100000.output +++ b/pytorch/output_test/epyc_7313p_20_10_10_tn2010.output @@ -1,8 +1,8 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3471795 queued and waiting for resources -srun: job 3471795 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476486 queued and waiting for resources +srun: job 3476486 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 4, 20, ..., 1193961, 1193963, 1193966]), col_indices=tensor([ 1152, 1272, 1961, ..., 238254, 239142, @@ -10,12 +10,11 @@ tensor(crow_indices=tensor([ 0, 4, 20, ..., 1193961, values=tensor([ 5728., 2871., 418449., ..., 10058., 33324., 34928.]), size=(240116, 240116), nnz=1193966, layout=torch.sparse_csr) -tensor([3.0726e-01, 6.8295e-01, 5.4596e-01, ..., 3.6135e-01, 9.3459e-01, - 3.1018e-04]) +tensor([0.9402, 0.2615, 0.4528, ..., 0.1016, 0.6532, 0.2144]) Matrix: tn2010 Shape: torch.Size([240116, 240116]) Size: 57655693456 NNZ: 1193966 Density: 2.070855328296721e-05 -Time: 17.589671850204468 seconds +Time: 21.533467769622803 seconds diff --git a/pytorch/output_test/epyc_7313p_20_10_10_ut2010.json b/pytorch/output_test/epyc_7313p_20_10_10_ut2010.json new file mode 100644 index 0000000..cfb8563 --- /dev/null +++ b/pytorch/output_test/epyc_7313p_20_10_10_ut2010.json @@ -0,0 +1 @@ +{"CPU": "EPYC_7313P", "ITERATIONS": 271870, "MATRIX_FILE": "ut2010", "MATRIX_SHAPE": [115406, 115406], "MATRIX_SIZE": 13318544836, "MATRIX_NNZ": 572066, "MATRIX_DENSITY": 4.295259032005559e-05, "TIME_S": 20.712276458740234, "TIME_S_1KI": 0.07618448691926374, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 3097.387019622326, "W": 139.21, "J_1KI": 11.392897412816149, "W_1KI": 0.5120461985507779, "W_D": 103.01350000000001, "J_D": 2292.0241200047735, "W_D_1KI": 0.378907198293302, "J_D_1KI": 0.0013937072802931623} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_ut2010_10000.output b/pytorch/output_test/epyc_7313p_20_10_10_ut2010.output similarity index 72% rename from pytorch/output_cpu/epyc_7313p_10_10_ut2010_10000.output rename to pytorch/output_test/epyc_7313p_20_10_10_ut2010.output index 0f866b2..5457d68 100644 --- a/pytorch/output_cpu/epyc_7313p_10_10_ut2010_10000.output +++ b/pytorch/output_test/epyc_7313p_20_10_10_ut2010.output @@ -1,8 +1,8 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3470887 queued and waiting for resources -srun: job 3470887 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476520 queued and waiting for resources +srun: job 3476520 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 3, 9, ..., 572056, 572061, 572066]), col_indices=tensor([ 453, 1291, 1979, ..., 113521, 114509, @@ -10,11 +10,11 @@ tensor(crow_indices=tensor([ 0, 3, 9, ..., 572056, 572061, values=tensor([160642., 31335., 282373., ..., 88393., 99485., 18651.]), size=(115406, 115406), nnz=572066, layout=torch.sparse_csr) -tensor([0.5921, 0.3895, 0.8812, ..., 0.2001, 0.1496, 0.7049]) +tensor([0.6306, 0.2406, 0.4388, ..., 0.5283, 0.9288, 0.4627]) Matrix: ut2010 Shape: torch.Size([115406, 115406]) Size: 13318544836 NNZ: 572066 Density: 4.295259032005559e-05 -Time: 0.7757325172424316 seconds +Time: 20.712276458740234 seconds diff --git a/pytorch/output_test/epyc_7313p_20_10_10_va2010.json b/pytorch/output_test/epyc_7313p_20_10_10_va2010.json new file mode 100644 index 0000000..2817e42 --- /dev/null +++ b/pytorch/output_test/epyc_7313p_20_10_10_va2010.json @@ -0,0 +1 @@ +{"CPU": "EPYC_7313P", "ITERATIONS": 89073, "MATRIX_FILE": "va2010", "MATRIX_SHAPE": [285762, 285762], "MATRIX_SIZE": 81659920644, "MATRIX_NNZ": 1402128, "MATRIX_DENSITY": 1.717033263003816e-05, "TIME_S": 21.420818090438843, "TIME_S_1KI": 0.2404860966896685, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 3078.995559251309, "W": 146.27, "J_1KI": 34.56710292963422, "W_1KI": 1.6421362253432579, "W_D": 110.60275000000001, "J_D": 2328.197006159723, "W_D_1KI": 1.2417090476350858, "J_D_1KI": 0.01394035283009538} diff --git a/pytorch/output_test2/epyc_7313p_10_10_va2010_100000.output b/pytorch/output_test/epyc_7313p_20_10_10_va2010.output similarity index 72% rename from pytorch/output_test2/epyc_7313p_10_10_va2010_100000.output rename to pytorch/output_test/epyc_7313p_20_10_10_va2010.output index 9a57254..565c864 100644 --- a/pytorch/output_test2/epyc_7313p_10_10_va2010_100000.output +++ b/pytorch/output_test/epyc_7313p_20_10_10_va2010.output @@ -1,8 +1,8 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3471814 queued and waiting for resources -srun: job 3471814 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476512 queued and waiting for resources +srun: job 3476512 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 2, 8, ..., 1402119, 1402123, 1402128]), col_indices=tensor([ 2006, 2464, 1166, ..., 285581, 285634, @@ -10,11 +10,11 @@ tensor(crow_indices=tensor([ 0, 2, 8, ..., 1402119, values=tensor([125334., 3558., 1192., ..., 10148., 1763., 9832.]), size=(285762, 285762), nnz=1402128, layout=torch.sparse_csr) -tensor([0.7314, 0.5884, 0.7739, ..., 0.0933, 0.1510, 0.6060]) +tensor([0.9919, 0.7660, 0.5385, ..., 0.7873, 0.6966, 0.7083]) Matrix: va2010 Shape: torch.Size([285762, 285762]) Size: 81659920644 NNZ: 1402128 Density: 1.717033263003816e-05 -Time: 24.332262754440308 seconds +Time: 21.420818090438843 seconds diff --git a/pytorch/output_test/epyc_7313p_20_10_10_vt2010.json b/pytorch/output_test/epyc_7313p_20_10_10_vt2010.json new file mode 100644 index 0000000..a68d651 --- /dev/null +++ b/pytorch/output_test/epyc_7313p_20_10_10_vt2010.json @@ -0,0 +1 @@ +{"CPU": "EPYC_7313P", "ITERATIONS": 266269, "MATRIX_FILE": "vt2010", "MATRIX_SHAPE": [32580, 32580], "MATRIX_SIZE": 1061456400, "MATRIX_NNZ": 155598, "MATRIX_DENSITY": 0.00014658915806621921, "TIME_S": 26.02103877067566, "TIME_S_1KI": 0.09772462724040598, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 3227.329737491608, "W": 105.08, "J_1KI": 12.120561302636085, "W_1KI": 0.39463850467008926, "W_D": 69.62725, "J_D": 2138.4668296989203, "W_D_1KI": 0.261492137650271, "J_D_1KI": 0.0009820600131831758} diff --git a/pytorch/output_test2/epyc_7313p_10_10_vt2010_100000.output b/pytorch/output_test/epyc_7313p_20_10_10_vt2010.output similarity index 71% rename from pytorch/output_test2/epyc_7313p_10_10_vt2010_100000.output rename to pytorch/output_test/epyc_7313p_20_10_10_vt2010.output index c31abbe..b34ffe9 100644 --- a/pytorch/output_test2/epyc_7313p_10_10_vt2010_100000.output +++ b/pytorch/output_test/epyc_7313p_20_10_10_vt2010.output @@ -1,18 +1,18 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3471811 queued and waiting for resources -srun: job 3471811 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476513 queued and waiting for resources +srun: job 3476513 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 4, 7, ..., 155588, 155592, 155598]), col_indices=tensor([ 131, 561, 996, ..., 32237, 32238, 32570]), values=tensor([79040., 7820., 15136., ..., 2828., 17986., 2482.]), size=(32580, 32580), nnz=155598, layout=torch.sparse_csr) -tensor([0.9036, 0.7985, 0.5047, ..., 0.6402, 0.1482, 0.0115]) +tensor([0.0233, 0.8665, 0.6176, ..., 0.5417, 0.8614, 0.4079]) Matrix: vt2010 Shape: torch.Size([32580, 32580]) Size: 1061456400 NNZ: 155598 Density: 0.00014658915806621921 -Time: 7.804270267486572 seconds +Time: 26.02103877067566 seconds diff --git a/pytorch/output_test/xeon_4216_20_10_10_ASIC_680k.json b/pytorch/output_test/xeon_4216_20_10_10_ASIC_680k.json new file mode 100644 index 0000000..823ec93 --- /dev/null +++ b/pytorch/output_test/xeon_4216_20_10_10_ASIC_680k.json @@ -0,0 +1 @@ +{"CPU": "XEON_4216", "ITERATIONS": 8359, "MATRIX_FILE": "ASIC_680k", "MATRIX_SHAPE": [682862, 682862], "MATRIX_SIZE": 466300511044, "MATRIX_NNZ": 3871773, "MATRIX_DENSITY": 8.303171256088674e-06, "TIME_S": 21.087437391281128, "TIME_S_1KI": 2.522722501648657, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1858.5772673797608, "W": 89.92, "J_1KI": 222.34445117594936, "W_1KI": 10.75726761574351, "W_D": 72.8635, "J_D": 1506.0325258198977, "W_D_1KI": 8.716772341189136, "J_D_1KI": 1.0428008543114171} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_ASIC_680k_10000.output b/pytorch/output_test/xeon_4216_20_10_10_ASIC_680k.output similarity index 73% rename from pytorch/output_cpu/epyc_7313p_10_10_ASIC_680k_10000.output rename to pytorch/output_test/xeon_4216_20_10_10_ASIC_680k.output index 02b19c9..62c8471 100644 --- a/pytorch/output_cpu/epyc_7313p_10_10_ASIC_680k_10000.output +++ b/pytorch/output_test/xeon_4216_20_10_10_ASIC_680k.output @@ -1,8 +1,8 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3470899 queued and waiting for resources -srun: job 3470899 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476550 queued and waiting for resources +srun: job 3476550 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 3, 4, ..., 3871767, 3871770, 3871773]), col_indices=tensor([ 0, 11698, 11699, ..., 169456, 645874, @@ -10,11 +10,11 @@ tensor(crow_indices=tensor([ 0, 3, 4, ..., 3871767, values=tensor([ 3.8333e-04, -3.3333e-04, -5.0000e-05, ..., 0.0000e+00, 0.0000e+00, 7.9289e-02]), size=(682862, 682862), nnz=3871773, layout=torch.sparse_csr) -tensor([0.6720, 0.5431, 0.4163, ..., 0.4625, 0.4662, 0.2085]) +tensor([0.6315, 0.0620, 0.1454, ..., 0.8255, 0.8982, 0.8293]) Matrix: ASIC_680k Shape: torch.Size([682862, 682862]) Size: 466300511044 NNZ: 3871773 Density: 8.303171256088674e-06 -Time: 7.5851967334747314 seconds +Time: 21.087437391281128 seconds diff --git a/pytorch/output_test/xeon_4216_20_10_10_Oregon-2.json b/pytorch/output_test/xeon_4216_20_10_10_Oregon-2.json new file mode 100644 index 0000000..c854150 --- /dev/null +++ b/pytorch/output_test/xeon_4216_20_10_10_Oregon-2.json @@ -0,0 +1 @@ +{"CPU": "XEON_4216", "ITERATIONS": 342492, "MATRIX_FILE": "Oregon-2", "MATRIX_SHAPE": [11806, 11806], "MATRIX_SIZE": 139381636, "MATRIX_NNZ": 65460, "MATRIX_DENSITY": 0.0004696458003979807, "TIME_S": 21.41038227081299, "TIME_S_1KI": 0.06251352519420304, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1782.7575745725633, "W": 85.14, "J_1KI": 5.205253187147622, "W_1KI": 0.248589748081707, "W_D": 67.88925, "J_D": 1421.541868329227, "W_D_1KI": 0.1982214183105007, "J_D_1KI": 0.0005787621851327933} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_Oregon-2_10000.output b/pytorch/output_test/xeon_4216_20_10_10_Oregon-2.output similarity index 69% rename from pytorch/output_cpu/epyc_7313p_10_10_Oregon-2_10000.output rename to pytorch/output_test/xeon_4216_20_10_10_Oregon-2.output index 018245e..51bc010 100644 --- a/pytorch/output_cpu/epyc_7313p_10_10_Oregon-2_10000.output +++ b/pytorch/output_test/xeon_4216_20_10_10_Oregon-2.output @@ -1,17 +1,17 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3470898 queued and waiting for resources -srun: job 3470898 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476551 queued and waiting for resources +srun: job 3476551 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 583, 584, ..., 65459, 65460, 65460]), col_indices=tensor([ 2, 23, 27, ..., 3324, 958, 841]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(11806, 11806), nnz=65460, layout=torch.sparse_csr) -tensor([0.6755, 0.7426, 0.3350, ..., 0.5898, 0.3954, 0.3897]) +tensor([0.6724, 0.2164, 0.1485, ..., 0.2563, 0.2809, 0.1323]) Matrix: Oregon-2 Shape: torch.Size([11806, 11806]) Size: 139381636 NNZ: 65460 Density: 0.0004696458003979807 -Time: 0.4882948398590088 seconds +Time: 21.41038227081299 seconds diff --git a/pytorch/output_test/xeon_4216_20_10_10_as-caida.json b/pytorch/output_test/xeon_4216_20_10_10_as-caida.json new file mode 100644 index 0000000..1857f1f --- /dev/null +++ b/pytorch/output_test/xeon_4216_20_10_10_as-caida.json @@ -0,0 +1 @@ +{"CPU": "XEON_4216", "ITERATIONS": 191471, "MATRIX_FILE": "as-caida", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 106762, "MATRIX_DENSITY": 0.00010842726485909405, "TIME_S": 21.31909680366516, "TIME_S_1KI": 0.111343737713101, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1842.3941788578034, "W": 82.28, "J_1KI": 9.622314495969642, "W_1KI": 0.4297256503595845, "W_D": 66.23825000000001, "J_D": 1483.1911305022838, "W_D_1KI": 0.34594403330008205, "J_D_1KI": 0.0018067698674999453} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_as-caida_10000.output b/pytorch/output_test/xeon_4216_20_10_10_as-caida.output similarity index 70% rename from pytorch/output_cpu/epyc_7313p_10_10_as-caida_10000.output rename to pytorch/output_test/xeon_4216_20_10_10_as-caida.output index 604b029..9f2d3b0 100644 --- a/pytorch/output_cpu/epyc_7313p_10_10_as-caida_10000.output +++ b/pytorch/output_test/xeon_4216_20_10_10_as-caida.output @@ -1,18 +1,18 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3470879 queued and waiting for resources -srun: job 3470879 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476410 queued and waiting for resources +srun: job 3476410 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 0, 0, ..., 106761, 106761, 106762]), col_indices=tensor([ 106, 329, 1040, ..., 155, 160, 12170]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379), nnz=106762, layout=torch.sparse_csr) -tensor([0.6412, 0.3070, 0.9642, ..., 0.0959, 0.1216, 0.7825]) +tensor([0.4011, 0.5137, 0.9095, ..., 0.8172, 0.5463, 0.1847]) Matrix: as-caida Shape: torch.Size([31379, 31379]) Size: 984641641 NNZ: 106762 Density: 0.00010842726485909405 -Time: 0.6748511791229248 seconds +Time: 21.31909680366516 seconds diff --git a/pytorch/output_test/xeon_4216_20_10_10_dc2.json