as-caida results and batch.py update to remove distribution of workload

This commit is contained in:
cephi 2024-12-12 15:24:15 -05:00
parent b402503c31
commit 752ec8b9cd
739 changed files with 5682 additions and 27 deletions

View File

@ -16,7 +16,7 @@ parser.add_argument('base_iterations', type=int)
parser.add_argument('min_time_s', type=int) parser.add_argument('min_time_s', type=int)
parser.add_argument('baseline_time_s', type=int) parser.add_argument('baseline_time_s', type=int)
parser.add_argument('baseline_delay_s', type=int) parser.add_argument('baseline_delay_s', type=int)
parser.add_argument('--perf', action='store_const', const='--perf') #parser.add_argument('--perf', action='store_const', const='--perf')
parser.add_argument('--power', action='store_const', const='--power') parser.add_argument('--power', action='store_const', const='--power')
parser.add_argument('--distribute', action='store_true') parser.add_argument('--distribute', action='store_true')
args = parser.parse_args() args = parser.parse_args()
@ -32,7 +32,8 @@ srun_args = {
'--cpus-per-task', '160', '--cpus-per-task', '160',
#'--mem 28114', #'--mem 28114',
'--mem', '16G', '--mem', '16G',
'--ntasks-per-node', '1'#, '--ntasks-per-node', '1',
'--time', '1-00:00:00'
#'--exclusive', #'--exclusive',
#'--output', '/dev/null', #'--output', '/dev/null',
#'--error', '/dev/null' #'--error', '/dev/null'
@ -43,6 +44,7 @@ srun_args = {
'--qos', 'high', '--qos', 'high',
'--cpus-per-task', '16', '--cpus-per-task', '16',
'--ntasks-per-node', '1', '--ntasks-per-node', '1',
'--time', '1-00:00:00',
'--prefer', 'EPYC-7313P' '--prefer', 'EPYC-7313P'
], ],
Cpu.XEON_4216: [ Cpu.XEON_4216: [
@ -51,6 +53,7 @@ srun_args = {
'--qos', 'tron-exempt', '--qos', 'tron-exempt',
'--cpus-per-task', '32', '--cpus-per-task', '32',
'--ntasks-per-node', '1', '--ntasks-per-node', '1',
'--time', '1-00:00:00',
'--prefer', 'Xeon,4216' '--prefer', 'Xeon,4216'
] ]
} }
@ -60,7 +63,7 @@ python = {
Cpu.XEON_4216: 'python3.11' Cpu.XEON_4216: 'python3.11'
} }
def srun(srun_args_list: list, run_args, matrix_file: str) -> list: def run(run_args, matrix_file: str, srun_args_list: list = None) -> list:
run_args_list = [ run_args_list = [
args.cpu.name.lower(), args.cpu.name.lower(),
matrix_file, matrix_file,
@ -69,22 +72,29 @@ def srun(srun_args_list: list, run_args, matrix_file: str) -> list:
str(args.min_time_s), str(args.min_time_s),
str(args.baseline_time_s), str(args.baseline_time_s),
str(args.baseline_delay_s)] str(args.baseline_delay_s)]
if args.perf is not None: # if args.perf is not None:
run_args_list += [args.perf] # run_args_list += [args.perf]
if args.power is not None: if args.power is not None:
run_args_list += [args.power] run_args_list += [args.power]
return ['srun'] + srun_args_list + [python[args.cpu], 'run.py'] + run_args_list
if srun_args_list is None:
command = [python[args.cpu], 'run.py'] + run_args_list
else:
command = ['srun'] + srun_args_list + [python[args.cpu], 'run.py'] + run_args_list
print(command)
return command
processes = list() processes = list()
for i, matrix in enumerate(glob.glob(f'{args.matrix_dir.rstrip("/")}/*.mtx')): for i, matrix in enumerate(glob.glob(f'{args.matrix_dir.rstrip("/")}/*.mtx')):
if args.distribute: #if args.distribute:
if args.cpu == Cpu.ALTRA: # if args.cpu == Cpu.ALTRA:
i = i % 40 # i = i % 40
srun_args_temp = srun_args[args.cpu] + ['--nodelist', f'oasis{i:02}'] # srun_args_temp = srun_args[args.cpu] + ['--nodelist', f'oasis{i:02}']
elif args.cpu == Cpu.EPYC_7313P: # elif args.cpu == Cpu.EPYC_7313P:
srun_args_temp = srun_args[args.cpu] # srun_args_temp = srun_args[args.cpu]
else: #else:
srun_args_temp = srun_args[args.cpu] srun_args_temp = srun_args[args.cpu]
output_filename = '_'.join([ output_filename = '_'.join([
@ -97,15 +107,21 @@ for i, matrix in enumerate(glob.glob(f'{args.matrix_dir.rstrip("/")}/*.mtx')):
json_filepath = f'{args.output_dir.rstrip("/")}/{output_filename}.json' json_filepath = f'{args.output_dir.rstrip("/")}/{output_filename}.json'
raw_filepath = f'{args.output_dir.rstrip("/")}/{output_filename}.output' raw_filepath = f'{args.output_dir.rstrip("/")}/{output_filename}.output'
with open(json_filepath, 'w') as json_file, open(raw_filepath, 'w') as raw_file: with open(json_filepath, 'w') as json_file, open(raw_filepath, 'w') as raw_file:
print(srun(srun_args_temp, args, matrix))
print(json_filepath) print(json_filepath)
print(raw_filepath) print(raw_filepath)
if args.distribute:
processes.append(subprocess.Popen( processes.append(subprocess.Popen(
srun(srun_args_temp, args, matrix), run(args, matrix, srun_args_temp),
stdout=json_file, stdout=json_file,
stderr=raw_file)) stderr=raw_file))
else:
subprocess.run(
run(args, matrix),
stdout=json_file,
stderr=raw_file)
if args.distribute:
# Wait on every 10 jobs to avoid socket timeout. # Wait on every 10 jobs to avoid socket timeout.
if i % 10 == 9: if i % 10 == 9:
print("Waiting on 10 jobs") print("Waiting on 10 jobs")
@ -114,5 +130,6 @@ for i, matrix in enumerate(glob.glob(f'{args.matrix_dir.rstrip("/")}/*.mtx')):
processes = list() processes = list()
if args.distribute:
for process in processes: for process in processes:
process.wait() process.wait()

