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