b/pytorch/output_test/xeon_4216_20_10_10_dc2.json new file mode 100644 index 0000000..103437d --- /dev/null +++ b/pytorch/output_test/xeon_4216_20_10_10_dc2.json @@ -0,0 +1 @@ +{"CPU": "XEON_4216", "ITERATIONS": 49841, "MATRIX_FILE": "dc2", "MATRIX_SHAPE": [116835, 116835], "MATRIX_SIZE": 13650417225, "MATRIX_NNZ": 766396, "MATRIX_DENSITY": 5.614451099680581e-05, "TIME_S": 20.360435247421265, "TIME_S_1KI": 0.4085077596240297, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1816.120480132103, "W": 88.97, "J_1KI": 36.43828334367495, "W_1KI": 1.7850765434080376, "W_D": 71.93825, "J_D": 1468.455986623168, "W_D_1KI": 1.4433548684817719, "J_D_1KI": 0.028959187586159424} diff --git a/pytorch/output_test2/epyc_7313p_10_10_dc2_100000.output b/pytorch/output_test/xeon_4216_20_10_10_dc2.output similarity index 73% rename from pytorch/output_test2/epyc_7313p_10_10_dc2_100000.output rename to pytorch/output_test/xeon_4216_20_10_10_dc2.output index a8105ee..1e7f089 100644 --- a/pytorch/output_test2/epyc_7313p_10_10_dc2_100000.output +++ b/pytorch/output_test/xeon_4216_20_10_10_dc2.output @@ -1,8 +1,8 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3471792 queued and waiting for resources -srun: job 3471792 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476417 queued and waiting for resources +srun: job 3476417 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 1, 2, ..., 766390, 766394, 766396]), col_indices=tensor([ 0, 1, 2, ..., 116833, 89, @@ -10,11 +10,11 @@ tensor(crow_indices=tensor([ 0, 1, 2, ..., 766390, 766394, values=tensor([-1.0000e+00, -1.0000e+00, -1.0000e+00, ..., 1.0331e+01, -1.0000e-03, 1.0000e-03]), size=(116835, 116835), nnz=766396, layout=torch.sparse_csr) -tensor([0.7371, 0.0385, 0.5513, ..., 0.8646, 0.4043, 0.6262]) +tensor([0.6884, 0.3977, 0.1702, ..., 0.4512, 0.7891, 0.8337]) Matrix: dc2 Shape: torch.Size([116835, 116835]) Size: 13650417225 NNZ: 766396 Density: 5.614451099680581e-05 -Time: 20.959667444229126 seconds +Time: 20.360435247421265 seconds diff --git a/pytorch/output_test/xeon_4216_20_10_10_de2010.json b/pytorch/output_test/xeon_4216_20_10_10_de2010.json new file mode 100644 index 0000000..3d987ba --- /dev/null +++ b/pytorch/output_test/xeon_4216_20_10_10_de2010.json @@ -0,0 +1 @@ +{"CPU": "XEON_4216", "ITERATIONS": 423754, "MATRIX_FILE": "de2010", "MATRIX_SHAPE": [24115, 24115], "MATRIX_SIZE": 581533225, "MATRIX_NNZ": 116056, "MATRIX_DENSITY": 0.0001995689928120616, "TIME_S": 21.152066707611084, "TIME_S_1KI": 0.0499159104282463, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1923.06989515543, "W": 91.37, "J_1KI": 4.538175203432723, "W_1KI": 0.21562038352440333, "W_D": 74.42125000000001, "J_D": 1566.3485327222947, "W_D_1KI": 0.17562371092662254, "J_D_1KI": 0.00041444732303794787} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_de2010_10000.output b/pytorch/output_test/xeon_4216_20_10_10_de2010.output similarity index 71% rename from pytorch/output_cpu/epyc_7313p_10_10_de2010_10000.output rename to pytorch/output_test/xeon_4216_20_10_10_de2010.output index 19c0d6c..36bb395 100644 --- a/pytorch/output_cpu/epyc_7313p_10_10_de2010_10000.output +++ b/pytorch/output_test/xeon_4216_20_10_10_de2010.output @@ -1,19 +1,19 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3470877 queued and waiting for resources -srun: job 3470877 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476416 queued and waiting for resources +srun: job 3476416 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 13, 21, ..., 116047, 116051, 116056]), col_indices=tensor([ 250, 251, 757, ..., 23334, 23553, 24050]), values=tensor([ 14900., 33341., 20255., ..., 164227., 52413., 16949.]), size=(24115, 24115), nnz=116056, layout=torch.sparse_csr) -tensor([0.4656, 0.5143, 0.3514, ..., 0.7050, 0.9241, 0.6135]) +tensor([0.2150, 0.5992, 0.9896, ..., 0.4383, 0.4421, 0.9824]) Matrix: de2010 Shape: torch.Size([24115, 24115]) Size: 581533225 NNZ: 116056 Density: 0.0001995689928120616 -Time: 0.5970535278320312 seconds +Time: 21.152066707611084 seconds diff --git a/pytorch/output_test/xeon_4216_20_10_10_email-Enron.json b/pytorch/output_test/xeon_4216_20_10_10_email-Enron.json new file mode 100644 index 0000000..b36a698 --- /dev/null +++ b/pytorch/output_test/xeon_4216_20_10_10_email-Enron.json @@ -0,0 +1 @@ +{"CPU": "XEON_4216", "ITERATIONS": 89822, "MATRIX_FILE": "email-Enron", "MATRIX_SHAPE": [36692, 36692], "MATRIX_SIZE": 1346302864, "MATRIX_NNZ": 367662, "MATRIX_DENSITY": 0.0002730901120626302, "TIME_S": 21.060860872268677, "TIME_S_1KI": 0.23447330133228694, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1824.7371728897094, "W": 85.08, "J_1KI": 20.315036103512607, "W_1KI": 0.9472066976909888, "W_D": 67.90899999999999, "J_D": 1456.465405192375, "W_D_1KI": 0.7560397230077263, "J_D_1KI": 0.008417088497336134} diff --git a/pytorch/output_test2/epyc_7313p_10_10_email-Enron_100000.output b/pytorch/output_test/xeon_4216_20_10_10_email-Enron.output similarity index 70% rename from pytorch/output_test2/epyc_7313p_10_10_email-Enron_100000.output rename to pytorch/output_test/xeon_4216_20_10_10_email-Enron.output index 75a8121..9f68b39 100644 --- a/pytorch/output_test2/epyc_7313p_10_10_email-Enron_100000.output +++ b/pytorch/output_test/xeon_4216_20_10_10_email-Enron.output @@ -1,18 +1,18 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3471780 queued and waiting for resources -srun: job 3471780 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476413 queued and waiting for resources +srun: job 3476413 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 1, 71, ..., 367660, 367661, 367662]), col_indices=tensor([ 1, 0, 2, ..., 36690, 36689, 8203]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(36692, 36692), nnz=367662, layout=torch.sparse_csr) -tensor([0.2683, 0.9357, 0.3150, ..., 0.9101, 0.5382, 0.3808]) +tensor([0.4981, 0.3503, 0.5022, ..., 0.5730, 0.6139, 0.1017]) Matrix: email-Enron Shape: torch.Size([36692, 36692]) Size: 1346302864 NNZ: 367662 Density: 0.0002730901120626302 -Time: 12.811992168426514 seconds +Time: 21.060860872268677 seconds diff --git a/pytorch/output_test/xeon_4216_20_10_10_fl2010.json b/pytorch/output_test/xeon_4216_20_10_10_fl2010.json new file mode 100644 index 0000000..29f571f --- /dev/null +++ b/pytorch/output_test/xeon_4216_20_10_10_fl2010.json @@ -0,0 +1 @@ +{"CPU": "XEON_4216", "ITERATIONS": 22899, "MATRIX_FILE": "fl2010", "MATRIX_SHAPE": [484481, 484481], "MATRIX_SIZE": 234721839361, "MATRIX_NNZ": 2346294, "MATRIX_DENSITY": 9.99606174861054e-06, "TIME_S": 21.102646827697754, "TIME_S_1KI": 0.9215532044062079, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1979.5554644393922, "W": 94.17, "J_1KI": 86.44724505172243, "W_1KI": 4.112406655312459, "W_D": 76.87475, "J_D": 1615.9905642976762, "W_D_1KI": 3.35712258177213, "J_D_1KI": 0.14660564137176865} diff --git a/pytorch/output_test2/epyc_7313p_10_10_fl2010_100000.output b/pytorch/output_test/xeon_4216_20_10_10_fl2010.output similarity index 72% rename from pytorch/output_test2/epyc_7313p_10_10_fl2010_100000.output rename to pytorch/output_test/xeon_4216_20_10_10_fl2010.output index fbb1249..4af0d8f 100644 --- a/pytorch/output_test2/epyc_7313p_10_10_fl2010_100000.output +++ b/pytorch/output_test/xeon_4216_20_10_10_fl2010.output @@ -1,19 +1,19 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3471812 queued and waiting for resources -srun: job 3471812 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476507 queued and waiting for resources +srun: job 3476507 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 2, 5, ..., 2346288, 2346292, 2346294]), col_indices=tensor([ 1513, 5311, 947, ..., 484460, 482463, 484022]), values=tensor([28364., 12497., 11567., ..., 8532., 22622., 35914.]), size=(484481, 484481), nnz=2346294, layout=torch.sparse_csr) -tensor([0.8274, 0.0613, 0.8619, ..., 0.9125, 0.2679, 0.3813]) +tensor([0.9877, 0.2442, 0.6939, ..., 0.8588, 0.9632, 0.3911]) Matrix: fl2010 Shape: torch.Size([484481, 484481]) Size: 234721839361 NNZ: 2346294 Density: 9.99606174861054e-06 -Time: 38.57364296913147 seconds +Time: 21.102646827697754 seconds diff --git a/pytorch/output_test/xeon_4216_20_10_10_ga2010.json b/pytorch/output_test/xeon_4216_20_10_10_ga2010.json new file mode 100644 index 0000000..cd42722 --- /dev/null +++ b/pytorch/output_test/xeon_4216_20_10_10_ga2010.json @@ -0,0 +1 @@ +{"CPU": "XEON_4216", "ITERATIONS": 43450, "MATRIX_FILE": "ga2010", "MATRIX_SHAPE": [291086, 291086], "MATRIX_SIZE": 84731059396, "MATRIX_NNZ": 1418056, "MATRIX_DENSITY": 1.6735964475229304e-05, "TIME_S": 21.27401065826416, "TIME_S_1KI": 0.4896204984640774, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1655.9512757730486, "W": 78.34, "J_1KI": 38.11165191652586, "W_1KI": 1.8029919447640967, "W_D": 61.378750000000004, "J_D": 1297.424296245277, "W_D_1KI": 1.4126294591484465, "J_D_1KI": 0.03251161010698381} diff --git a/pytorch/output_test2/epyc_7313p_10_10_ga2010_100000.output b/pytorch/output_test/xeon_4216_20_10_10_ga2010.output similarity index 72% rename from pytorch/output_test2/epyc_7313p_10_10_ga2010_100000.output rename to pytorch/output_test/xeon_4216_20_10_10_ga2010.output index 940efc8..a8b376e 100644 --- a/pytorch/output_test2/epyc_7313p_10_10_ga2010_100000.output +++ b/pytorch/output_test/xeon_4216_20_10_10_ga2010.output @@ -1,19 +1,19 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3471787 queued and waiting for resources -srun: job 3471787 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476412 queued and waiting for resources +srun: job 3476412 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 3, 10, ..., 1418047, 1418054, 1418056]), col_indices=tensor([ 1566, 1871, 1997, ..., 291064, 289820, 290176]), values=tensor([18760., 17851., 18847., ..., 65219., 56729., 77629.]), size=(291086, 291086), nnz=1418056, layout=torch.sparse_csr) -tensor([0.9989, 0.8982, 0.5822, ..., 0.5453, 0.7727, 0.8878]) +tensor([0.3863, 0.8469, 0.0606, ..., 0.1284, 0.0807, 0.3967]) Matrix: ga2010 Shape: torch.Size([291086, 291086]) Size: 84731059396 NNZ: 1418056 Density: 1.6735964475229304e-05 -Time: 33.085010051727295 seconds +Time: 21.27401065826416 seconds diff --git a/pytorch/output_test/xeon_4216_20_10_10_mac_econ_fwd500.json b/pytorch/output_test/xeon_4216_20_10_10_mac_econ_fwd500.json new file mode 100644 index 0000000..7f9686a --- /dev/null +++ b/pytorch/output_test/xeon_4216_20_10_10_mac_econ_fwd500.json @@ -0,0 +1 @@ +{"CPU": "XEON_4216", "ITERATIONS": 95140, "MATRIX_FILE": "mac_econ_fwd500", "MATRIX_SHAPE": [206500, 206500], "MATRIX_SIZE": 42642250000, "MATRIX_NNZ": 1273389, "MATRIX_DENSITY": 2.9862143765866013e-05, "TIME_S": 21.01556706428528, "TIME_S_1KI": 0.22089097187602774, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2015.6398850345613, "W": 94.51, "J_1KI": 21.186040414489817, "W_1KI": 0.9933781795249106, "W_D": 77.14250000000001, "J_D": 1645.2385973048213, "W_D_1KI": 0.8108314063485391, "J_D_1KI": 0.008522507949848004} diff --git a/pytorch/output_test2/epyc_7313p_10_10_mac_econ_fwd500_100000.output b/pytorch/output_test/xeon_4216_20_10_10_mac_econ_fwd500.output similarity index 73% rename from pytorch/output_test2/epyc_7313p_10_10_mac_econ_fwd500_100000.output rename to pytorch/output_test/xeon_4216_20_10_10_mac_econ_fwd500.output index 3aa52d7..7a3741a 100644 --- a/pytorch/output_test2/epyc_7313p_10_10_mac_econ_fwd500_100000.output +++ b/pytorch/output_test/xeon_4216_20_10_10_mac_econ_fwd500.output @@ -1,8 +1,8 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3471816 queued and waiting for resources -srun: job 3471816 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476502 queued and waiting for resources +srun: job 3476502 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 3, 8, ..., 1273376, 1273379, 1273389]), col_indices=tensor([ 3, 30, 44, ..., 206363, 206408, @@ -10,11 +10,11 @@ tensor(crow_indices=tensor([ 0, 3, 8, ..., 1273376, values=tensor([-3.7877e-03, -1.5420e-01, 9.5305e-04, ..., 1.2290e-01, 2.2235e-01, -1.0000e+00]), size=(206500, 206500), nnz=1273389, layout=torch.sparse_csr) -tensor([0.1058, 0.5873, 0.8242, ..., 0.1574, 0.8351, 0.1537]) +tensor([0.6065, 0.5629, 0.6231, ..., 0.8894, 0.3274, 0.9878]) Matrix: mac_econ_fwd500 Shape: torch.Size([206500, 206500]) Size: 42642250000 NNZ: 1273389 Density: 2.9862143765866013e-05 -Time: 10.857311248779297 seconds +Time: 21.01556706428528 seconds diff --git a/pytorch/output_test/xeon_4216_20_10_10_mc2depi.json b/pytorch/output_test/xeon_4216_20_10_10_mc2depi.json new file mode 100644 index 0000000..a5fe2c3 --- /dev/null +++ b/pytorch/output_test/xeon_4216_20_10_10_mc2depi.json @@ -0,0 +1 @@ +{"CPU": "XEON_4216", "ITERATIONS": 47795, "MATRIX_FILE": "mc2depi", "MATRIX_SHAPE": [525825, 525825], "MATRIX_SIZE": 276491930625, "MATRIX_NNZ": 2100225, "MATRIX_DENSITY": 7.595972132902821e-06, "TIME_S": 21.546258449554443, "TIME_S_1KI": 0.45080570037774753, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2051.987362103462, "W": 94.42, "J_1KI": 42.93309681145438, "W_1KI": 1.975520451930118, "W_D": 77.2705, "J_D": 1679.28499749434, "W_D_1KI": 1.616706768490428, "J_D_1KI": 0.03382585560185015} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_mc2depi_10000.output b/pytorch/output_test/xeon_4216_20_10_10_mc2depi.output similarity index 72% rename from pytorch/output_cpu/epyc_7313p_10_10_mc2depi_10000.output rename to pytorch/output_test/xeon_4216_20_10_10_mc2depi.output index 0da53ce..b64e5c4 100644 --- a/pytorch/output_cpu/epyc_7313p_10_10_mc2depi_10000.output +++ b/pytorch/output_test/xeon_4216_20_10_10_mc2depi.output @@ -1,19 +1,19 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3470883 queued and waiting for resources -srun: job 3470883 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476409 queued and waiting for resources +srun: job 3476409 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 2, 5, ..., 2100220, 2100223, 2100225]), col_indices=tensor([ 0, 1, 1, ..., 525824, 525821, 525824]), values=tensor([-2025., 2025., -2026., ..., 2025., 1024., -1024.]), size=(525825, 525825), nnz=2100225, layout=torch.sparse_csr) -tensor([0.1112, 0.0723, 0.0629, ..., 0.0188, 0.2120, 0.5563]) +tensor([0.0758, 0.7859, 0.0925, ..., 0.0430, 0.0592, 0.4957]) Matrix: mc2depi Shape: torch.Size([525825, 525825]) Size: 276491930625 NNZ: 2100225 Density: 7.595972132902821e-06 -Time: 1.4909443855285645 seconds +Time: 21.546258449554443 seconds diff --git a/pytorch/output_test/xeon_4216_20_10_10_p2p-Gnutella04.json b/pytorch/output_test/xeon_4216_20_10_10_p2p-Gnutella04.json new file mode 100644 index 0000000..98725e1 --- /dev/null +++ b/pytorch/output_test/xeon_4216_20_10_10_p2p-Gnutella04.json @@ -0,0 +1 @@ +{"CPU": "XEON_4216", "ITERATIONS": 591873, "MATRIX_FILE": "p2p-Gnutella04", "MATRIX_SHAPE": [10879, 10879], "MATRIX_SIZE": 118352641, "MATRIX_NNZ": 39994, "MATRIX_DENSITY": 0.0003379223282393842, "TIME_S": 21.74970245361328, "TIME_S_1KI": 0.03674724553006014, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1805.301536707878, "W": 85.74, "J_1KI": 3.050150178683397, "W_1KI": 0.14486215792915033, "W_D": 68.5855, "J_D": 1444.1043683913945, "W_D_1KI": 0.11587874425763635, "J_D_1KI": 0.00019578312282810055} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_p2p-Gnutella04_10000.output b/pytorch/output_test/xeon_4216_20_10_10_p2p-Gnutella04.output similarity index 69% rename from pytorch/output_cpu/epyc_7313p_10_10_p2p-Gnutella04_10000.output rename to pytorch/output_test/xeon_4216_20_10_10_p2p-Gnutella04.output index 23b38e4..b0d711a 100644 --- a/pytorch/output_cpu/epyc_7313p_10_10_p2p-Gnutella04_10000.output +++ b/pytorch/output_test/xeon_4216_20_10_10_p2p-Gnutella04.output @@ -1,17 +1,17 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3470891 queued and waiting for resources -srun: job 3470891 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476504 queued and waiting for resources +srun: job 3476504 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 10, 20, ..., 39994, 39994, 39994]), col_indices=tensor([ 1, 2, 3, ..., 9711, 10875, 10876]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(10879, 10879), nnz=39994, layout=torch.sparse_csr) -tensor([0.7836, 0.6102, 0.9911, ..., 0.3070, 0.4164, 0.1677]) +tensor([0.2334, 0.1020, 0.8896, ..., 0.8020, 0.9592, 0.7319]) Matrix: p2p-Gnutella04 Shape: torch.Size([10879, 10879]) Size: 118352641 NNZ: 39994 Density: 0.0003379223282393842 -Time: 0.3917062282562256 seconds +Time: 21.74970245361328 seconds diff --git a/pytorch/output_test/xeon_4216_20_10_10_p2p-Gnutella24.json b/pytorch/output_test/xeon_4216_20_10_10_p2p-Gnutella24.json new file mode 100644 index 0000000..9d93767 --- /dev/null +++ b/pytorch/output_test/xeon_4216_20_10_10_p2p-Gnutella24.json @@ -0,0 +1 @@ +{"CPU": "XEON_4216", "ITERATIONS": 358390, "MATRIX_FILE": "p2p-Gnutella24", "MATRIX_SHAPE": [26518, 26518], "MATRIX_SIZE": 703204324, "MATRIX_NNZ": 65369, "MATRIX_DENSITY": 9.295875717624285e-05, "TIME_S": 20.834523916244507, "TIME_S_1KI": 0.05813366421006308, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1800.9590803027154, "W": 87.09, "J_1KI": 5.025137644194078, "W_1KI": 0.24300343201540223, "W_D": 69.9445, "J_D": 1446.4023698729277, "W_D_1KI": 0.19516309048801586, "J_D_1KI": 0.0005445550670722281} diff --git a/pytorch/output_test2/epyc_7313p_10_10_p2p-Gnutella24_100000.output b/pytorch/output_test/xeon_4216_20_10_10_p2p-Gnutella24.output similarity index 69% rename from pytorch/output_test2/epyc_7313p_10_10_p2p-Gnutella24_100000.output rename to pytorch/output_test/xeon_4216_20_10_10_p2p-Gnutella24.output index c23a0b2..3b042fd 100644 --- a/pytorch/output_test2/epyc_7313p_10_10_p2p-Gnutella24_100000.output +++ b/pytorch/output_test/xeon_4216_20_10_10_p2p-Gnutella24.output @@ -1,17 +1,17 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3471834 queued and waiting for resources -srun: job 3471834 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476549 queued and waiting for resources +srun: job 3476549 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 9, 9, ..., 65369, 65369, 65369]), col_indices=tensor([ 1, 2, 3, ..., 15065, 9401, 26517]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(26518, 26518), nnz=65369, layout=torch.sparse_csr) -tensor([0.1894, 0.0975, 0.5835, ..., 0.5367, 0.6746, 0.5669]) +tensor([0.9412, 0.5504, 0.9733, ..., 0.7745, 0.1801, 0.5360]) Matrix: p2p-Gnutella24 Shape: torch.Size([26518, 26518]) Size: 703204324 NNZ: 65369 Density: 9.295875717624285e-05 -Time: 5.905890703201294 seconds +Time: 20.834523916244507 seconds diff --git a/pytorch/output_test/xeon_4216_20_10_10_p2p-Gnutella25.json b/pytorch/output_test/xeon_4216_20_10_10_p2p-Gnutella25.json new file mode 100644 index 0000000..4105ec6 --- /dev/null +++ b/pytorch/output_test/xeon_4216_20_10_10_p2p-Gnutella25.json @@ -0,0 +1 @@ +{"CPU": "XEON_4216", "ITERATIONS": 426716, "MATRIX_FILE": "p2p-Gnutella25", "MATRIX_SHAPE": [22687, 22687], "MATRIX_SIZE": 514699969, "MATRIX_NNZ": 54705, "MATRIX_DENSITY": 0.00010628522108964806, "TIME_S": 20.72452974319458, "TIME_S_1KI": 0.04856750096831284, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1875.1789210319519, "W": 87.32, "J_1KI": 4.394442488755875, "W_1KI": 0.2046325893568556, "W_D": 70.04149999999998, "J_D": 1504.1267109191415, "W_D_1KI": 0.16414078684652084, "J_D_1KI": 0.000384660492802053} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_p2p-Gnutella25_10000.output b/pytorch/output_test/xeon_4216_20_10_10_p2p-Gnutella25.output similarity index 69% rename from pytorch/output_cpu/epyc_7313p_10_10_p2p-Gnutella25_10000.output rename to pytorch/output_test/xeon_4216_20_10_10_p2p-Gnutella25.output index ebb9072..921cb49 100644 --- a/pytorch/output_cpu/epyc_7313p_10_10_p2p-Gnutella25_10000.output +++ b/pytorch/output_test/xeon_4216_20_10_10_p2p-Gnutella25.output @@ -1,17 +1,17 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3470893 queued and waiting for resources -srun: job 3470893 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476505 queued and waiting for resources +srun: job 3476505 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 9, 9, ..., 54704, 54704, 54705]), col_indices=tensor([ 1, 2, 3, ..., 17949, 22685, 144]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(22687, 22687), nnz=54705, layout=torch.sparse_csr) -tensor([0.1041, 0.8814, 0.8390, ..., 0.9768, 0.6919, 0.6912]) +tensor([0.2160, 0.3964, 0.8952, ..., 0.9231, 0.1940, 0.2732]) Matrix: p2p-Gnutella25 Shape: torch.Size([22687, 22687]) Size: 514699969 NNZ: 54705 Density: 0.00010628522108964806 -Time: 0.5463590621948242 seconds +Time: 20.72452974319458 seconds diff --git a/pytorch/output_test/xeon_4216_20_10_10_p2p-Gnutella30.json b/pytorch/output_test/xeon_4216_20_10_10_p2p-Gnutella30.json new file mode 100644 index 0000000..34db7f7 --- /dev/null +++ b/pytorch/output_test/xeon_4216_20_10_10_p2p-Gnutella30.json @@ -0,0 +1 @@ +{"CPU": "XEON_4216", "ITERATIONS": 413059, "MATRIX_FILE": "p2p-Gnutella30", "MATRIX_SHAPE": [36682, 36682], "MATRIX_SIZE": 1345569124, "MATRIX_NNZ": 88328, "MATRIX_DENSITY": 6.564359899804003e-05, "TIME_S": 20.61072611808777, "TIME_S_1KI": 0.04989777760099107, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1884.4326162958146, "W": 90.69, "J_1KI": 4.562139104330893, "W_1KI": 0.21955701243648001, "W_D": 73.50825, "J_D": 1527.4158547450304, "W_D_1KI": 0.1779606545311929, "J_D_1KI": 0.00043083592060987145} diff --git a/pytorch/output_test2/epyc_7313p_10_10_p2p-Gnutella30_100000.output b/pytorch/output_test/xeon_4216_20_10_10_p2p-Gnutella30.output similarity index 69% rename from pytorch/output_test2/epyc_7313p_10_10_p2p-Gnutella30_100000.output rename to pytorch/output_test/xeon_4216_20_10_10_p2p-Gnutella30.output index 1505e8b..6d800fd 100644 --- a/pytorch/output_test2/epyc_7313p_10_10_p2p-Gnutella30_100000.output +++ b/pytorch/output_test/xeon_4216_20_10_10_p2p-Gnutella30.output @@ -1,17 +1,17 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3471815 queued and waiting for resources -srun: job 3471815 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476503 queued and waiting for resources +srun: job 3476503 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 10, 10, ..., 88328, 88328, 88328]), col_indices=tensor([ 1, 2, 3, ..., 36675, 36676, 36677]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(36682, 36682), nnz=88328, layout=torch.sparse_csr) -tensor([0.5178, 0.0281, 0.3608, ..., 0.6911, 0.2357, 0.6596]) +tensor([0.3731, 0.0521, 0.2564, ..., 0.6292, 0.0263, 0.7518]) Matrix: p2p-Gnutella30 Shape: torch.Size([36682, 36682]) Size: 1345569124 NNZ: 88328 Density: 6.564359899804003e-05 -Time: 5.514855861663818 seconds +Time: 20.61072611808777 seconds diff --git a/pytorch/output_test/xeon_4216_20_10_10_ri2010.json b/pytorch/output_test/xeon_4216_20_10_10_ri2010.json new file mode 100644 index 0000000..e43c998 --- /dev/null +++ b/pytorch/output_test/xeon_4216_20_10_10_ri2010.json @@ -0,0 +1 @@ +{"CPU": "XEON_4216", "ITERATIONS": 403912, "MATRIX_FILE": "ri2010", "MATRIX_SHAPE": [25181, 25181], "MATRIX_SIZE": 634082761, "MATRIX_NNZ": 125750, "MATRIX_DENSITY": 0.00019831796057928155, "TIME_S": 21.314026355743408, "TIME_S_1KI": 0.05276898521396593, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1981.9616649270056, "W": 91.85, "J_1KI": 4.906914538134558, "W_1KI": 0.22740101804353421, "W_D": 74.8665, "J_D": 1615.4875665460825, "W_D_1KI": 0.18535349284002456, "J_D_1KI": 0.0004588957318426404} diff --git a/pytorch/output_test2/epyc_7313p_10_10_ri2010_100000.output b/pytorch/output_test/xeon_4216_20_10_10_ri2010.output similarity index 70% rename from pytorch/output_test2/epyc_7313p_10_10_ri2010_100000.output rename to pytorch/output_test/xeon_4216_20_10_10_ri2010.output index 9a4dc4d..bb58dd8 100644 --- a/pytorch/output_test2/epyc_7313p_10_10_ri2010_100000.output +++ b/pytorch/output_test/xeon_4216_20_10_10_ri2010.output @@ -1,18 +1,18 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3471781 queued and waiting for resources -srun: job 3471781 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476414 queued and waiting for resources +srun: job 3476414 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 3, 8, ..., 125742, 125747, 125750]), col_indices=tensor([ 25, 56, 662, ..., 21738, 22279, 23882]), values=tensor([17171., 37318., 5284., ..., 25993., 24918., 803.]), size=(25181, 25181), nnz=125750, layout=torch.sparse_csr) -tensor([0.1942, 0.1978, 0.3462, ..., 0.1743, 0.2436, 0.9955]) +tensor([0.5620, 0.5155, 0.6623, ..., 0.4422, 0.0933, 0.8792]) Matrix: ri2010 Shape: torch.Size([25181, 25181]) Size: 634082761 NNZ: 125750 Density: 0.00019831796057928155 -Time: 6.2883217334747314 seconds +Time: 21.314026355743408 seconds diff --git a/pytorch/output_test/xeon_4216_20_10_10_rma10.json b/pytorch/output_test/xeon_4216_20_10_10_rma10.json new file mode 100644 index 0000000..1db12d1 --- /dev/null +++ b/pytorch/output_test/xeon_4216_20_10_10_rma10.json @@ -0,0 +1 @@ +{"CPU": "XEON_4216", "ITERATIONS": 41800, "MATRIX_FILE": "rma10", "MATRIX_SHAPE": [46835, 46835], "MATRIX_SIZE": 2193517225, "MATRIX_NNZ": 2374001, "MATRIX_DENSITY": 0.0010822805369125833, "TIME_S": 21.44151496887207, "TIME_S_1KI": 0.5129549035615328, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2021.3178783774376, "W": 94.05, "J_1KI": 48.35688704252243, "W_1KI": 2.25, "W_D": 76.95474999999999, "J_D": 1653.9076236158012, "W_D_1KI": 1.8410227272727269, "J_D_1KI": 0.04404360591561548} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_rma10_10000.output b/pytorch/output_test/xeon_4216_20_10_10_rma10.output similarity index 72% rename from pytorch/output_cpu/epyc_7313p_10_10_rma10_10000.output rename to pytorch/output_test/xeon_4216_20_10_10_rma10.output index a52f402..0d6e71d 100644 --- a/pytorch/output_cpu/epyc_7313p_10_10_rma10_10000.output +++ b/pytorch/output_test/xeon_4216_20_10_10_rma10.output @@ -1,19 +1,19 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3470888 queued and waiting for resources -srun: job 3470888 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476506 queued and waiting for resources +srun: job 3476506 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 17, 34, ..., 2373939, 2373970, 2374001]), col_indices=tensor([ 0, 1, 2, ..., 46831, 46833, 46834]), values=tensor([ 1.2636e+05, -1.6615e+07, -8.2015e+04, ..., 8.3378e+01, 2.5138e+00, 1.2184e+03]), size=(46835, 46835), nnz=2374001, layout=torch.sparse_csr) -tensor([0.0694, 0.3886, 0.4209, ..., 0.6373, 0.6766, 0.6929]) +tensor([0.9139, 0.5351, 0.8928, ..., 0.3550, 0.3135, 0.2632]) Matrix: rma10 Shape: torch.Size([46835, 46835]) Size: 2193517225 NNZ: 2374001 Density: 0.0010822805369125833 -Time: 3.9620909690856934 seconds +Time: 21.44151496887207 seconds diff --git a/pytorch/output_test/xeon_4216_20_10_10_soc-sign-Slashdot090216.json b/pytorch/output_test/xeon_4216_20_10_10_soc-sign-Slashdot090216.json new file mode 100644 index 0000000..a689e80 --- /dev/null +++ b/pytorch/output_test/xeon_4216_20_10_10_soc-sign-Slashdot090216.json @@ -0,0 +1 @@ +{"CPU": "XEON_4216", "ITERATIONS": 76573, "MATRIX_FILE": "soc-sign-Slashdot090216", "MATRIX_SHAPE": [81871, 81871], "MATRIX_SIZE": 6702860641, "MATRIX_NNZ": 545671, "MATRIX_DENSITY": 8.140867447881048e-05, "TIME_S": 21.045573234558105, "TIME_S_1KI": 0.2748432637425477, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1924.5487319564818, "W": 90.32, "J_1KI": 25.133516147421176, "W_1KI": 1.1795280320739685, "W_D": 73.1285, "J_D": 1558.2303138272762, "W_D_1KI": 0.9550167813720241, "J_D_1KI": 0.012471978130307341} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_soc-sign-Slashdot090216_10000.output b/pytorch/output_test/xeon_4216_20_10_10_soc-sign-Slashdot090216.output similarity index 70% rename from pytorch/output_cpu/epyc_7313p_10_10_soc-sign-Slashdot090216_10000.output rename to pytorch/output_test/xeon_4216_20_10_10_soc-sign-Slashdot090216.output index 1a4ac52..1643cfa 100644 --- a/pytorch/output_cpu/epyc_7313p_10_10_soc-sign-Slashdot090216_10000.output +++ b/pytorch/output_test/xeon_4216_20_10_10_soc-sign-Slashdot090216.output @@ -1,18 +1,18 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3470897 queued and waiting for resources -srun: job 3470897 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476552 queued and waiting for resources +srun: job 3476552 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 29, 124, ..., 545669, 545669, 545671]), col_indices=tensor([ 1, 2, 3, ..., 81869, 81699, 81863]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(81871, 81871), nnz=545671, layout=torch.sparse_csr) -tensor([0.6601, 0.6699, 0.4597, ..., 0.4006, 0.0724, 0.2095]) +tensor([0.4154, 0.0852, 0.3251, ..., 0.7418, 0.5922, 0.5972]) Matrix: soc-sign-Slashdot090216 Shape: torch.Size([81871, 81871]) Size: 6702860641 NNZ: 545671 Density: 8.140867447881048e-05 -Time: 1.620380163192749 seconds +Time: 21.045573234558105 seconds diff --git a/pytorch/output_test/xeon_4216_20_10_10_soc-sign-Slashdot090221.json b/pytorch/output_test/xeon_4216_20_10_10_soc-sign-Slashdot090221.json new file mode 100644 index 0000000..e287ebd --- /dev/null +++ b/pytorch/output_test/xeon_4216_20_10_10_soc-sign-Slashdot090221.json @@ -0,0 +1 @@ +{"CPU": "XEON_4216", "ITERATIONS": 75939, "MATRIX_FILE": "soc-sign-Slashdot090221", "MATRIX_SHAPE": [82144, 82144], "MATRIX_SIZE": 6747636736, "MATRIX_NNZ": 549202, "MATRIX_DENSITY": 8.13917555860553e-05, "TIME_S": 21.407752990722656, "TIME_S_1KI": 0.28190722804781015, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1896.3932697772977, "W": 90.6, "J_1KI": 24.972586810167343, "W_1KI": 1.1930628530794454, "W_D": 73.37625, "J_D": 1535.8744664624332, "W_D_1KI": 0.966252518468771, "J_D_1KI": 0.0127240616609222} diff --git a/pytorch/output_test2/epyc_7313p_10_10_soc-sign-Slashdot090221_100000.output b/pytorch/output_test/xeon_4216_20_10_10_soc-sign-Slashdot090221.output similarity index 70% rename from pytorch/output_test2/epyc_7313p_10_10_soc-sign-Slashdot090221_100000.output rename to pytorch/output_test/xeon_4216_20_10_10_soc-sign-Slashdot090221.output index 19faa39..1ef860d 100644 --- a/pytorch/output_test2/epyc_7313p_10_10_soc-sign-Slashdot090221_100000.output +++ b/pytorch/output_test/xeon_4216_20_10_10_soc-sign-Slashdot090221.output @@ -1,18 +1,18 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3471777 queued and waiting for resources -srun: job 3471777 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476408 queued and waiting for resources +srun: job 3476408 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 29, 124, ..., 549200, 549200, 549202]), col_indices=tensor([ 1, 2, 3, ..., 82142, 81974, 82136]), values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(82144, 82144), nnz=549202, layout=torch.sparse_csr) -tensor([0.7544, 0.2107, 0.1548, ..., 0.8853, 0.5512, 0.8288]) +tensor([0.5088, 0.5274, 0.1979, ..., 0.1078, 0.2507, 0.2157]) Matrix: soc-sign-Slashdot090221 Shape: torch.Size([82144, 82144]) Size: 6747636736 NNZ: 549202 Density: 8.13917555860553e-05 -Time: 16.887201070785522 seconds +Time: 21.407752990722656 seconds diff --git a/pytorch/output_test/xeon_4216_20_10_10_soc-sign-epinions.json b/pytorch/output_test/xeon_4216_20_10_10_soc-sign-epinions.json new file mode 100644 index 0000000..5e37148 --- /dev/null +++ b/pytorch/output_test/xeon_4216_20_10_10_soc-sign-epinions.json @@ -0,0 +1 @@ +{"CPU": "XEON_4216", "ITERATIONS": 41188, "MATRIX_FILE": "soc-sign-epinions", "MATRIX_SHAPE": [131828, 131828], "MATRIX_SIZE": 17378621584, "MATRIX_NNZ": 841372, "MATRIX_DENSITY": 4.841419648464106e-05, "TIME_S": 20.150396585464478, "TIME_S_1KI": 0.48922978987725746, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1860.381990737915, "W": 91.56, "J_1KI": 45.168058432988126, "W_1KI": 2.222977566281441, "W_D": 74.42875000000001, "J_D": 1512.2969210696222, "W_D_1KI": 1.807049383315529, "J_D_1KI": 0.04387320052722951} diff --git a/pytorch/output_test2/epyc_7313p_10_10_soc-sign-epinions_100000.output b/pytorch/output_test/xeon_4216_20_10_10_soc-sign-epinions.output similarity index 71% rename from pytorch/output_test2/epyc_7313p_10_10_soc-sign-epinions_100000.output rename to pytorch/output_test/xeon_4216_20_10_10_soc-sign-epinions.output index 7cd92e2..9a66335 100644 --- a/pytorch/output_test2/epyc_7313p_10_10_soc-sign-epinions_100000.output +++ b/pytorch/output_test/xeon_4216_20_10_10_soc-sign-epinions.output @@ -1,19 +1,19 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3471807 queued and waiting for resources -srun: job 3471807 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476499 queued and waiting for resources +srun: job 3476499 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 1, 2, ..., 841371, 841371, 841372]), col_indices=tensor([ 1, 128552, 3, ..., 131824, 131826, 7714]), values=tensor([-1., -1., 1., ..., 1., 1., 1.]), size=(131828, 131828), nnz=841372, layout=torch.sparse_csr) -tensor([0.8235, 0.7193, 0.6409, ..., 0.8649, 0.4430, 0.4079]) +tensor([0.0679, 0.0714, 0.0893, ..., 0.5812, 0.2775, 0.7418]) Matrix: soc-sign-epinions Shape: torch.Size([131828, 131828]) Size: 17378621584 NNZ: 841372 Density: 4.841419648464106e-05 -Time: 28.74003553390503 seconds +Time: 20.150396585464478 seconds diff --git a/pytorch/output_test/xeon_4216_20_10_10_sx-mathoverflow.json b/pytorch/output_test/xeon_4216_20_10_10_sx-mathoverflow.json new file mode 100644 index 0000000..27dd2cd --- /dev/null +++ b/pytorch/output_test/xeon_4216_20_10_10_sx-mathoverflow.json @@ -0,0 +1 @@ +{"CPU": "XEON_4216", "ITERATIONS": 142684, "MATRIX_FILE": "sx-mathoverflow", "MATRIX_SHAPE": [24818, 24818], "MATRIX_SIZE": 615933124, "MATRIX_NNZ": 239978, "MATRIX_DENSITY": 0.00038961697406616504, "TIME_S": 20.939576387405396, "TIME_S_1KI": 0.14675490165264077, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1765.209924120903, "W": 84.19, "J_1KI": 12.371463682829912, "W_1KI": 0.5900451347032604, "W_D": 67.227, "J_D": 1409.547066977978, "W_D_1KI": 0.47116004597572264, "J_D_1KI": 0.003302122494293142} diff --git a/pytorch/output_test2/epyc_7313p_10_10_sx-mathoverflow_100000.output b/pytorch/output_test/xeon_4216_20_10_10_sx-mathoverflow.output similarity index 70% rename from pytorch/output_test2/epyc_7313p_10_10_sx-mathoverflow_100000.output rename to pytorch/output_test/xeon_4216_20_10_10_sx-mathoverflow.output index b179d74..326658e 100644 --- a/pytorch/output_test2/epyc_7313p_10_10_sx-mathoverflow_100000.output +++ b/pytorch/output_test/xeon_4216_20_10_10_sx-mathoverflow.output @@ -1,18 +1,18 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3471793 queued and waiting for resources -srun: job 3471793 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476411 queued and waiting for resources +srun: job 3476411 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 317, 416, ..., 239976, 239977, 239978]), col_indices=tensor([ 0, 1, 2, ..., 1483, 2179, 24817]), values=tensor([151., 17., 6., ..., 1., 1., 1.]), size=(24818, 24818), nnz=239978, layout=torch.sparse_csr) -tensor([0.4836, 0.4937, 0.4802, ..., 0.5967, 0.6196, 0.8699]) +tensor([0.4572, 0.7870, 0.6302, ..., 0.8277, 0.5418, 0.0159]) Matrix: sx-mathoverflow Shape: torch.Size([24818, 24818]) Size: 615933124 NNZ: 239978 Density: 0.00038961697406616504 -Time: 9.806709051132202 seconds +Time: 20.939576387405396 seconds diff --git a/pytorch/output_test/xeon_4216_20_10_10_tn2010.json b/pytorch/output_test/xeon_4216_20_10_10_tn2010.json new file mode 100644 index 0000000..5004ad4 --- /dev/null +++ b/pytorch/output_test/xeon_4216_20_10_10_tn2010.json @@ -0,0 +1 @@ +{"CPU": "XEON_4216", "ITERATIONS": 52501, "MATRIX_FILE": "tn2010", "MATRIX_SHAPE": [240116, 240116], "MATRIX_SIZE": 57655693456, "MATRIX_NNZ": 1193966, "MATRIX_DENSITY": 2.070855328296721e-05, "TIME_S": 21.56684708595276, "TIME_S_1KI": 0.4107892627940946, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2042.7551445579527, "W": 94.38, "J_1KI": 38.908880679567105, "W_1KI": 1.7976800441896343, "W_D": 77.35999999999999, "J_D": 1674.3752700042721, "W_D_1KI": 1.4734957429382294, "J_D_1KI": 0.028066050988328404} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_tn2010_10000.output b/pytorch/output_test/xeon_4216_20_10_10_tn2010.output similarity index 72% rename from pytorch/output_cpu/epyc_7313p_10_10_tn2010_10000.output rename to pytorch/output_test/xeon_4216_20_10_10_tn2010.output index c8e88f5..4763085 100644 --- a/pytorch/output_cpu/epyc_7313p_10_10_tn2010_10000.output +++ b/pytorch/output_test/xeon_4216_20_10_10_tn2010.output @@ -1,8 +1,8 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3470881 queued and waiting for resources -srun: job 3470881 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476415 queued and waiting for resources +srun: job 3476415 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 4, 20, ..., 1193961, 1193963, 1193966]), col_indices=tensor([ 1152, 1272, 1961, ..., 238254, 239142, @@ -10,11 +10,11 @@ tensor(crow_indices=tensor([ 0, 4, 20, ..., 1193961, values=tensor([ 5728., 2871., 418449., ..., 10058., 33324., 34928.]), size=(240116, 240116), nnz=1193966, layout=torch.sparse_csr) -tensor([0.0655, 0.4633, 0.1355, ..., 0.7193, 0.0926, 0.7299]) +tensor([0.8288, 0.6918, 0.1632, ..., 0.1033, 0.8045, 0.3588]) Matrix: tn2010 Shape: torch.Size([240116, 240116]) Size: 57655693456 NNZ: 1193966 Density: 2.070855328296721e-05 -Time: 1.7836709022521973 seconds +Time: 21.56684708595276 seconds diff --git a/pytorch/output_test/xeon_4216_20_10_10_ut2010.json b/pytorch/output_test/xeon_4216_20_10_10_ut2010.json new file mode 100644 index 0000000..2a56406 --- /dev/null +++ b/pytorch/output_test/xeon_4216_20_10_10_ut2010.json @@ -0,0 +1 @@ +{"CPU": "XEON_4216", "ITERATIONS": 127924, "MATRIX_FILE": "ut2010", "MATRIX_SHAPE": [115406, 115406], "MATRIX_SIZE": 13318544836, "MATRIX_NNZ": 572066, "MATRIX_DENSITY": 4.295259032005559e-05, "TIME_S": 21.34226894378662, "TIME_S_1KI": 0.1668355347220742, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2007.6915735912323, "W": 94.36, "J_1KI": 15.694408974009821, "W_1KI": 0.7376254651199149, "W_D": 77.17775, "J_D": 1642.1059595562817, "W_D_1KI": 0.6033093868234265, "J_D_1KI": 0.004716154801471393} diff --git a/pytorch/output_test2/epyc_7313p_10_10_ut2010_100000.output b/pytorch/output_test/xeon_4216_20_10_10_ut2010.output similarity index 72% rename from pytorch/output_test2/epyc_7313p_10_10_ut2010_100000.output rename to pytorch/output_test/xeon_4216_20_10_10_ut2010.output index eb6b60c..881538d 100644 --- a/pytorch/output_test2/epyc_7313p_10_10_ut2010_100000.output +++ b/pytorch/output_test/xeon_4216_20_10_10_ut2010.output @@ -1,8 +1,8 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3471813 queued and waiting for resources -srun: job 3471813 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476508 queued and waiting for resources +srun: job 3476508 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 3, 9, ..., 572056, 572061, 572066]), col_indices=tensor([ 453, 1291, 1979, ..., 113521, 114509, @@ -10,11 +10,11 @@ tensor(crow_indices=tensor([ 0, 3, 9, ..., 572056, 572061, values=tensor([160642., 31335., 282373., ..., 88393., 99485., 18651.]), size=(115406, 115406), nnz=572066, layout=torch.sparse_csr) -tensor([0.9535, 0.8300, 0.1451, ..., 0.8613, 0.5153, 0.2159]) +tensor([0.6411, 0.3961, 0.2005, ..., 0.7969, 0.3944, 0.9610]) Matrix: ut2010 Shape: torch.Size([115406, 115406]) Size: 13318544836 NNZ: 572066 Density: 4.295259032005559e-05 -Time: 7.522066831588745 seconds +Time: 21.34226894378662 seconds diff --git a/pytorch/output_test/xeon_4216_20_10_10_va2010.json b/pytorch/output_test/xeon_4216_20_10_10_va2010.json new file mode 100644 index 0000000..0c589b6 --- /dev/null +++ b/pytorch/output_test/xeon_4216_20_10_10_va2010.json @@ -0,0 +1 @@ +{"CPU": "XEON_4216", "ITERATIONS": 42803, "MATRIX_FILE": "va2010", "MATRIX_SHAPE": [285762, 285762], "MATRIX_SIZE": 81659920644, "MATRIX_NNZ": 1402128, "MATRIX_DENSITY": 1.717033263003816e-05, "TIME_S": 21.75146985054016, "TIME_S_1KI": 0.5081762925622073, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2052.319766521454, "W": 94.5, "J_1KI": 47.94803557043791, "W_1KI": 2.207789173656052, "W_D": 77.42575, "J_D": 1681.5068482830522, "W_D_1KI": 1.8088860593883604, "J_D_1KI": 0.04226073077560827} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_va2010_10000.output b/pytorch/output_test/xeon_4216_20_10_10_va2010.output similarity