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 120692, "MATRIX_FILE": "as-caida", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 106762, "MATRIX_DENSITY": 0.00010842726485909405, "TIME_S": 12.845242738723755, "TIME_S_1KI": 0.10642994348195203, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 724.7658282470703, "W": 70.22348643601961, "J_1KI": 6.005085906663824, "W_1KI": 0.5818404404270342, "W_D": 51.50448643601961, "J_D": 531.5699015350341, "W_D_1KI": 0.4267431680311836, "J_D_1KI": 0.0035358032680805986}

View File

@ -0,0 +1,15 @@
/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([0.0273, 0.7256, 0.4946, ..., 0.3773, 0.8985, 0.8937])
Matrix: as-caida
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 106762
Density: 0.00010842726485909405
Time: 12.845242738723755 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 119741, "MATRIX_FILE": "as-caida_G_001", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 65910, "MATRIX_DENSITY": 6.693805873684353e-05, "TIME_S": 10.126415729522705, "TIME_S_1KI": 0.08456932654247672, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 934.9993497276305, "W": 78.36349396737387, "J_1KI": 7.808514625129492, "W_1KI": 0.6544416195569928, "W_D": 59.51949396737387, "J_D": 710.1608840818404, "W_D_1KI": 0.49706862283907655, "J_D_1KI": 0.004151198193092396}

View File

@ -0,0 +1,14 @@
/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, 148, 148, ..., 65909, 65909, 65910]),
col_indices=tensor([ 51, 111, 117, ..., 978, 8230, 12170]),
values=tensor([3., 4., 3., ..., 2., 1., 1.]), size=(31379, 31379),
nnz=65910, layout=torch.sparse_csr)
tensor([0.9176, 0.4871, 0.4502, ..., 0.9545, 0.3229, 0.4089])
Matrix: as-caida_G_001
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 65910
Density: 6.693805873684353e-05
Time: 10.126415729522705 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 127504, "MATRIX_FILE": "as-caida_G_002", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 66744, "MATRIX_DENSITY": 6.77850673999679e-05, "TIME_S": 11.147692203521729, "TIME_S_1KI": 0.08743013712135876, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 822.7334973049162, "W": 72.58391126827506, "J_1KI": 6.452609308766127, "W_1KI": 0.5692677191952806, "W_D": 53.68591126827506, "J_D": 608.5260047574042, "W_D_1KI": 0.42105276123317753, "J_D_1KI": 0.0033022709972485376}

View File

@ -0,0 +1,14 @@
/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, 129, 129, ..., 66743, 66743, 66744]),
col_indices=tensor([ 51, 111, 117, ..., 978, 8230, 12170]),
values=tensor([3., 4., 3., ..., 2., 1., 1.]), size=(31379, 31379),
nnz=66744, layout=torch.sparse_csr)
tensor([0.5711, 0.1553, 0.3490, ..., 0.7582, 0.3633, 0.1833])
Matrix: as-caida_G_002
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 66744
Density: 6.77850673999679e-05
Time: 11.147692203521729 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 143449, "MATRIX_FILE": "as-caida_G_003", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 66680, "MATRIX_DENSITY": 6.772006913325332e-05, "TIME_S": 14.305955648422241, "TIME_S_1KI": 0.09972851430419342, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 973.3464808654785, "W": 81.56949469547968, "J_1KI": 6.785313810939627, "W_1KI": 0.5686306261840771, "W_D": 62.753494695479674, "J_D": 748.8202967529297, "W_D_1KI": 0.43746205756387063, "J_D_1KI": 0.003049599910517819}

