169 lines
10 KiB
Plaintext
169 lines
10 KiB
Plaintext
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 3394149 queued and waiting for resources
|
|
srun: job 3394149 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([0.4749, 0.3788, 0.8812, ..., 0.8281, 0.8889, 0.4945])
|
|
Shape: torch.Size([116835, 116835])
|
|
NNZ: 766396
|
|
Density: 5.614451099680581e-05
|
|
Time: 2.2480316162109375 seconds
|
|
|
|
Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/dc2.mtx 100':
|
|
|
|
50.43 msec task-clock:u # 0.009 CPUs utilized
|
|
0 context-switches:u # 0.000 /sec
|
|
0 cpu-migrations:u # 0.000 /sec
|
|
3,285 page-faults:u # 65.135 K/sec
|
|
54,118,679 cycles:u # 1.073 GHz (60.92%)
|
|
77,692,421 instructions:u # 1.44 insn per cycle (82.73%)
|
|
<not supported> branches:u
|
|
367,999 branch-misses:u
|
|
32,182,371 L1-dcache-loads:u # 638.112 M/sec
|
|
491,960 L1-dcache-load-misses:u # 1.53% of all L1-dcache accesses
|
|
<not supported> LLC-loads:u
|
|
<not supported> LLC-load-misses:u
|
|
30,682,258 L1-icache-loads:u # 608.367 M/sec
|
|
300,874 L1-icache-load-misses:u # 0.98% of all L1-icache accesses
|
|
55,244,523 dTLB-loads:u # 1.095 G/sec (19.09%)
|
|
<not counted> dTLB-load-misses:u (0.00%)
|
|
<not counted> iTLB-loads:u (0.00%)
|
|
<not counted> iTLB-load-misses:u (0.00%)
|
|
|
|
5.813837947 seconds time elapsed
|
|
|
|
28.815118000 seconds user
|
|
213.749674000 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([0.9715, 0.3920, 0.0297, ..., 0.1819, 0.5744, 0.8105])
|
|
Shape: torch.Size([116835, 116835])
|
|
NNZ: 766396
|
|
Density: 5.614451099680581e-05
|
|
Time: 2.2333595752716064 seconds
|
|
|
|
Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/dc2.mtx 100':
|
|
|
|
325,039 BR_MIS_PRED_RETIRED:u # 0.0 per branch branch_misprediction_ratio
|
|
19,383,216 BR_RETIRED:u
|
|
|
|
5.973132269 seconds time elapsed
|
|
|
|
29.719778000 seconds user
|
|
213.706315000 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([0.3371, 0.4985, 0.9905, ..., 0.6075, 0.1568, 0.3782])
|
|
Shape: torch.Size([116835, 116835])
|
|
NNZ: 766396
|
|
Density: 5.614451099680581e-05
|
|
Time: 1.9790923595428467 seconds
|
|
|
|
Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/dc2.mtx 100':
|
|
|
|
26,060,519 L1I_TLB:u # 0.0 per TLB access itlb_walk_ratio
|
|
4,749 ITLB_WALK:u
|
|
16,865 DTLB_WALK:u # 0.0 per TLB access dtlb_walk_ratio
|
|
34,819,729 L1D_TLB:u
|
|
|
|
5.575020445 seconds time elapsed
|
|
|
|
26.769391000 seconds user
|
|
188.138935000 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([0.6806, 0.8858, 0.7035, ..., 0.6007, 0.0880, 0.4550])
|
|
Shape: torch.Size([116835, 116835])
|
|
NNZ: 766396
|
|
Density: 5.614451099680581e-05
|
|
Time: 1.5306556224822998 seconds
|
|
|
|
Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/dc2.mtx 100':
|
|
|
|
30,777,115 L1I_CACHE:u # 0.0 per cache access l1i_cache_miss_ratio
|
|
293,980 L1I_CACHE_REFILL:u
|
|
461,522 L1D_CACHE_REFILL:u # 0.0 per cache access l1d_cache_miss_ratio
|
|
32,216,597 L1D_CACHE:u
|
|
|
|
4.961298684 seconds time elapsed
|
|
|
|
23.946357000 seconds user
|
|
156.598674000 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([0.3029, 0.1908, 0.9816, ..., 0.0418, 0.8182, 0.5474])
|
|
Shape: torch.Size([116835, 116835])
|
|
NNZ: 766396
|
|
Density: 5.614451099680581e-05
|
|
Time: 2.28926944732666 seconds
|
|
|
|
Performance counter stats for 'apptainer run pytorch-altra.sif -c numactl --cpunodebind=0 --membind=0 python spmv.py matrices/dc2.mtx 100':
|
|
|
|
567,700 LL_CACHE_MISS_RD:u # 1.0 per cache access ll_cache_read_miss_ratio
|
|
588,689 LL_CACHE_RD:u
|
|
189,417 L2D_TLB:u # 0.1 per TLB access l2_tlb_miss_ratio
|
|
22,360 L2D_TLB_REFILL:u
|
|
328,306 L2D_CACHE_REFILL:u # 0.2 per cache access l2_cache_miss_ratio
|
|
1,908,607 L2D_CACHE:u
|
|
|
|
5.710829283 seconds time elapsed
|
|
|
|
28.671301000 seconds user
|
|
213.960421000 seconds sys
|
|
|
|
|
|
|