index 72% rename from pytorch/output_cpu/epyc_7313p_10_10_va2010_10000.output rename to pytorch/output_test/xeon_4216_20_10_10_va2010.output index aa37cd6..0a5045a 100644 --- a/pytorch/output_cpu/epyc_7313p_10_10_va2010_10000.output +++ b/pytorch/output_test/xeon_4216_20_10_10_va2010.output @@ -1,8 +1,8 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3470895 queued and waiting for resources -srun: job 3470895 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476501 queued and waiting for resources +srun: job 3476501 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 2, 8, ..., 1402119, 1402123, 1402128]), col_indices=tensor([ 2006, 2464, 1166, ..., 285581, 285634, @@ -10,11 +10,11 @@ tensor(crow_indices=tensor([ 0, 2, 8, ..., 1402119, values=tensor([125334., 3558., 1192., ..., 10148., 1763., 9832.]), size=(285762, 285762), nnz=1402128, layout=torch.sparse_csr) -tensor([0.7826, 0.8027, 0.4606, ..., 0.4410, 0.5591, 0.5693]) +tensor([0.6170, 0.3785, 0.5272, ..., 0.7518, 0.8278, 0.9372]) Matrix: va2010 Shape: torch.Size([285762, 285762]) Size: 81659920644 NNZ: 1402128 Density: 1.717033263003816e-05 -Time: 2.389526844024658 seconds +Time: 21.75146985054016 seconds diff --git a/pytorch/output_test/xeon_4216_20_10_10_vt2010.json b/pytorch/output_test/xeon_4216_20_10_10_vt2010.json new file mode 100644 index 0000000..5c24a48 --- /dev/null +++ b/pytorch/output_test/xeon_4216_20_10_10_vt2010.json @@ -0,0 +1 @@ +{"CPU": "XEON_4216", "ITERATIONS": 345897, "MATRIX_FILE": "vt2010", "MATRIX_SHAPE": [32580, 32580], "MATRIX_SIZE": 1061456400, "MATRIX_NNZ": 155598, "MATRIX_DENSITY": 0.00014658915806621921, "TIME_S": 20.66720962524414, "TIME_S_1KI": 0.059749606458697646, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 2033.30526542902, "W": 90.89, "J_1KI": 5.878354728225513, "W_1KI": 0.26276608354510156, "W_D": 73.72274999999999, "J_D": 1649.2557570349572, "W_D_1KI": 0.2131349794881135, "J_D_1KI": 0.0006161804799929272} diff --git a/pytorch/output_cpu/epyc_7313p_10_10_vt2010_10000.output b/pytorch/output_test/xeon_4216_20_10_10_vt2010.output similarity index 71% rename from pytorch/output_cpu/epyc_7313p_10_10_vt2010_10000.output rename to pytorch/output_test/xeon_4216_20_10_10_vt2010.output index b5a2376..c900a9f 100644 --- a/pytorch/output_cpu/epyc_7313p_10_10_vt2010_10000.output +++ b/pytorch/output_test/xeon_4216_20_10_10_vt2010.output @@ -1,18 +1,18 @@ srun: Job time limit was unset; set to partition default of 60 minutes -srun: job 3470894 queued and waiting for resources -srun: job 3470894 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) - ).to_sparse_csr().type(torch.float) +srun: job 3476500 queued and waiting for resources +srun: job 3476500 has been allocated resources +/nfshomes/vut/ampere_research/pytorch/spmv.py:47: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at ../aten/src/ATen/SparseCsrTensorImpl.cpp:53.) + matrix = matrix.to_sparse_csr().type(torch.float32) tensor(crow_indices=tensor([ 0, 4, 7, ..., 155588, 155592, 155598]), col_indices=tensor([ 131, 561, 996, ..., 32237, 32238, 32570]), values=tensor([79040., 7820., 15136., ..., 2828., 17986., 2482.]), size=(32580, 32580), nnz=155598, layout=torch.sparse_csr) -tensor([0.7387, 0.6923, 0.8988, ..., 0.9956, 0.0628, 0.8750]) +tensor([0.6666, 0.1292, 0.0273, ..., 0.0169, 0.0420, 0.0601]) Matrix: vt2010 Shape: torch.Size([32580, 32580]) Size: 1061456400 NNZ: 155598 Density: 0.00014658915806621921 -Time: 0.8104038238525391 seconds +Time: 20.66720962524414 seconds diff --git a/pytorch/output_test2/altra_10_10_ASIC_680k_100000.json b/pytorch/output_test2/altra_10_10_ASIC_680k_100000.json deleted file mode 100644 index 94de9d0..0000000 --- a/pytorch/output_test2/altra_10_10_ASIC_680k_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 100000, "MATRIX_FILE": "ASIC_680k", "MATRIX_SHAPE": [682862, 682862], "MATRIX_SIZE": 466300511044, "MATRIX_NNZ": 3871773, "MATRIX_DENSITY": 8.303171256088674e-06, "TIME_S": 143.85276532173157, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [22.6, 22.52, 22.76, 22.68, 22.88, 22.84, 22.72, 22.48, 22.36, 22.64], "POWER": [101.6, 97.76, 86.32, 69.24, 56.32, 52.24, 57.16, 57.16, 75.88, 89.6, 100.72, 101.6, 103.6, 103.0, 104.08, 109.28, 107.8, 106.44, 104.68, 104.68, 100.04, 98.72, 98.28, 101.4, 98.96, 97.04, 94.12, 92.84, 88.4, 88.8, 93.36, 93.36, 94.12, 94.92, 95.6, 92.16, 91.6, 94.88, 95.88, 97.28, 98.36, 98.64, 96.52, 96.52, 97.24, 98.36, 95.12, 94.92, 98.72, 97.52, 94.56, 96.2, 98.04, 98.52, 102.44, 106.08, 106.08, 108.4, 107.52, 104.56, 103.16, 101.8, 103.24, 107.64, 105.52, 103.64, 104.84, 101.64, 101.64, 98.92, 95.64, 96.16, 100.24, 104.36, 105.52, 105.64, 102.0, 97.16, 95.4, 98.28, 98.28, 100.16, 102.76, 101.96, 103.16, 101.8, 105.32, 100.96, 98.44, 97.68, 97.6, 97.16, 97.16, 100.4, 101.48, 100.6, 98.96, 95.88, 93.68, 93.72, 94.44, 98.64, 100.44, 99.52, 101.52, 101.52, 98.16, 97.68, 98.28, 101.56, 100.2, 102.72, 103.8, 100.68, 103.12, 102.24, 101.28, 101.28, 100.04, 97.48, 95.08, 95.8, 94.92, 96.12, 95.16, 100.08, 104.08, 104.48, 107.4, 107.4, 109.68, 102.6, 100.44, 102.16, 99.48, 97.88, 95.96, 98.92, 102.84, 101.36, 102.48, 102.48, 100.92, 100.68, 96.48, 100.0, 102.04], "JOULES": 14125.656173429492, "POWER_AFTER": [23.04, 23.28, 23.0, 22.88, 22.84, 22.84, 22.76, 22.8, 22.96, 22.8]} diff --git a/pytorch/output_test2/altra_10_10_ASIC_680k_100000.output b/pytorch/output_test2/altra_10_10_ASIC_680k_100000.output deleted file mode 100644 index 145bbe8..0000000 --- a/pytorch/output_test2/altra_10_10_ASIC_680k_100000.output +++ /dev/null @@ -1,26 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3471856 queued and waiting for resources -srun: job 3471856 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 4, ..., 3871767, - 3871770, 3871773]), - col_indices=tensor([ 0, 11698, 11699, ..., 169456, 645874, - 682861]), - values=tensor([ 3.8333e-04, -3.3333e-04, -5.0000e-05, ..., - 0.0000e+00, 0.0000e+00, 7.9289e-02]), - size=(682862, 682862), nnz=3871773, layout=torch.sparse_csr) -tensor([0.6052, 0.7917, 0.7066, ..., 0.3876, 0.8366, 0.5267]) -Matrix: ASIC_680k -Shape: torch.Size([682862, 682862]) -Size: 466300511044 -NNZ: 3871773 -Density: 8.303171256088674e-06 -Time: 143.85276532173157 seconds - diff --git a/pytorch/output_test2/altra_10_10_Oregon-2_100000.json b/pytorch/output_test2/altra_10_10_Oregon-2_100000.json deleted file mode 100644 index ebc7194..0000000 --- a/pytorch/output_test2/altra_10_10_Oregon-2_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 100000, "MATRIX_FILE": "Oregon-2", "MATRIX_SHAPE": [11806, 11806], "MATRIX_SIZE": 139381636, "MATRIX_NNZ": 65460, "MATRIX_DENSITY": 0.0004696458003979807, "TIME_S": 8.373449563980103, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [22.8, 22.64, 22.52, 22.56, 22.68, 22.44, 22.44, 22.36, 22.24, 22.36], "POWER": [97.2, 97.84, 89.08, 72.32, 58.36, 58.4, 58.84, 74.52, 74.52, 88.96, 99.68], "JOULES": 612.2254525375366, "POWER_AFTER": [22.16, 21.92, 21.92, 21.96, 21.96, 22.36, 22.24, 22.24, 22.04, 21.88]} diff --git a/pytorch/output_test2/altra_10_10_as-caida_100000.json b/pytorch/output_test2/altra_10_10_as-caida_100000.json deleted file mode 100644 index 76f9282..0000000 --- a/pytorch/output_test2/altra_10_10_as-caida_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 100000, "MATRIX_FILE": "as-caida", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 106762, "MATRIX_DENSITY": 0.00010842726485909405, "TIME_S": 7.69922399520874, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [21.24, 21.28, 21.28, 21.12, 21.28, 21.0, 21.0, 21.16, 21.16, 21.08], "POWER": [101.04, 100.48, 90.04, 74.8, 62.16, 59.8, 62.6, 62.6, 77.84, 92.96, 103.24, 101.24, 101.84], "JOULES": 631.4889716720581, "POWER_AFTER": [21.24, 21.44, 21.36, 21.4, 21.4, 21.36, 21.44, 21.36, 21.28, 21.36]} diff --git a/pytorch/output_test2/altra_10_10_dc2_100000.json b/pytorch/output_test2/altra_10_10_dc2_100000.json deleted file mode 100644 index f164ec6..0000000 --- a/pytorch/output_test2/altra_10_10_dc2_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 100000, "MATRIX_FILE": "dc2", "MATRIX_SHAPE": [116835, 116835], "MATRIX_SIZE": 13650417225, "MATRIX_NNZ": 766396, "MATRIX_DENSITY": 5.614451099680581e-05, "TIME_S": 37.14217662811279, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [21.24, 20.92, 21.08, 21.16, 21.36, 21.6, 21.6, 21.28, 21.04, 21.0], "POWER": [102.16, 103.44, 92.84, 76.08, 59.2, 59.48, 59.48, 66.88, 79.48, 96.96, 106.32, 105.92, 102.72, 102.68, 100.88, 99.92, 99.32, 98.04, 97.4, 97.4, 97.08, 94.72, 94.08, 96.16, 94.52, 95.32, 94.76, 92.16, 92.76, 95.88, 96.48, 96.48, 97.4, 98.08, 97.92, 97.56, 98.44, 97.36, 97.88, 99.72, 99.52, 99.0, 97.76, 96.36, 96.36], "JOULES": 3585.0201398849495, "POWER_AFTER": [23.36, 23.08, 22.72, 22.52, 22.52, 22.12, 22.0, 21.96, 21.72, 21.72]} diff --git a/pytorch/output_test2/altra_10_10_de2010_100000.json b/pytorch/output_test2/altra_10_10_de2010_100000.json deleted file mode 100644 index 3d331a8..0000000 --- a/pytorch/output_test2/altra_10_10_de2010_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 100000, "MATRIX_FILE": "de2010", "MATRIX_SHAPE": [24115, 24115], "MATRIX_SIZE": 581533225, "MATRIX_NNZ": 116056, "MATRIX_DENSITY": 0.0001995689928120616, "TIME_S": 8.169610738754272, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [21.48, 21.44, 21.32, 21.44, 21.6, 21.64, 21.44, 21.44, 21.4, 21.24], "POWER": [104.88, 103.2, 87.28, 72.12, 56.64, 56.64, 57.16, 62.64, 82.24, 97.56, 105.76, 102.64, 99.36], "JOULES": 638.1325230026245, "POWER_AFTER": [21.24, 21.32, 21.32, 21.24, 20.84, 20.84, 20.84, 20.84, 21.16, 21.56]} diff --git a/pytorch/output_test2/altra_10_10_de2010_100000.output b/pytorch/output_test2/altra_10_10_de2010_100000.output deleted file mode 100644 index bbe38c5..0000000 --- a/pytorch/output_test2/altra_10_10_de2010_100000.output +++ /dev/null @@ -1,25 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3471783 queued and waiting for resources -srun: job 3471783 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 13, 21, ..., 116047, 116051, - 116056]), - col_indices=tensor([ 250, 251, 757, ..., 23334, 23553, 24050]), - values=tensor([ 14900., 33341., 20255., ..., 164227., 52413., - 16949.]), size=(24115, 24115), nnz=116056, - layout=torch.sparse_csr) -tensor([0.3069, 0.2208, 0.9592, ..., 0.2726, 0.0490, 0.9363]) -Matrix: de2010 -Shape: torch.Size([24115, 24115]) -Size: 581533225 -NNZ: 116056 -Density: 0.0001995689928120616 -Time: 8.169610738754272 seconds - diff --git a/pytorch/output_test2/altra_10_10_email-Enron_100000.json b/pytorch/output_test2/altra_10_10_email-Enron_100000.json deleted file mode 100644 index 7897789..0000000 --- a/pytorch/output_test2/altra_10_10_email-Enron_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 100000, "MATRIX_FILE": "email-Enron", "MATRIX_SHAPE": [36692, 36692], "MATRIX_SIZE": 1346302864, "MATRIX_NNZ": 367662, "MATRIX_DENSITY": 0.0002730901120626302, "TIME_S": 12.88691234588623, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [20.76, 20.88, 20.84, 21.0, 21.16, 21.24, 21.44, 21.44, 21.48, 21.36], "POWER": [99.08, 99.68, 91.12, 77.0, 63.04, 57.0, 58.84, 75.68, 90.8, 90.8, 104.72, 103.32, 100.28, 97.32, 95.44, 93.56, 93.36], "JOULES": 1113.6021366119385, "POWER_AFTER": [21.72, 21.72, 21.8, 21.72, 21.92, 21.88, 21.88, 21.76, 21.44, 21.28]} diff --git a/pytorch/output_test2/altra_10_10_fl2010_100000.json b/pytorch/output_test2/altra_10_10_fl2010_100000.json deleted file mode 100644 index 92ea71d..0000000 --- a/pytorch/output_test2/altra_10_10_fl2010_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 100000, "MATRIX_FILE": "fl2010", "MATRIX_SHAPE": [484481, 484481], "MATRIX_SIZE": 234721839361, "MATRIX_NNZ": 2346294, "MATRIX_DENSITY": 9.99606174861054e-06, "TIME_S": 31.069382905960083, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [21.64, 21.72, 21.64, 21.68, 21.4, 21.44, 21.24, 21.12, 21.12, 21.04], "POWER": [120.04, 120.92, 98.92, 79.32, 58.8, 57.84, 62.68, 62.68, 77.64, 100.44, 114.44, 116.72, 118.8, 121.36, 118.48, 116.4, 114.24, 113.88, 109.72, 109.72, 117.92, 119.64, 115.56, 112.28, 107.04, 104.52, 105.32, 109.56, 109.76, 110.6, 113.36, 113.36, 116.64], "JOULES": 3220.0128221511845, "POWER_AFTER": [22.36, 22.24, 22.52, 22.4, 22.44, 22.4, 22.28, 23.56, 23.56, 25.52]} diff --git a/pytorch/output_test2/altra_10_10_fl2010_100000.output b/pytorch/output_test2/altra_10_10_fl2010_100000.output deleted file mode 100644 index 163aa12..0000000 --- a/pytorch/output_test2/altra_10_10_fl2010_100000.output +++ /dev/null @@ -1,25 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3471829 queued and waiting for resources -srun: job 3471829 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 2, 5, ..., 2346288, - 2346292, 2346294]), - col_indices=tensor([ 1513, 5311, 947, ..., 484460, 482463, - 484022]), - values=tensor([28364., 12497., 11567., ..., 8532., 22622., 35914.]), - size=(484481, 484481), nnz=2346294, layout=torch.sparse_csr) -tensor([0.5157, 0.5811, 0.2529, ..., 0.9249, 0.1469, 0.4136]) -Matrix: fl2010 -Shape: torch.Size([484481, 484481]) -Size: 234721839361 -NNZ: 2346294 -Density: 9.99606174861054e-06 -Time: 31.069382905960083 seconds - diff --git a/pytorch/output_test2/altra_10_10_ga2010_100000.json b/pytorch/output_test2/altra_10_10_ga2010_100000.json deleted file mode 100644 index d45c8be..0000000 --- a/pytorch/output_test2/altra_10_10_ga2010_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 