View File

@ -0,0 +1,14 @@
/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, 89, 89, ..., 66679, 66679, 66680]),
col_indices=tensor([ 51, 111, 117, ..., 978, 978, 12170]),
values=tensor([3., 4., 3., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=66680, layout=torch.sparse_csr)
tensor([0.2059, 0.9834, 0.9558, ..., 0.2703, 0.8687, 0.4075])
Matrix: as-caida_G_003
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 66680
Density: 6.772006913325332e-05
Time: 14.305955648422241 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 103560, "MATRIX_FILE": "as-caida_G_004", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 68670, "MATRIX_DENSITY": 6.97411089889098e-05, "TIME_S": 10.448071718215942, "TIME_S_1KI": 0.10088906641768967, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 629.299670639038, "W": 68.7216518977672, "J_1KI": 6.076667348774025, "W_1KI": 0.6635926216470375, "W_D": 49.7746518977672, "J_D": 455.79771702861774, "W_D_1KI": 0.4806358815929625, "J_D_1KI": 0.004641134430214007}

View File

@ -0,0 +1,14 @@
/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, 62, 62, ..., 68669, 68669, 68670]),
col_indices=tensor([ 111, 822, 875, ..., 160, 4019, 12170]),
values=tensor([4., 3., 3., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=68670, layout=torch.sparse_csr)
tensor([0.8601, 0.7413, 0.4398, ..., 0.7803, 0.6991, 0.8078])
Matrix: as-caida_G_004
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 68670
Density: 6.97411089889098e-05
Time: 10.448071718215942 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 120363, "MATRIX_FILE": "as-caida_G_005", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 70026, "MATRIX_DENSITY": 7.111825976492498e-05, "TIME_S": 10.082340002059937, "TIME_S_1KI": 0.08376610754185204, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 809.4879877853394, "W": 72.0223387430858, "J_1KI": 6.7253889300311505, "W_1KI": 0.5983760685849122, "W_D": 53.3753387430858, "J_D": 599.9068665435316, "W_D_1KI": 0.4434530440674111, "J_D_1KI": 0.003684297035363119}

View File

@ -0,0 +1,14 @@
/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, 63, 63, ..., 70025, 70025, 70026]),
col_indices=tensor([ 111, 761, 822, ..., 978, 978, 12170]),
values=tensor([4., 3., 3., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=70026, layout=torch.sparse_csr)
tensor([0.7040, 0.6399, 0.5818, ..., 0.1186, 0.7545, 0.7756])
Matrix: as-caida_G_005
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 70026
Density: 7.111825976492498e-05
Time: 10.082340002059937 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 118086, "MATRIX_FILE": "as-caida_G_006", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 71094, "MATRIX_DENSITY": 7.220291834072453e-05, "TIME_S": 10.220538854598999, "TIME_S_1KI": 0.08655165603542332, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 715.6333376598358, "W": 68.29903449226836, "J_1KI": 6.060272493435596, "W_1KI": 0.5783838430658025, "W_D": 49.44803449226836, "J_D": 518.1136487138272, "W_D_1KI": 0.4187459520372302, "J_D_1KI": 0.0035461100556986452}

View File

@ -0,0 +1,14 @@
/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, 60, 60, ..., 71093, 71093, 71094]),
col_indices=tensor([ 111, 761, 822, ..., 978, 978, 12170]),
values=tensor([4., 3., 3., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=71094, layout=torch.sparse_csr)
tensor([0.1653, 0.6941, 0.2553, ..., 0.0566, 0.9880, 0.5165])
Matrix: as-caida_G_006
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 71094
Density: 7.220291834072453e-05
Time: 10.220538854598999 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 142731, "MATRIX_FILE": "as-caida_G_007", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 71658, "MATRIX_DENSITY": 7.277571556614678e-05, "TIME_S": 12.983977794647217, "TIME_S_1KI": 0.09096816945616032, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 918.3490377235412, "W": 75.87233515955411, "J_1KI": 6.434124596083129, "W_1KI": 0.5315757274842474, "W_D": 57.07933515955411, "J_D": 690.8809700853824, "W_D_1KI": 0.3999084652917314, "J_D_1KI": 0.002801833275824673}

View File

@ -0,0 +1,14 @@
/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, 57, 57, ..., 71657, 71657, 71658]),
col_indices=tensor([ 111, 761, 822, ..., 978, 978, 12170]),
values=tensor([4., 3., 3., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=71658, layout=torch.sparse_csr)
tensor([0.9538, 0.1962, 0.6080, ..., 0.4987, 0.0086, 0.8802])
Matrix: as-caida_G_007
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 71658
Density: 7.277571556614678e-05
Time: 12.983977794647217 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 127900, "MATRIX_FILE": "as-caida_G_008", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 72140, "MATRIX_DENSITY": 7.326523376234096e-05, "TIME_S": 12.409928798675537, "TIME_S_1KI": 0.09702837215539903, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 921.8805131340027, "W": 78.83067805993187, "J_1KI": 7.207822620281491, "W_1KI": 0.6163461928063476, "W_D": 60.15167805993187, "J_D": 703.4400971865654, "W_D_1KI": 0.4703024085999364, "J_D_1KI": 0.003677110309616391}

