174 lines
10 KiB
Plaintext
174 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 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%)
|
|
<not supported> 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
|
|
<not supported> LLC-loads:u
|
|
<not supported> 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%)
|
|
<not counted> 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
|
|
|
|
|
|
|