100000, "MATRIX_FILE": "ga2010", "MATRIX_SHAPE": [291086, 291086], "MATRIX_SIZE": 84731059396, "MATRIX_NNZ": 1418056, "MATRIX_DENSITY": 1.6735964475229304e-05, "TIME_S": 17.813313722610474, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [21.56, 21.72, 21.68, 21.56, 21.56, 21.76, 21.2, 21.16, 20.96, 20.96], "POWER": [115.36, 114.88, 106.84, 89.28, 71.36, 58.8, 57.92, 74.0, 91.48, 110.52, 113.56, 113.56, 116.72, 113.88, 117.16, 119.4, 113.4, 113.76, 111.48, 110.64, 115.04, 121.64], "JOULES": 1821.6114812183382, "POWER_AFTER": [21.56, 21.56, 21.4, 21.28, 21.36, 21.48, 21.64, 21.84, 21.68, 21.88]} diff --git a/pytorch/output_test2/altra_10_10_mac_econ_fwd500_100000.json b/pytorch/output_test2/altra_10_10_mac_econ_fwd500_100000.json deleted file mode 100644 index 2fa64f9..0000000 --- a/pytorch/output_test2/altra_10_10_mac_econ_fwd500_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 100000, "MATRIX_FILE": "mac_econ_fwd500", "MATRIX_SHAPE": [206500, 206500], "MATRIX_SIZE": 42642250000, "MATRIX_NNZ": 1273389, "MATRIX_DENSITY": 2.9862143765866013e-05, "TIME_S": 13.249896049499512, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [21.04, 21.08, 21.12, 21.24, 21.32, 21.32, 21.2, 21.04, 21.08, 21.08], "POWER": [109.92, 114.24, 106.48, 89.76, 70.72, 60.84, 56.4, 74.08, 74.08, 94.04, 110.84, 111.28, 109.44, 109.48, 110.12, 107.64, 114.36, 119.48, 121.56, 121.4], "JOULES": 1343.1373804092407, "POWER_AFTER": [21.44, 21.4, 21.24, 21.28, 21.92, 22.56, 23.52, 24.28, 24.28, 24.4]} diff --git a/pytorch/output_test2/altra_10_10_mac_econ_fwd500_100000.output b/pytorch/output_test2/altra_10_10_mac_econ_fwd500_100000.output deleted file mode 100644 index c038e87..0000000 --- a/pytorch/output_test2/altra_10_10_mac_econ_fwd500_100000.output +++ /dev/null @@ -1,26 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3471828 queued and waiting for resources -srun: job 3471828 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 8, ..., 1273376, - 1273379, 1273389]), - col_indices=tensor([ 3, 30, 44, ..., 206363, 206408, - 206459]), - values=tensor([-3.7877e-03, -1.5420e-01, 9.5305e-04, ..., - 1.2290e-01, 2.2235e-01, -1.0000e+00]), - size=(206500, 206500), nnz=1273389, layout=torch.sparse_csr) -tensor([0.7751, 0.0281, 0.9910, ..., 0.3020, 0.8213, 0.1857]) -Matrix: mac_econ_fwd500 -Shape: torch.Size([206500, 206500]) -Size: 42642250000 -NNZ: 1273389 -Density: 2.9862143765866013e-05 -Time: 13.249896049499512 seconds - diff --git a/pytorch/output_test2/altra_10_10_mc2depi_100000.json b/pytorch/output_test2/altra_10_10_mc2depi_100000.json deleted file mode 100644 index 25bae52..0000000 --- a/pytorch/output_test2/altra_10_10_mc2depi_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 100000, "MATRIX_FILE": "mc2depi", "MATRIX_SHAPE": [525825, 525825], "MATRIX_SIZE": 276491930625, "MATRIX_NNZ": 2100225, "MATRIX_DENSITY": 7.595972132902821e-06, "TIME_S": 19.404656887054443, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [21.72, 21.48, 21.32, 21.52, 21.52, 21.28, 21.4, 21.56, 21.56, 21.4], "POWER": [116.0, 110.56, 110.56, 91.76, 72.12, 61.36, 61.52, 74.2, 90.92, 117.8, 125.96, 128.68, 130.72, 124.72, 124.72, 115.08, 119.88, 117.36, 115.44, 110.44, 111.24, 110.4], "JOULES": 2049.1541203308107, "POWER_AFTER": [21.6, 21.64, 21.64, 21.84, 21.84, 21.76, 21.76, 21.72, 21.72, 21.4]} diff --git a/pytorch/output_test2/altra_10_10_p2p-Gnutella04_100000.json b/pytorch/output_test2/altra_10_10_p2p-Gnutella04_100000.json deleted file mode 100644 index e13712a..0000000 --- a/pytorch/output_test2/altra_10_10_p2p-Gnutella04_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 100000, "MATRIX_FILE": "p2p-Gnutella04", "MATRIX_SHAPE": [10879, 10879], "MATRIX_SIZE": 118352641, "MATRIX_NNZ": 39994, "MATRIX_DENSITY": 0.0003379223282393842, "TIME_S": 7.197759389877319, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [21.12, 21.2, 21.04, 20.92, 21.0, 21.04, 21.04, 20.72, 20.76, 21.12], "POWER": [100.84, 100.04, 85.4, 71.68, 71.68, 56.24, 57.76, 66.84, 79.88, 94.24, 101.2, 100.24, 98.4, 96.36, 95.08], "JOULES": 655.9629627895355, "POWER_AFTER": [21.68, 21.04, 20.88, 21.4, 21.28, 21.28, 21.32, 21.04, 21.04, 21.04]} diff --git a/pytorch/output_test2/altra_10_10_p2p-Gnutella04_100000.output b/pytorch/output_test2/altra_10_10_p2p-Gnutella04_100000.output deleted file mode 100644 index d5225b0..0000000 --- a/pytorch/output_test2/altra_10_10_p2p-Gnutella04_100000.output +++ /dev/null @@ -1,23 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3471826 queued and waiting for resources -srun: job 3471826 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 10, 20, ..., 39994, 39994, 39994]), - col_indices=tensor([ 1, 2, 3, ..., 9711, 10875, 10876]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(10879, 10879), - nnz=39994, layout=torch.sparse_csr) -tensor([0.2810, 0.9768, 0.5232, ..., 0.2583, 0.8876, 0.2861]) -Matrix: p2p-Gnutella04 -Shape: torch.Size([10879, 10879]) -Size: 118352641 -NNZ: 39994 -Density: 0.0003379223282393842 -Time: 7.197759389877319 seconds - diff --git a/pytorch/output_test2/altra_10_10_p2p-Gnutella24_100000.json b/pytorch/output_test2/altra_10_10_p2p-Gnutella24_100000.json deleted file mode 100644 index ef0ec71..0000000 --- a/pytorch/output_test2/altra_10_10_p2p-Gnutella24_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 100000, "MATRIX_FILE": "p2p-Gnutella24", "MATRIX_SHAPE": [26518, 26518], "MATRIX_SIZE": 703204324, "MATRIX_NNZ": 65369, "MATRIX_DENSITY": 9.295875717624285e-05, "TIME_S": 8.68448281288147, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [21.84, 22.0, 21.92, 22.0, 21.92, 21.56, 21.52, 21.56, 21.48, 21.48], "POWER": [94.8, 95.36, 83.32, 69.76, 57.92, 59.48, 65.64, 65.64, 82.24, 99.92, 105.68, 103.0, 101.92, 99.32], "JOULES": 751.5028329753875, "POWER_AFTER": [24.12, 24.56, 24.56, 24.56, 24.28, 24.16, 24.24, 24.28, 24.28, 24.08]} diff --git a/pytorch/output_test2/altra_10_10_p2p-Gnutella25_100000.json b/pytorch/output_test2/altra_10_10_p2p-Gnutella25_100000.json deleted file mode 100644 index c6e61c4..0000000 --- a/pytorch/output_test2/altra_10_10_p2p-Gnutella25_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 100000, "MATRIX_FILE": "p2p-Gnutella25", "MATRIX_SHAPE": [22687, 22687], "MATRIX_SIZE": 514699969, "MATRIX_NNZ": 54705, "MATRIX_DENSITY": 0.00010628522108964806, "TIME_S": 8.185347080230713, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [21.32, 21.32, 21.2, 21.04, 20.96, 21.04, 21.16, 21.28, 21.24, 21.2], "POWER": [97.72, 94.2, 94.2, 82.44, 65.96, 58.0, 62.48, 71.32, 83.8, 97.0, 99.48, 98.56, 99.76], "JOULES": 655.0902247238158, "POWER_AFTER": [21.28, 21.28, 21.32, 21.48, 21.28, 21.48, 21.44, 21.12, 20.8, 20.76]} diff --git a/pytorch/output_test2/altra_10_10_p2p-Gnutella30_100000.json b/pytorch/output_test2/altra_10_10_p2p-Gnutella30_100000.json deleted file mode 100644 index 4f06737..0000000 --- a/pytorch/output_test2/altra_10_10_p2p-Gnutella30_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 100000, "MATRIX_FILE": "p2p-Gnutella30", "MATRIX_SHAPE": [36682, 36682], "MATRIX_SIZE": 1345569124, "MATRIX_NNZ": 88328, "MATRIX_DENSITY": 6.564359899804003e-05, "TIME_S": 9.74808645248413, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [21.08, 21.04, 21.16, 21.16, 21.24, 21.04, 21.08, 21.04, 21.0, 21.08], "POWER": [97.28, 94.08, 85.16, 71.12, 58.08, 59.6, 62.64, 79.28, 97.68, 97.68, 108.48, 106.6, 104.24, 101.56], "JOULES": 850.2556601142884, "POWER_AFTER": [20.92, 20.84, 20.96, 21.32, 21.28, 21.4, 21.44, 21.16, 21.16, 21.24]} diff --git a/pytorch/output_test2/altra_10_10_p2p-Gnutella30_100000.output b/pytorch/output_test2/altra_10_10_p2p-Gnutella30_100000.output deleted file mode 100644 index f8a4d57..0000000 --- a/pytorch/output_test2/altra_10_10_p2p-Gnutella30_100000.output +++ /dev/null @@ -1,23 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3471825 queued and waiting for resources -srun: job 3471825 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 10, 10, ..., 88328, 88328, 88328]), - col_indices=tensor([ 1, 2, 3, ..., 36675, 36676, 36677]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(36682, 36682), - nnz=88328, layout=torch.sparse_csr) -tensor([0.5280, 0.0933, 0.8124, ..., 0.0433, 0.2447, 0.2625]) -Matrix: p2p-Gnutella30 -Shape: torch.Size([36682, 36682]) -Size: 1345569124 -NNZ: 88328 -Density: 6.564359899804003e-05 -Time: 9.74808645248413 seconds - diff --git a/pytorch/output_test2/altra_10_10_ri2010_100000.json b/pytorch/output_test2/altra_10_10_ri2010_100000.json deleted file mode 100644 index 9ae93fc..0000000 --- a/pytorch/output_test2/altra_10_10_ri2010_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 100000, "MATRIX_FILE": "ri2010", "MATRIX_SHAPE": [25181, 25181], "MATRIX_SIZE": 634082761, "MATRIX_NNZ": 125750, "MATRIX_DENSITY": 0.00019831796057928155, "TIME_S": 7.650730133056641, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [17.16, 16.92, 16.6, 16.44, 16.52, 16.56, 16.84, 16.76, 16.84, 17.12], "POWER": [100.16, 98.52, 90.92, 76.44, 58.2, 51.52, 52.8, 65.68, 82.84, 97.64, 96.8, 94.84, 94.84], "JOULES": 603.8352458190918, "POWER_AFTER": [16.72, 16.96, 17.0, 16.92, 17.28, 17.12, 17.0, 17.04, 17.28, 17.2]} diff --git a/pytorch/output_test2/altra_10_10_ri2010_100000.output b/pytorch/output_test2/altra_10_10_ri2010_100000.output deleted file mode 100644 index 320f268..0000000 --- a/pytorch/output_test2/altra_10_10_ri2010_100000.output +++ /dev/null @@ -1,24 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3471779 queued and waiting for resources -srun: job 3471779 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 3, 8, ..., 125742, 125747, - 125750]), - col_indices=tensor([ 25, 56, 662, ..., 21738, 22279, 23882]), - values=tensor([17171., 37318., 5284., ..., 25993., 24918., 803.]), - size=(25181, 25181), nnz=125750, layout=torch.sparse_csr) -tensor([0.2875, 0.2982, 0.0876, ..., 0.4058, 0.8442, 0.7364]) -Matrix: ri2010 -Shape: torch.Size([25181, 25181]) -Size: 634082761 -NNZ: 125750 -Density: 0.00019831796057928155 -Time: 7.650730133056641 seconds - diff --git a/pytorch/output_test2/altra_10_10_rma10_100000.json b/pytorch/output_test2/altra_10_10_rma10_100000.json deleted file mode 100644 index 0de6415..0000000 --- a/pytorch/output_test2/altra_10_10_rma10_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 100000, "MATRIX_FILE": "rma10", "MATRIX_SHAPE": [46835, 46835], "MATRIX_SIZE": 2193517225, "MATRIX_NNZ": 2374001, "MATRIX_DENSITY": 0.0010822805369125833, "TIME_S": 18.274461030960083, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [22.08, 21.84, 22.0, 22.0, 21.84, 21.96, 21.92, 21.6, 21.68, 21.84], "POWER": [117.12, 115.04, 110.24, 95.24, 76.96, 64.8, 58.44, 69.48, 90.84, 109.8, 115.72, 121.2, 121.2, 120.8, 116.48, 112.48, 110.0, 110.48, 109.6, 107.6, 108.0, 110.32], "JOULES": 1849.3985409355162, "POWER_AFTER": [22.12, 22.0, 21.84, 21.76, 21.72, 21.64, 21.6, 21.56, 21.64, 21.64]} diff --git a/pytorch/output_test2/altra_10_10_rma10_100000.output b/pytorch/output_test2/altra_10_10_rma10_100000.output deleted file mode 100644 index dda985b..0000000 --- a/pytorch/output_test2/altra_10_10_rma10_100000.output +++ /dev/null @@ -1,25 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3471832 queued and waiting for resources -srun: job 3471832 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 17, 34, ..., 2373939, - 2373970, 2374001]), - col_indices=tensor([ 0, 1, 2, ..., 46831, 46833, 46834]), - values=tensor([ 1.2636e+05, -1.6615e+07, -8.2015e+04, ..., - 8.3378e+01, 2.5138e+00, 1.2184e+03]), - size=(46835, 46835), nnz=2374001, layout=torch.sparse_csr) -tensor([0.9389, 0.2472, 0.7378, ..., 0.8609, 0.3319, 0.1508]) -Matrix: rma10 -Shape: torch.Size([46835, 46835]) -Size: 2193517225 -NNZ: 2374001 -Density: 0.0010822805369125833 -Time: 18.274461030960083 seconds - diff --git a/pytorch/output_test2/altra_10_10_soc-sign-Slashdot090216_100000.json b/pytorch/output_test2/altra_10_10_soc-sign-Slashdot090216_100000.json deleted file mode 100644 index 2e64644..0000000 --- a/pytorch/output_test2/altra_10_10_soc-sign-Slashdot090216_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 100000, "MATRIX_FILE": "soc-sign-Slashdot090216", "MATRIX_SHAPE": [81871, 81871], "MATRIX_SIZE": 6702860641, "MATRIX_NNZ": 545671, "MATRIX_DENSITY": 8.140867447881048e-05, "TIME_S": 21.485024452209473, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [21.56, 21.36, 21.36, 21.72, 21.72, 21.88, 21.88, 22.0, 21.76, 21.36], "POWER": [101.32, 102.64, 100.4, 82.28, 66.8, 62.8, 62.88, 75.04, 91.52, 107.12, 104.28, 103.72, 102.2, 102.2, 103.2, 103.24, 107.08, 108.24, 106.2], "JOULES": 1738.469596824646, "POWER_AFTER": [21.72, 21.68, 21.52, 21.4, 21.56, 21.72, 21.84, 22.08, 22.08, 22.08]} diff --git a/pytorch/output_test2/altra_10_10_soc-sign-Slashdot090216_100000.output b/pytorch/output_test2/altra_10_10_soc-sign-Slashdot090216_100000.output deleted file mode 100644 index ba7cba7..0000000 --- a/pytorch/output_test2/altra_10_10_soc-sign-Slashdot090216_100000.output +++ /dev/null @@ -1,24 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3471855 queued and waiting for resources -srun: job 3471855 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 29, 124, ..., 545669, 545669, - 545671]), - col_indices=tensor([ 1, 2, 3, ..., 81869, 81699, 81863]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(81871, 81871), - nnz=545671, layout=torch.sparse_csr) -tensor([0.7599, 0.3131, 0.1356, ..., 0.5599, 0.7303, 0.7084]) -Matrix: soc-sign-Slashdot090216 -Shape: torch.Size([81871, 81871]) -Size: 6702860641 -NNZ: 545671 -Density: 8.140867447881048e-05 -Time: 21.485024452209473 seconds - diff --git a/pytorch/output_test2/altra_10_10_soc-sign-Slashdot090221_100000.json b/pytorch/output_test2/altra_10_10_soc-sign-Slashdot090221_100000.json deleted file mode 100644 index 644846e..0000000 --- a/pytorch/output_test2/altra_10_10_soc-sign-Slashdot090221_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 100000, "MATRIX_FILE": "soc-sign-Slashdot090221", "MATRIX_SHAPE": [82144, 82144], "MATRIX_SIZE": 6747636736, "MATRIX_NNZ": 549202, "MATRIX_DENSITY": 8.13917555860553e-05, "TIME_S": 9.906620264053345, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [21.48, 21.28, 21.6, 21.6, 21.88, 21.96, 21.96, 21.84, 21.4, 21.44], "POWER": [102.12, 100.8, 88.12, 72.96, 55.76, 56.68, 60.92, 60.92, 77.88, 98.52, 109.4, 109.76, 111.0, 109.04, 106.48, 104.8, 105.32, 102.52], "JOULES": 1025.1467094707486, "POWER_AFTER": [21.88, 21.84, 21.76, 21.24, 21.36, 21.4, 21.32, 21.48, 21.6, 21.48]} diff --git a/pytorch/output_test2/altra_10_10_soc-sign-Slashdot090221_100000.output b/pytorch/output_test2/altra_10_10_soc-sign-Slashdot090221_100000.output deleted file mode 100644 index bdcacf0..0000000 --- a/pytorch/output_test2/altra_10_10_soc-sign-Slashdot090221_100000.output +++ /dev/null @@ -1,24 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3471791 queued and waiting for resources -srun: job 3471791 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 29, 124, ..., 549200, 549200, - 549202]), - col_indices=tensor([ 1, 2, 3, ..., 82142, 81974, 82136]), - values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(82144, 82144), - nnz=549202, layout=torch.sparse_csr) -tensor([0.7291, 0.8277, 0.0975, ..., 0.0057, 0.6109, 0.6944]) -Matrix: soc-sign-Slashdot090221 -Shape: torch.Size([82144, 82144]) -Size: 6747636736 -NNZ: 549202 -Density: 8.13917555860553e-05 -Time: 9.906620264053345 seconds - diff --git a/pytorch/output_test2/altra_10_10_soc-sign-epinions_100000.json b/pytorch/output_test2/altra_10_10_soc-sign-epinions_100000.json deleted file mode 100644 index 7a4388d..0000000 --- a/pytorch/output_test2/altra_10_10_soc-sign-epinions_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 100000, "MATRIX_FILE": "soc-sign-epinions", "MATRIX_SHAPE": [131828, 131828], "MATRIX_SIZE": 17378621584, "MATRIX_NNZ": 841372, "MATRIX_DENSITY": 4.841419648464106e-05, "TIME_S": 31.47378420829773, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [17.16, 17.16, 17.12, 17.2, 17.16, 17.36, 17.52, 17.6, 17.48, 17.44], "POWER": [93.96, 93.96, 93.72, 75.92, 62.56, 53.68, 55.84, 68.12, 84.12, 103.88, 107.88, 106.84, 107.16, 107.16, 104.4, 100.96, 96.64, 96.08, 98.24, 100.16, 97.92, 97.44, 96.24, 94.64, 90.36, 92.96, 92.96, 92.96, 91.56, 90.88, 91.24, 93.72, 95.72], "JOULES": 2885.550624418259, "POWER_AFTER": [18.24, 18.04, 18.0, 18.0, 17.88, 17.8, 18.2, 18.28, 18.48, 18.52]} diff --git a/pytorch/output_test2/altra_10_10_soc-sign-epinions_100000.output b/pytorch/output_test2/altra_10_10_soc-sign-epinions_100000.output deleted file mode 100644 index 22be23c..0000000 --- a/pytorch/output_test2/altra_10_10_soc-sign-epinions_100000.output +++ /dev/null @@ -1,25 +0,0 @@ -srun: Job time limit was unset; set to partition default of 60 minutes -srun: ################################################################################ -srun: # Please note that the oasis compute nodes have aarch64 architecture CPUs. # -srun: # All submission nodes and all other compute nodes have x86_64 architecture # -srun: # CPUs. Programs, environments, or other software that was built on x86_64 # -srun: # nodes may need to be rebuilt to properly execute on these nodes. # -srun: ################################################################################ -srun: job 3471823 queued and waiting for resources -srun: job 3471823 has been allocated resources -/nfshomes/vut/ampere_research/pytorch/spmv.py:22: UserWarning: Sparse CSR tensor support is in beta state. If you miss a functionality in the sparse tensor support, please submit a feature request to https://github.com/pytorch/pytorch/issues. (Triggered internally at /space/jenkins/workspace/Releases/pytorch-dls/pytorch-dls/aten/src/ATen/SparseCsrTensorImpl.cpp:55.) - ).to_sparse_csr().type(torch.float) -tensor(crow_indices=tensor([ 0, 1, 2, ..., 841371, 841371, - 841372]), - col_indices=tensor([ 1, 128552, 3, ..., 131824, 131826, - 7714]), - values=tensor([-1., -1., 1., ..., 1., 1., 1.]), - size=(131828, 131828), nnz=841372, layout=torch.sparse_csr) -tensor([0.4186, 0.4768, 0.7650, ..., 0.7266, 0.5735, 0.6056]) -Matrix: soc-sign-epinions -Shape: torch.Size([131828, 131828]) -Size: 17378621584 -NNZ: 841372 -Density: 4.841419648464106e-05 -Time: 31.47378420829773 seconds - diff --git a/pytorch/output_test2/altra_10_10_sx-mathoverflow_100000.json b/pytorch/output_test2/altra_10_10_sx-mathoverflow_100000.json deleted file mode 100644 index 2f91ad3..0000000 --- a/pytorch/output_test2/altra_10_10_sx-mathoverflow_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 100000, "MATRIX_FILE": "sx-mathoverflow", "MATRIX_SHAPE": [24818, 24818], "MATRIX_SIZE": 615933124, "MATRIX_NNZ": 239978, "MATRIX_DENSITY": 0.00038961697406616504, "TIME_S": 9.512531042098999, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [21.16, 20.8, 20.84, 20.76, 21.0, 21.28, 21.36, 21.36, 21.56, 21.6], "POWER": [99.92, 100.08, 87.32, 72.88, 59.16, 50.88, 50.88, 54.48, 71.84, 89.72, 105.24, 106.76, 106.32, 104.48], "JOULES": 748.8292432785034, "POWER_AFTER": [21.2, 20.92, 20.92, 20.92, 21.2, 21.04, 21.08, 21.4, 21.08, 21.16]} diff --git a/pytorch/output_test2/altra_10_10_tn2010_100000.json b/pytorch/output_test2/altra_10_10_tn2010_100000.json deleted file mode 100644 index 740f027..0000000 --- a/pytorch/output_test2/altra_10_10_tn2010_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 100000, "MATRIX_FILE": "tn2010", "MATRIX_SHAPE": [240116, 240116], "MATRIX_SIZE": 57655693456, "MATRIX_NNZ": 1193966, "MATRIX_DENSITY": 2.070855328296721e-05, "TIME_S": 16.210495948791504, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [21.44, 21.16, 21.24, 21.36, 21.16, 21.48, 21.28, 21.16, 21.0, 20.84], "POWER": [107.88, 109.4, 94.08, 72.28, 58.64, 54.36, 59.12, 78.8, 78.8, 95.44, 110.56, 109.2, 109.28, 105.92, 108.24, 107.16, 106.4, 109.0, 111.52], "JOULES": 1480.7545082092288, "POWER_AFTER": [21.68, 21.64, 21.64, 21.6, 21.6, 21.52, 21.56, 21.76, 22.08, 22.48]} diff --git a/pytorch/output_test2/altra_10_10_ut2010_100000.json b/pytorch/output_test2/altra_10_10_ut2010_100000.json deleted file mode 100644 index 1ee8891..0000000 --- a/pytorch/output_test2/altra_10_10_ut2010_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 100000, "MATRIX_FILE": "ut2010", "MATRIX_SHAPE": [115406, 115406], "MATRIX_SIZE": 13318544836, "MATRIX_NNZ": 572066, "MATRIX_DENSITY": 4.295259032005559e-05, "TIME_S": 14.674797296524048, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [20.72, 21.0, 20.96, 20.92, 20.76, 20.68, 20.84, 20.64, 20.72, 20.96], "POWER": [105.44, 105.04, 90.04, 74.36, 58.48, 58.48, 56.4, 64.68, 79.0, 96.68, 106.04, 107.04, 105.64, 109.16, 108.88, 108.28, 106.32, 106.16, 106.16, 103.52], "JOULES": 1388.7750161361696, "POWER_AFTER": [21.24, 21.04, 21.08, 21.08, 21.12, 20.88, 20.88, 20.96, 20.88, 21.0]} diff --git a/pytorch/output_test2/altra_10_10_va2010_100000.json b/pytorch/output_test2/altra_10_10_va2010_100000.json deleted file mode 100644 index 90e8f6c..0000000 --- a/pytorch/output_test2/altra_10_10_va2010_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 100000, "MATRIX_FILE": "va2010", "MATRIX_SHAPE": [285762, 285762], "MATRIX_SIZE": 81659920644, "MATRIX_NNZ": 1402128, "MATRIX_DENSITY": 1.717033263003816e-05, "TIME_S": 21.11183762550354, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [21.8, 21.72, 21.76, 21.88, 21.84, 21.92, 21.96, 22.08, 22.08, 22.12], "POWER": [110.68, 110.72, 94.88, 76.76, 76.76, 61.96, 63.16, 70.68, 91.52, 111.84, 121.16, 120.44, 118.12, 116.68, 114.48, 116.84, 116.84, 113.72, 113.04, 110.4], "JOULES": 1932.6268738555905, "POWER_AFTER": [22.2, 22.44, 22.88, 22.52, 22.32, 22.32, 22.04, 21.76, 21.68, 21.92]} diff --git a/pytorch/output_test2/altra_10_10_vt2010_100000.json b/pytorch/output_test2/altra_10_10_vt2010_100000.json deleted file mode 100644 index eeec04a..0000000 --- a/pytorch/output_test2/altra_10_10_vt2010_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "ALTRA", "ITERATIONS": 100000, "MATRIX_FILE": "vt2010", "MATRIX_SHAPE": [32580, 32580], "MATRIX_SIZE": 1061456400, "MATRIX_NNZ": 155598, "MATRIX_DENSITY": 0.00014658915806621921, "TIME_S": 10.058021783828735, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [22.12, 21.92, 21.84, 21.48, 21.44, 21.6, 21.72, 21.48, 21.88, 22.0], "POWER": [107.8, 104.12, 91.0, 91.0, 73.2, 62.64, 62.2, 72.84, 88.0, 105.08, 109.68, 107.92, 107.12], "JOULES": 869.7752934837342, "POWER_AFTER": [21.96, 21.84, 22.04, 21.92, 21.96, 21.88, 21.96, 21.92, 21.84, 21.76]} diff --git a/pytorch/output_test2/epyc_7313p_10_10_ASIC_680k_100000.json b/pytorch/output_test2/epyc_7313p_10_10_ASIC_680k_100000.json deleted file mode 100644 index 4e4efee..0000000 --- a/pytorch/output_test2/epyc_7313p_10_10_ASIC_680k_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 100000, "MATRIX_FILE": "ASIC_680k", "MATRIX_SHAPE": [682862, 682862], "MATRIX_SIZE": 466300511044, "MATRIX_NNZ": 3871773, "MATRIX_DENSITY": 8.303171256088674e-06, "TIME_S": 77.6055359840393, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [41.35, 40.33, 40.46, 40.22, 39.91, 39.8, 40.53, 39.93, 40.0, 39.65], "POWER": [139.19], "JOULES": 10801.914553618431, "POWER_AFTER": [42.62, 39.91, 41.87, 45.85, 40.24, 40.39, 40.2, 39.74, 40.32, 39.74]} diff --git a/pytorch/output_test2/epyc_7313p_10_10_Oregon-2_100000.json b/pytorch/output_test2/epyc_7313p_10_10_Oregon-2_100000.json deleted file mode 100644 index 9b1a164..0000000 --- a/pytorch/output_test2/epyc_7313p_10_10_Oregon-2_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 100000, "MATRIX_FILE": "Oregon-2", "MATRIX_SHAPE": [11806, 11806], "MATRIX_SIZE": 139381636, "MATRIX_NNZ": 65460, "MATRIX_DENSITY": 0.0004696458003979807, "TIME_S": 4.933578252792358, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [41.99, 39.07, 45.1, 40.09, 40.19, 39.0, 39.86, 40.01, 39.88, 38.97], "POWER": [95.59], "JOULES": 471.60074518442156, "POWER_AFTER": [42.25, 39.42, 39.09, 38.85, 39.96, 39.01, 40.48, 38.81, 39.77, 39.06]} diff --git a/pytorch/output_test2/epyc_7313p_10_10_as-caida_100000.json b/pytorch/output_test2/epyc_7313p_10_10_as-caida_100000.json deleted file mode 100644 index 5a08e97..0000000 --- a/pytorch/output_test2/epyc_7313p_10_10_as-caida_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 100000, "MATRIX_FILE": "as-caida", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 106762, "MATRIX_DENSITY": 0.00010842726485909405, "TIME_S": 7.285882234573364, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [40.52, 39.15, 38.66, 39.53, 38.32, 39.24, 38.32, 39.16, 38.39, 39.11], "POWER": [99.77], "JOULES": 726.9124705433845, "POWER_AFTER": [40.76, 39.27, 38.94, 41.07, 38.55, 38.52, 38.78, 39.4, 38.59, 39.27]} diff --git a/pytorch/output_test2/epyc_7313p_10_10_dc2_100000.json b/pytorch/output_test2/epyc_7313p_10_10_dc2_100000.json deleted file mode 100644 index 56bc384..0000000 --- a/pytorch/output_test2/epyc_7313p_10_10_dc2_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 100000, "MATRIX_FILE": "dc2", "MATRIX_SHAPE": [116835, 116835], "MATRIX_SIZE": 13650417225, "MATRIX_NNZ": 766396, "MATRIX_DENSITY": 5.614451099680581e-05, "TIME_S": 20.959667444229126, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [41.86, 41.11, 41.33, 45.34, 39.89, 39.64, 39.81, 39.46, 39.95, 39.44], "POWER": [130.0], "JOULES": 2724.7567677497864, "POWER_AFTER": [41.62, 39.66, 40.09, 40.08, 39.79, 40.13, 39.9, 39.48, 39.76, 39.46]} diff --git a/pytorch/output_test2/epyc_7313p_10_10_de2010_100000.json b/pytorch/output_test2/epyc_7313p_10_10_de2010_100000.json deleted file mode 100644 index b6adaa9..0000000 --- a/pytorch/output_test2/epyc_7313p_10_10_de2010_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 100000, "MATRIX_FILE": "de2010", "MATRIX_SHAPE": [24115, 24115], "MATRIX_SIZE": 581533225, "MATRIX_NNZ": 116056, "MATRIX_DENSITY": 0.0001995689928120616, "TIME_S": 5.716832399368286, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [42.3, 38.99, 39.9, 38.72, 39.6, 38.7, 39.72, 39.07, 40.88, 38.68], "POWER": [102.08], "JOULES": 583.5742513275146, "POWER_AFTER": [41.13, 39.37, 39.29, 39.48, 39.23, 39.31, 38.94, 48.65, 45.1, 39.44]} diff --git a/pytorch/output_test2/epyc_7313p_10_10_email-Enron_100000.json b/pytorch/output_test2/epyc_7313p_10_10_email-Enron_100000.json deleted file mode 100644 index d492d5b..0000000 --- a/pytorch/output_test2/epyc_7313p_10_10_email-Enron_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 100000, "MATRIX_FILE": "email-Enron", "MATRIX_SHAPE": [36692, 36692], "MATRIX_SIZE": 1346302864, "MATRIX_NNZ": 367662, "MATRIX_DENSITY": 0.0002730901120626302, "TIME_S": 12.811992168426514, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [40.69, 39.52, 38.51, 39.32, 39.3, 39.55, 38.71, 39.44, 38.55, 39.0], "POWER": [111.47], "JOULES": 1428.1527670145035, "POWER_AFTER": [40.08, 39.68, 38.83, 39.62, 38.73, 39.72, 38.93, 39.63, 38.72, 39.59]} diff --git a/pytorch/output_test2/epyc_7313p_10_10_fl2010_100000.json