View File

@ -0,0 +1,14 @@
/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, 55, 55, ..., 72139, 72139, 72140]),
col_indices=tensor([ 111, 761, 822, ..., 978, 978, 12170]),
values=tensor([4., 3., 3., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=72140, layout=torch.sparse_csr)
tensor([0.0748, 0.5344, 0.1913, ..., 0.6845, 0.0381, 0.0128])
Matrix: as-caida_G_008
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 72140
Density: 7.326523376234096e-05
Time: 12.409928798675537 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 129513, "MATRIX_FILE": "as-caida_G_009", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 74344, "MATRIX_DENSITY": 7.550361157232432e-05, "TIME_S": 10.678922891616821, "TIME_S_1KI": 0.08245444775132088, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 731.9839727020264, "W": 75.90516682079294, "J_1KI": 5.651818525569066, "W_1KI": 0.5860814498991834, "W_D": 57.05816682079294, "J_D": 550.2347913045884, "W_D_1KI": 0.4405593787557461, "J_D_1KI": 0.003401661445227476}

View File

@ -0,0 +1,14 @@
/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, 38, 38, ..., 74343, 74343, 74344]),
col_indices=tensor([ 111, 875, 1040, ..., 160, 4019, 12170]),
values=tensor([4., 3., 1., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=74344, layout=torch.sparse_csr)
tensor([0.5760, 0.2857, 0.5374, ..., 0.8141, 0.5093, 0.9535])
Matrix: as-caida_G_009
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 74344
Density: 7.550361157232432e-05
Time: 10.678922891616821 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 124622, "MATRIX_FILE": "as-caida_G_010", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 74994, "MATRIX_DENSITY": 7.616375021864427e-05, "TIME_S": 10.38208293914795, "TIME_S_1KI": 0.08330858868536815, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 957.5060850524903, "W": 71.48797724957048, "J_1KI": 7.683282927994177, "W_1KI": 0.5736385008230528, "W_D": 52.79097724957048, "J_D": 707.0794824123384, "W_D_1KI": 0.4236088110411523, "J_D_1KI": 0.0033991495164670143}

View File

@ -0,0 +1,14 @@
/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, 28, 28, ..., 74993, 74993, 74994]),
col_indices=tensor([ 1040, 2020, 2054, ..., 160, 160, 12170]),
values=tensor([1., 3., 3., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=74994, layout=torch.sparse_csr)
tensor([0.1218, 0.1233, 0.3562, ..., 0.2874, 0.3356, 0.1438])
Matrix: as-caida_G_010
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 74994
Density: 7.616375021864427e-05
Time: 10.38208293914795 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 142236, "MATRIX_FILE": "as-caida_G_011", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 75118, "MATRIX_DENSITY": 7.628968436040377e-05, "TIME_S": 11.37498664855957, "TIME_S_1KI": 0.07997262752439306, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 922.4947763061523, "W": 78.68239973433768, "J_1KI": 6.485663097290083, "W_1KI": 0.5531820336225547, "W_D": 59.73639973433768, "J_D": 700.3664974164961, "W_D_1KI": 0.41998087498479764, "J_D_1KI": 0.002952704483989972}

View File

@ -0,0 +1,14 @@
/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, 25, 25, ..., 75117, 75117, 75118]),
col_indices=tensor([ 1040, 2020, 2054, ..., 160, 160, 12170]),
values=tensor([1., 3., 3., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=75118, layout=torch.sparse_csr)
tensor([0.9968, 0.0059, 0.7418, ..., 0.5181, 0.9717, 0.5988])
Matrix: as-caida_G_011
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 75118
Density: 7.628968436040377e-05
Time: 11.37498664855957 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 108283, "MATRIX_FILE": "as-caida_G_012", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 76530, "MATRIX_DENSITY": 7.772370861979419e-05, "TIME_S": 10.989976406097412, "TIME_S_1KI": 0.10149309130793766, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 780.6230658721925, "W": 78.37336875797533, "J_1KI": 7.20910083643963, "W_1KI": 0.7237827614489377, "W_D": 59.35736875797533, "J_D": 591.2178067150118, "W_D_1KI": 0.5481688608366533, "J_D_1KI": 0.005062372309934646}

View File

@ -0,0 +1,14 @@
/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, 3, ..., 76530, 76530, 76530]),
col_indices=tensor([1040, 2054, 5699, ..., 160, 160, 7454]),
values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=76530, layout=torch.sparse_csr)
tensor([0.7299, 0.6893, 0.7667, ..., 0.2441, 0.4109, 0.3491])
Matrix: as-caida_G_012
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 76530
Density: 7.772370861979419e-05
Time: 10.989976406097412 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 123279, "MATRIX_FILE": "as-caida_G_013", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 77002, "MATRIX_DENSITY": 7.820307083681422e-05, "TIME_S": 10.23405146598816, "TIME_S_1KI": 0.08301536730495995, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 676.166474685669, "W": 70.81417020223847, "J_1KI": 5.484847173368286, "W_1KI": 0.5744220037657547, "W_D": 50.85017020223847, "J_D": 485.5409619941712, "W_D_1KI": 0.412480391650147, "J_D_1KI": 0.003345909616805352}