b/pytorch/output_test2/epyc_7313p_10_10_fl2010_100000.json deleted file mode 100644 index 9ce2194..0000000 --- a/pytorch/output_test2/epyc_7313p_10_10_fl2010_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 100000, "MATRIX_FILE": "fl2010", "MATRIX_SHAPE": [484481, 484481], "MATRIX_SIZE": 234721839361, "MATRIX_NNZ": 2346294, "MATRIX_DENSITY": 9.99606174861054e-06, "TIME_S": 38.57364296913147, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [43.41, 39.71, 40.64, 39.57, 40.01, 39.59, 40.67, 39.73, 41.95, 40.44], "POWER": [146.49], "JOULES": 5650.652958548069, "POWER_AFTER": [43.12, 40.88, 40.74, 46.15, 40.03, 39.97, 39.96, 40.9, 40.37, 40.98]} diff --git a/pytorch/output_test2/epyc_7313p_10_10_ga2010_100000.json b/pytorch/output_test2/epyc_7313p_10_10_ga2010_100000.json deleted file mode 100644 index 40e95da..0000000 --- a/pytorch/output_test2/epyc_7313p_10_10_ga2010_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 100000, "MATRIX_FILE": "ga2010", "MATRIX_SHAPE": [291086, 291086], "MATRIX_SIZE": 84731059396, "MATRIX_NNZ": 1418056, "MATRIX_DENSITY": 1.6735964475229304e-05, "TIME_S": 33.085010051727295, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [68.04, 68.74, 71.06, 68.14, 65.41, 69.01, 74.03, 67.13, 66.27, 68.83], "POWER": [151.87], "JOULES": 5024.620476555824, "POWER_AFTER": [42.56, 40.18, 39.86, 39.64, 39.71, 40.05, 40.15, 40.33, 39.91, 39.93]} diff --git a/pytorch/output_test2/epyc_7313p_10_10_mac_econ_fwd500_100000.json b/pytorch/output_test2/epyc_7313p_10_10_mac_econ_fwd500_100000.json deleted file mode 100644 index a0e52b3..0000000 --- a/pytorch/output_test2/epyc_7313p_10_10_mac_econ_fwd500_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 100000, "MATRIX_FILE": "mac_econ_fwd500", "MATRIX_SHAPE": [206500, 206500], "MATRIX_SIZE": 42642250000, "MATRIX_NNZ": 1273389, "MATRIX_DENSITY": 2.9862143765866013e-05, "TIME_S": 10.857311248779297, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [41.53, 40.0, 40.81, 39.62, 40.1, 39.65, 39.53, 39.92, 39.46, 39.53], "POWER": [155.03], "JOULES": 1683.2089628982544, "POWER_AFTER": [41.01, 40.13, 40.25, 39.99, 39.62, 40.85, 40.86, 45.62, 40.19, 40.22]} diff --git a/pytorch/output_test2/epyc_7313p_10_10_mc2depi_100000.json b/pytorch/output_test2/epyc_7313p_10_10_mc2depi_100000.json deleted file mode 100644 index c57d4b5..0000000 --- a/pytorch/output_test2/epyc_7313p_10_10_mc2depi_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 100000, "MATRIX_FILE": "mc2depi", "MATRIX_SHAPE": [525825, 525825], "MATRIX_SIZE": 276491930625, "MATRIX_NNZ": 2100225, "MATRIX_DENSITY": 7.595972132902821e-06, "TIME_S": 14.108525037765503, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [50.67, 65.78, 62.81, 66.06, 69.62, 62.45, 50.62, 61.89, 58.86, 60.32], "POWER": [159.06], "JOULES": 2244.101992506981, "POWER_AFTER": [41.84, 39.6, 40.57, 39.58, 40.13, 39.78, 40.32, 39.37, 41.46, 39.79]} diff --git a/pytorch/output_test2/epyc_7313p_10_10_p2p-Gnutella04_100000.json b/pytorch/output_test2/epyc_7313p_10_10_p2p-Gnutella04_100000.json deleted file mode 100644 index 67f66d8..0000000 --- a/pytorch/output_test2/epyc_7313p_10_10_p2p-Gnutella04_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 100000, "MATRIX_FILE": "p2p-Gnutella04", "MATRIX_SHAPE": [10879, 10879], "MATRIX_SIZE": 118352641, "MATRIX_NNZ": 39994, "MATRIX_DENSITY": 0.0003379223282393842, "TIME_S": 3.682297468185425, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [40.77, 39.53, 40.23, 39.78, 39.57, 39.96, 39.93, 40.04, 40.05, 39.92], "POWER": [96.71], "JOULES": 356.11498814821243, "POWER_AFTER": [40.97, 39.73, 39.97, 39.94, 39.55, 39.74, 39.76, 40.24, 39.22, 39.1]} diff --git a/pytorch/output_test2/epyc_7313p_10_10_p2p-Gnutella24_100000.json b/pytorch/output_test2/epyc_7313p_10_10_p2p-Gnutella24_100000.json deleted file mode 100644 index 121d587..0000000 --- a/pytorch/output_test2/epyc_7313p_10_10_p2p-Gnutella24_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 100000, "MATRIX_FILE": "p2p-Gnutella24", "MATRIX_SHAPE": [26518, 26518], "MATRIX_SIZE": 703204324, "MATRIX_NNZ": 65369, "MATRIX_DENSITY": 9.295875717624285e-05, "TIME_S": 5.905890703201294, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [40.57, 39.41, 39.78, 39.28, 45.08, 41.07, 39.52, 39.33, 39.72, 39.39], "POWER": [100.57], "JOULES": 593.9554280209541, "POWER_AFTER": [41.15, 39.21, 39.39, 39.22, 39.59, 39.76, 39.22, 39.3, 39.72, 39.16]} diff --git a/pytorch/output_test2/epyc_7313p_10_10_p2p-Gnutella25_100000.json b/pytorch/output_test2/epyc_7313p_10_10_p2p-Gnutella25_100000.json deleted file mode 100644 index 895385c..0000000 --- a/pytorch/output_test2/epyc_7313p_10_10_p2p-Gnutella25_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 100000, "MATRIX_FILE": "p2p-Gnutella25", "MATRIX_SHAPE": [22687, 22687], "MATRIX_SIZE": 514699969, "MATRIX_NNZ": 54705, "MATRIX_DENSITY": 0.00010628522108964806, "TIME_S": 5.152007102966309, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [42.83, 40.24, 39.56, 40.23, 39.39, 40.63, 39.47, 40.26, 40.06, 41.04], "POWER": [97.57], "JOULES": 502.6813330364227, "POWER_AFTER": [42.03, 40.41, 39.25, 39.24, 39.63, 39.78, 40.31, 39.31, 39.76, 39.21]} diff --git a/pytorch/output_test2/epyc_7313p_10_10_p2p-Gnutella30_100000.json b/pytorch/output_test2/epyc_7313p_10_10_p2p-Gnutella30_100000.json deleted file mode 100644 index 2c738d3..0000000 --- a/pytorch/output_test2/epyc_7313p_10_10_p2p-Gnutella30_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 100000, "MATRIX_FILE": "p2p-Gnutella30", "MATRIX_SHAPE": [36682, 36682], "MATRIX_SIZE": 1345569124, "MATRIX_NNZ": 88328, "MATRIX_DENSITY": 6.564359899804003e-05, "TIME_S": 5.514855861663818, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [41.78, 39.68, 40.37, 39.72, 40.3, 45.14, 41.88, 39.66, 39.77, 39.66], "POWER": [102.52], "JOULES": 565.3830229377746, "POWER_AFTER": [41.48, 40.47, 41.27, 39.44, 39.67, 40.54, 39.51, 40.24, 39.31, 41.6]} diff --git a/pytorch/output_test2/epyc_7313p_10_10_ri2010_100000.json b/pytorch/output_test2/epyc_7313p_10_10_ri2010_100000.json deleted file mode 100644 index 2f4b486..0000000 --- a/pytorch/output_test2/epyc_7313p_10_10_ri2010_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 100000, "MATRIX_FILE": "ri2010", "MATRIX_SHAPE": [25181, 25181], "MATRIX_SIZE": 634082761, "MATRIX_NNZ": 125750, "MATRIX_DENSITY": 0.00019831796057928155, "TIME_S": 6.2883217334747314, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [40.5, 38.99, 38.82, 39.05, 38.88, 38.78, 39.03, 39.21, 38.98, 39.37], "POWER": [104.04], "JOULES": 654.2369931507111, "POWER_AFTER": [40.71, 38.8, 39.41, 38.92, 39.31, 38.76, 38.78, 40.12, 39.21, 46.09]} diff --git a/pytorch/output_test2/epyc_7313p_10_10_rma10_100000.json b/pytorch/output_test2/epyc_7313p_10_10_rma10_100000.json deleted file mode 100644 index 4014dcb..0000000 --- a/pytorch/output_test2/epyc_7313p_10_10_rma10_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 100000, "MATRIX_FILE": "rma10", "MATRIX_SHAPE": [46835, 46835], "MATRIX_SIZE": 2193517225, "MATRIX_NNZ": 2374001, "MATRIX_DENSITY": 0.0010822805369125833, "TIME_S": 38.7850341796875, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [40.96, 39.06, 40.4, 39.15, 40.82, 44.54, 39.97, 39.79, 39.76, 39.05], "POWER": [136.56], "JOULES": 5296.484267578125, "POWER_AFTER": [42.19, 40.68, 40.19, 40.6, 39.77, 40.74, 39.92, 40.55, 39.61, 40.63]} diff --git a/pytorch/output_test2/epyc_7313p_10_10_soc-sign-Slashdot090216_100000.json b/pytorch/output_test2/epyc_7313p_10_10_soc-sign-Slashdot090216_100000.json deleted file mode 100644 index ae5d036..0000000 --- a/pytorch/output_test2/epyc_7313p_10_10_soc-sign-Slashdot090216_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 100000, "MATRIX_FILE": "soc-sign-Slashdot090216", "MATRIX_SHAPE": [81871, 81871], "MATRIX_SIZE": 6702860641, "MATRIX_NNZ": 545671, "MATRIX_DENSITY": 8.140867447881048e-05, "TIME_S": 16.565748691558838, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [42.54, 39.13, 39.85, 38.79, 40.21, 38.98, 39.52, 38.92, 39.48, 39.32], "POWER": [122.93], "JOULES": 2036.427486653328, "POWER_AFTER": [45.86, 39.56, 39.69, 40.31, 39.53, 40.55, 39.15, 39.99, 39.13, 39.67]} diff --git a/pytorch/output_test2/epyc_7313p_10_10_soc-sign-Slashdot090221_100000.json b/pytorch/output_test2/epyc_7313p_10_10_soc-sign-Slashdot090221_100000.json deleted file mode 100644 index e12f006..0000000 --- a/pytorch/output_test2/epyc_7313p_10_10_soc-sign-Slashdot090221_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 100000, "MATRIX_FILE": "soc-sign-Slashdot090221", "MATRIX_SHAPE": [82144, 82144], "MATRIX_SIZE": 6747636736, "MATRIX_NNZ": 549202, "MATRIX_DENSITY": 8.13917555860553e-05, "TIME_S": 16.887201070785522, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [39.82, 40.65, 38.66, 38.38, 38.81, 38.83, 38.56, 38.43, 38.56, 38.4], "POWER": [121.66], "JOULES": 2054.4968822717665, "POWER_AFTER": [40.16, 38.88, 38.98, 39.3, 38.76, 38.72, 39.42, 38.78, 39.25, 38.73]} diff --git a/pytorch/output_test2/epyc_7313p_10_10_soc-sign-epinions_100000.json b/pytorch/output_test2/epyc_7313p_10_10_soc-sign-epinions_100000.json deleted file mode 100644 index efc26c5..0000000 --- a/pytorch/output_test2/epyc_7313p_10_10_soc-sign-epinions_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 100000, "MATRIX_FILE": "soc-sign-epinions", "MATRIX_SHAPE": [131828, 131828], "MATRIX_SIZE": 17378621584, "MATRIX_NNZ": 841372, "MATRIX_DENSITY": 4.841419648464106e-05, "TIME_S": 28.74003553390503, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [40.92, 39.62, 39.81, 39.63, 39.69, 39.93, 40.13, 39.49, 40.22, 39.73], "POWER": [128.06], "JOULES": 3680.448950471878, "POWER_AFTER": [42.14, 40.44, 40.38, 40.33, 40.04, 39.73, 40.03, 39.89, 39.97, 39.65]} diff --git a/pytorch/output_test2/epyc_7313p_10_10_sx-mathoverflow_100000.json b/pytorch/output_test2/epyc_7313p_10_10_sx-mathoverflow_100000.json deleted file mode 100644 index bcd157a..0000000 --- a/pytorch/output_test2/epyc_7313p_10_10_sx-mathoverflow_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 100000, "MATRIX_FILE": "sx-mathoverflow", "MATRIX_SHAPE": [24818, 24818], "MATRIX_SIZE": 615933124, "MATRIX_NNZ": 239978, "MATRIX_DENSITY": 0.00038961697406616504, "TIME_S": 9.806709051132202, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [41.51, 39.84, 40.47, 39.69, 39.5, 39.43, 40.48, 39.36, 40.49, 39.18], "POWER": [110.18], "JOULES": 1080.503203253746, "POWER_AFTER": [46.74, 40.47, 40.85, 39.36, 40.38, 40.43, 40.13, 39.54, 39.53, 39.11]} diff --git a/pytorch/output_test2/epyc_7313p_10_10_tn2010_100000.json b/pytorch/output_test2/epyc_7313p_10_10_tn2010_100000.json deleted file mode 100644 index 3881548..0000000 --- a/pytorch/output_test2/epyc_7313p_10_10_tn2010_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 100000, "MATRIX_FILE": "tn2010", "MATRIX_SHAPE": [240116, 240116], "MATRIX_SIZE": 57655693456, "MATRIX_NNZ": 1193966, "MATRIX_DENSITY": 2.070855328296721e-05, "TIME_S": 17.589671850204468, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [42.04, 40.09, 40.08, 39.99, 39.91, 39.54, 39.73, 39.62, 39.97, 40.07], "POWER": [151.86], "JOULES": 2671.167567172051, "POWER_AFTER": [40.91, 40.25, 40.2, 40.04, 40.2, 39.75, 40.04, 39.72, 39.49, 39.74]} diff --git a/pytorch/output_test2/epyc_7313p_10_10_ut2010_100000.json b/pytorch/output_test2/epyc_7313p_10_10_ut2010_100000.json deleted file mode 100644 index 39245b5..0000000 --- a/pytorch/output_test2/epyc_7313p_10_10_ut2010_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 100000, "MATRIX_FILE": "ut2010", "MATRIX_SHAPE": [115406, 115406], "MATRIX_SIZE": 13318544836, "MATRIX_NNZ": 572066, "MATRIX_DENSITY": 4.295259032005559e-05, "TIME_S": 7.522066831588745, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [41.26, 39.12, 39.52, 39.06, 39.73, 40.19, 40.7, 41.76, 41.47, 45.37], "POWER": [133.71], "JOULES": 1005.7755560517312, "POWER_AFTER": [40.48, 39.63, 40.0, 39.19, 39.77, 39.15, 39.44, 39.69, 39.23, 39.42]} diff --git a/pytorch/output_test2/epyc_7313p_10_10_va2010_100000.json b/pytorch/output_test2/epyc_7313p_10_10_va2010_100000.json deleted file mode 100644 index 1a8be31..0000000 --- a/pytorch/output_test2/epyc_7313p_10_10_va2010_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 100000, "MATRIX_FILE": "va2010", "MATRIX_SHAPE": [285762, 285762], "MATRIX_SIZE": 81659920644, "MATRIX_NNZ": 1402128, "MATRIX_DENSITY": 1.717033263003816e-05, "TIME_S": 24.332262754440308, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [40.82, 39.63, 39.96, 39.41, 39.7, 39.59, 39.95, 39.95, 39.67, 39.46], "POWER": [151.38], "JOULES": 3683.4179357671737, "POWER_AFTER": [42.41, 39.81, 39.8, 40.35, 39.68, 39.68, 40.54, 39.6, 39.96, 40.45]} diff --git a/pytorch/output_test2/epyc_7313p_10_10_vt2010_100000.json b/pytorch/output_test2/epyc_7313p_10_10_vt2010_100000.json deleted file mode 100644 index f761b92..0000000 --- a/pytorch/output_test2/epyc_7313p_10_10_vt2010_100000.json +++ /dev/null @@ -1 +0,0 @@ -{"CPU": "EPYC_7313P", "ITERATIONS": 100000, "MATRIX_FILE": "vt2010", "MATRIX_SHAPE": [32580, 32580], "MATRIX_SIZE": 1061456400, "MATRIX_NNZ": 155598, "MATRIX_DENSITY": 0.00014658915806621921, "TIME_S": 7.804270267486572, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "POWER_BEFORE": [40.49, 39.46, 39.18, 39.38, 39.2, 39.17, 39.84, 39.64, 39.67, 39.25], "POWER": [106.01], "JOULES": 827.3306910562516, "POWER_AFTER": [40.57, 39.15, 39.36, 39.09, 39.14, 39.45, 39.76, 39.02, 39.12, 44.6]}