View File

@ -0,0 +1,14 @@
/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, 5, 5, ..., 77002, 77002, 77002]),
col_indices=tensor([1040, 2054, 2150, ..., 160, 8230, 7454]),
values=tensor([1., 1., 3., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=77002, layout=torch.sparse_csr)
tensor([0.5264, 0.1057, 0.6362, ..., 0.0751, 0.3658, 0.4784])
Matrix: as-caida_G_013
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 77002
Density: 7.820307083681422e-05
Time: 10.23405146598816 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 121400, "MATRIX_FILE": "as-caida_G_014", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 75982, "MATRIX_DENSITY": 7.71671609610506e-05, "TIME_S": 10.063333511352539, "TIME_S_1KI": 0.08289401574425485, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1041.4476663208006, "W": 81.6299534814742, "J_1KI": 8.578646345311371, "W_1KI": 0.6724048886447629, "W_D": 62.811953481474205, "J_D": 801.3646900482177, "W_D_1KI": 0.5173966514124728, "J_D_1KI": 0.004261916403727124}

View File

@ -0,0 +1,14 @@
/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, 5, 5, ..., 75982, 75982, 75982]),
col_indices=tensor([1040, 2054, 2150, ..., 160, 160, 8230]),
values=tensor([1., 1., 3., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=75982, layout=torch.sparse_csr)
tensor([0.6612, 0.7577, 0.1054, ..., 0.0684, 0.9601, 0.4212])
Matrix: as-caida_G_014
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 75982
Density: 7.71671609610506e-05
Time: 10.063333511352539 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 127630, "MATRIX_FILE": "as-caida_G_015", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 77124, "MATRIX_DENSITY": 7.832697378273889e-05, "TIME_S": 12.151241302490234, "TIME_S_1KI": 0.0952067797734877, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1043.6709648132326, "W": 76.83403885753171, "J_1KI": 8.177316969468249, "W_1KI": 0.6020061024644026, "W_D": 58.006038857531706, "J_D": 787.9218565053942, "W_D_1KI": 0.4544859269570768, "J_D_1KI": 0.0035609647179901025}

View File

@ -0,0 +1,14 @@
/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, 4, ..., 77124, 77124, 77124]),
col_indices=tensor([1040, 2054, 4842, ..., 160, 160, 8230]),
values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=77124, layout=torch.sparse_csr)
tensor([0.6540, 0.7414, 0.3084, ..., 0.9960, 0.8018, 0.9618])
Matrix: as-caida_G_015
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 77124
Density: 7.832697378273889e-05
Time: 12.151241302490234 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 107877, "MATRIX_FILE": "as-caida_G_016", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 77860, "MATRIX_DENSITY": 7.907445384995657e-05, "TIME_S": 11.006463050842285, "TIME_S_1KI": 0.10202789334929861, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 513.9647554016113, "W": 62.33405857814145, "J_1KI": 4.764358995908409, "W_1KI": 0.5778252878569247, "W_D": 43.50305857814145, "J_D": 358.69698478412624, "W_D_1KI": 0.4032653723976515, "J_D_1KI": 0.0037381960232269295}

View File

@ -0,0 +1,14 @@
/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, 3, ..., 77856, 77858, 77860]),
col_indices=tensor([ 1040, 4842, 5699, ..., 31378, 17998, 31377]),
values=tensor([1., 1., 1., ..., 1., 1., 3.]), size=(31379, 31379),
nnz=77860, layout=torch.sparse_csr)
tensor([0.4735, 0.8183, 0.1183, ..., 0.2637, 0.0889, 0.4462])
Matrix: as-caida_G_016
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 77860
Density: 7.907445384995657e-05
Time: 11.006463050842285 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 116896, "MATRIX_FILE": "as-caida_G_017", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 79718, "MATRIX_DENSITY": 8.096143478051423e-05, "TIME_S": 12.301311254501343, "TIME_S_1KI": 0.10523295283415467, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 661.7432917404175, "W": 69.21782114353806, "J_1KI": 5.660957532682192, "W_1KI": 0.5921316481619393, "W_D": 49.28182114353806, "J_D": 471.14910593414305, "W_D_1KI": 0.4215868904285695, "J_D_1KI": 0.0036065125447283866}

View File

@ -0,0 +1,14 @@
/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, 2, ..., 79714, 79716, 79718]),
col_indices=tensor([ 1040, 5699, 33, ..., 31378, 17998, 31377]),
values=tensor([1., 1., 4., ..., 1., 1., 3.]), size=(31379, 31379),
nnz=79718, layout=torch.sparse_csr)
tensor([0.1143, 0.2051, 0.8521, ..., 0.2485, 0.3618, 0.1726])
Matrix: as-caida_G_017
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 79718
Density: 8.096143478051423e-05
Time: 12.301311254501343 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 116812, "MATRIX_FILE": "as-caida_G_018", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 80396, "MATRIX_DENSITY": 8.16500101685218e-05, "TIME_S": 10.66378378868103, "TIME_S_1KI": 0.09129013961477443, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 778.7428970146179, "W": 74.46749600000055, "J_1KI": 6.666634395563966, "W_1KI": 0.6374986816423017, "W_D": 55.170496000000554, "J_D": 576.9447637231351, "W_D_1KI": 0.4723016128479998, "J_D_1KI": 0.004043262788480634}

View File

@ -0,0 +1,14 @@
/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, 3, ..., 80392, 80394, 80396]),
col_indices=tensor([ 1040, 5282, 5699, ..., 31378, 17998, 31377]),
values=tensor([1., 1., 1., ..., 1., 1., 3.]), size=(31379, 31379),
nnz=80396, layout=torch.sparse_csr)
tensor([0.8702, 0.4804, 0.1121, ..., 0.8957, 0.0227, 0.0461])
Matrix: as-caida_G_018
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 80396
Density: 8.16500101685218e-05
Time: 10.66378378868103 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 129456, "MATRIX_FILE": "as-caida_G_019", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 80970, "MATRIX_DENSITY": 8.22329633731182e-05, "TIME_S": 12.674752235412598, "TIME_S_1KI": 0.0979078006072534, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 739.2131903076173, "W": 74.1279715637925, "J_1KI": 5.710150091981966, "W_1KI": 0.5726113240312731, "W_D": 55.1819715637925, "J_D": 550.2813632507325, "W_D_1KI": 0.4262604403333372, "J_D_1KI": 0.003292705168808995}

View File

@ -0,0 +1,14 @@
/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, 2, ..., 80966, 80968, 80970]),
col_indices=tensor([ 1040, 5699, 106, ..., 31378, 17998, 31377]),
values=tensor([1., 1., 1., ..., 1., 1., 3.]), size=(31379, 31379),
nnz=80970, layout=torch.sparse_csr)
tensor([0.1947, 0.2233, 0.6922, ..., 0.1662, 0.9424, 0.1964])
Matrix: as-caida_G_019
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 80970
Density: 8.22329633731182e-05
Time: 12.674752235412598 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 121477, "MATRIX_FILE": "as-caida_G_020", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 80948, "MATRIX_DENSITY": 8.221062021893506e-05, "TIME_S": 10.737972974777222, "TIME_S_1KI": 0.08839511162423522, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 957.4885776519776, "W": 79.88900379083044, "J_1KI": 7.882056501658566, "W_1KI": 0.6576471578227191, "W_D": 59.867003790830445, "J_D": 717.5201790981293, "W_D_1KI": 0.49282583362142995, "J_D_1KI": 0.004056947682453715}

View File

@ -0,0 +1,14 @@
/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, 2, ..., 80944, 80946, 80948]),
col_indices=tensor([ 1040, 5699, 106, ..., 31378, 17998, 31377]),
values=tensor([1., 1., 1., ..., 1., 1., 3.]), size=(31379, 31379),
nnz=80948, layout=torch.sparse_csr)
tensor([0.6576, 0.0654, 0.6667, ..., 0.9255, 0.6823, 0.1685])
Matrix: as-caida_G_020
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 80948
Density: 8.221062021893506e-05
Time: 10.737972974777222 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 101211, "MATRIX_FILE": "as-caida_G_021", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 81882, "MATRIX_DENSITY": 8.315918867380097e-05, "TIME_S": 10.29597282409668, "TIME_S_1KI": 0.10172780452813113, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 898.7896647548675, "W": 74.06007424440759, "J_1KI": 8.880355541935833, "W_1KI": 0.731739378569598, "W_D": 55.56207424440759, "J_D": 674.2987850432396, "W_D_1KI": 0.5489726832499194, "J_D_1KI": 0.005424041687661613}

View File

@ -0,0 +1,14 @@
/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, 2, ..., 81878, 81880, 81882]),
col_indices=tensor([ 1040, 5699, 106, ..., 31378, 17998, 31377]),
values=tensor([1., 1., 1., ..., 1., 1., 3.]), size=(31379, 31379),
nnz=81882, layout=torch.sparse_csr)
tensor([0.5868, 0.6239, 0.1382, ..., 0.7506, 0.4458, 0.8585])
Matrix: as-caida_G_021
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 81882
Density: 8.315918867380097e-05
Time: 10.29597282409668 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 97313, "MATRIX_FILE": "as-caida_G_022", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 82138, "MATRIX_DENSITY": 8.341918174065929e-05, "TIME_S": 10.294092178344727, "TIME_S_1KI": 0.10578331958057738, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 537.0419758224486, "W": 67.61006796553038, "J_1KI": 5.518707426782122, "W_1KI": 0.6947691260728821, "W_D": 48.796067965530376, "J_D": 387.59814242362967, "W_D_1KI": 0.5014342170679187, "J_D_1KI": 0.005152797848878553}

View File

@ -0,0 +1,14 @@
/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, ..., 82134, 82136, 82138]),
col_indices=tensor([ 106, 329, 1040, ..., 31378, 17998, 31377]),
values=tensor([1., 1., 1., ..., 1., 1., 3.]), size=(31379, 31379),
nnz=82138, layout=torch.sparse_csr)
tensor([0.5654, 0.6771, 0.6558, ..., 0.6752, 0.9803, 0.3834])
Matrix: as-caida_G_022
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 82138
Density: 8.341918174065929e-05
Time: 10.294092178344727 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 140111, "MATRIX_FILE": "as-caida_G_023", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 83524, "MATRIX_DENSITY": 8.482680045419692e-05, "TIME_S": 12.064657926559448, "TIME_S_1KI": 0.08610785681751931, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 904.5556343078613, "W": 76.45920851310774, "J_1KI": 6.455992993468473, "W_1KI": 0.5457045379242724, "W_D": 57.44720851310774, "J_D": 679.6329329891205, "W_D_1KI": 0.41001212262497405, "J_D_1KI": 0.002926337850882329}

View File

@ -0,0 +1,14 @@
/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, ..., 83520, 83522, 83524]),
col_indices=tensor([ 106, 329, 1040, ..., 31378, 17998, 31377]),
values=tensor([1., 1., 1., ..., 1., 1., 3.]), size=(31379, 31379),
nnz=83524, layout=torch.sparse_csr)
tensor([0.6514, 0.1380, 0.8213, ..., 0.6748, 0.3381, 0.7461])
Matrix: as-caida_G_023
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 83524
Density: 8.482680045419692e-05
Time: 12.064657926559448 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 116214, "MATRIX_FILE": "as-caida_G_024", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 83640, "MATRIX_DENSITY": 8.49446098126171e-05, "TIME_S": 11.284576892852783, "TIME_S_1KI": 0.09710169938951231, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1002.3741082382203, "W": 78.47596712397663, "J_1KI": 8.625244017400831, "W_1KI": 0.6752711990291757, "W_D": 59.85396712397663, "J_D": 764.5151645679475, "W_D_1KI": 0.5150323293577076, "J_D_1KI": 0.0044317580442778635}

View File

@ -0,0 +1,14 @@
/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, ..., 83636, 83638, 83640]),
col_indices=tensor([ 106, 329, 1040, ..., 31378, 17998, 31377]),
values=tensor([1., 1., 1., ..., 1., 1., 3.]), size=(31379, 31379),
nnz=83640, layout=torch.sparse_csr)
tensor([0.3311, 0.4574, 0.7903, ..., 0.2454, 0.1433, 0.1661])
Matrix: as-caida_G_024
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 83640
Density: 8.49446098126171e-05
Time: 11.284576892852783 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 115508, "MATRIX_FILE": "as-caida_G_025", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 85850, "MATRIX_DENSITY": 8.718908121010495e-05, "TIME_S": 11.6570463180542, "TIME_S_1KI": 0.10091981783126883, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 777.9111336517335, "W": 74.52510352442891, "J_1KI": 6.734694857946925, "W_1KI": 0.6451943027706212, "W_D": 55.619103524428915, "J_D": 580.5657131519318, "W_D_1KI": 0.48151732801562586, "J_D_1KI": 0.004168692454337586}

View File

@ -0,0 +1,14 @@
/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, 3, ..., 85845, 85847, 85850]),
col_indices=tensor([ 346, 13811, 21783, ..., 15310, 17998, 31377]),
values=tensor([1., 1., 1., ..., 1., 1., 3.]), size=(31379, 31379),
nnz=85850, layout=torch.sparse_csr)
tensor([0.9169, 0.1387, 0.2857, ..., 0.8309, 0.2879, 0.8450])
Matrix: as-caida_G_025
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 85850
Density: 8.718908121010495e-05
Time: 11.6570463180542 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 114440, "MATRIX_FILE": "as-caida_G_026", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 85468, "MATRIX_DENSITY": 8.68011228056523e-05, "TIME_S": 11.638119459152222, "TIME_S_1KI": 0.10169625532289603, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 906.8806961631776, "W": 78.96386440870926, "J_1KI": 7.924508005620217, "W_1KI": 0.6900023104570889, "W_D": 60.217864408709254, "J_D": 691.5874647898675, "W_D_1KI": 0.5261959490449952, "J_D_1KI": 0.004598007244363817}

View File

@ -0,0 +1,14 @@
/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, 1, ..., 85463, 85465, 85468]),
col_indices=tensor([21783, 106, 329, ..., 15310, 17998, 31377]),
values=tensor([1., 1., 1., ..., 1., 1., 3.]), size=(31379, 31379),
nnz=85468, layout=torch.sparse_csr)
tensor([0.1366, 0.2794, 0.6490, ..., 0.4198, 0.3634, 0.1789])
Matrix: as-caida_G_026
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 85468
Density: 8.68011228056523e-05
Time: 11.638119459152222 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 102448, "MATRIX_FILE": "as-caida_G_027", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 85744, "MATRIX_DENSITY": 8.708142783085892e-05, "TIME_S": 10.685343265533447, "TIME_S_1KI": 0.10430016462530696, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 623.8277059173585, "W": 66.90704171784165, "J_1KI": 6.089213121948291, "W_1KI": 0.6530829466445577, "W_D": 48.12704171784165, "J_D": 448.7267895364762, "W_D_1KI": 0.4697704368835082, "J_D_1KI": 0.004585452491834962}

View File

@ -0,0 +1,14 @@
/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, 1, ..., 85739, 85741, 85744]),
col_indices=tensor([21783, 106, 329, ..., 15310, 17998, 31377]),
values=tensor([1., 1., 1., ..., 1., 1., 3.]), size=(31379, 31379),
nnz=85744, layout=torch.sparse_csr)
tensor([0.2505, 0.1829, 0.9414, ..., 0.0525, 0.8527, 0.8419])
Matrix: as-caida_G_027
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 85744
Density: 8.708142783085892e-05
Time: 10.685343265533447 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 106948, "MATRIX_FILE": "as-caida_G_028", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 84962, "MATRIX_DENSITY": 8.628723025944015e-05, "TIME_S": 10.41547679901123, "TIME_S_1KI": 0.0973882335248086, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 648.8419272613526, "W": 67.54788598351732, "J_1KI": 6.066891641371065, "W_1KI": 0.631595597706524, "W_D": 48.72688598351732, "J_D": 468.05382804584514, "W_D_1KI": 0.45561287713203913, "J_D_1KI": 0.00426013461805774}

View File

@ -0,0 +1,14 @@
/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, 2, ..., 84957, 84959, 84962]),
col_indices=tensor([ 1809, 21783, 106, ..., 15310, 17998, 31377]),
values=tensor([1., 1., 1., ..., 1., 1., 3.]), size=(31379, 31379),
nnz=84962, layout=torch.sparse_csr)
tensor([0.5564, 0.8395, 0.7950, ..., 0.7334, 0.0065, 0.6442])
Matrix: as-caida_G_028
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 84962
Density: 8.628723025944015e-05
Time: 10.41547679901123 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 129559, "MATRIX_FILE": "as-caida_G_029", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 86566, "MATRIX_DENSITY": 8.791624931897431e-05, "TIME_S": 13.361659049987793, "TIME_S_1KI": 0.10313184765232668, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 850.7577786254884, "W": 72.8947653550395, "J_1KI": 6.566566418585266, "W_1KI": 0.5626376041420472, "W_D": 53.7847653550395, "J_D": 627.7241894459726, "W_D_1KI": 0.41513723751371573, "J_D_1KI": 0.0032042331101175196}

View File

@ -0,0 +1,14 @@
/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, 2, ..., 86561, 86563, 86566]),
col_indices=tensor([ 1809, 21783, 106, ..., 15310, 17998, 31377]),
values=tensor([1., 1., 1., ..., 1., 1., 3.]), size=(31379, 31379),
nnz=86566, layout=torch.sparse_csr)
tensor([0.7891, 0.7339, 0.3751, ..., 0.4286, 0.1932, 0.8774])
Matrix: as-caida_G_029
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 86566
Density: 8.791624931897431e-05
Time: 13.361659049987793 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 126211, "MATRIX_FILE": "as-caida_G_030", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 86850, "MATRIX_DENSITY": 8.820467912752026e-05, "TIME_S": 10.005269289016724, "TIME_S_1KI": 0.07927414638198511, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1141.9418496513367, "W": 81.59308951543089, "J_1KI": 9.04787894598202, "W_1KI": 0.6464816023597855, "W_D": 62.695089515430894, "J_D": 877.4535553255082, "W_D_1KI": 0.4967482193741503, "J_D_1KI": 0.003935855189913322}

View File

@ -0,0 +1,14 @@
/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, 2, ..., 86850, 86850, 86850]),
col_indices=tensor([ 1809, 21783, 106, ..., 7018, 160, 882]),
values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=86850, layout=torch.sparse_csr)
tensor([0.2301, 0.9372, 0.3059, ..., 0.8450, 0.5545, 0.5854])
Matrix: as-caida_G_030
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 86850
Density: 8.820467912752026e-05
Time: 10.005269289016724 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 105338, "MATRIX_FILE": "as-caida_G_031", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 87232, "MATRIX_DENSITY": 8.859263753197291e-05, "TIME_S": 10.614361763000488, "TIME_S_1KI": 0.10076479298069536, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 799.4572194194793, "W": 72.010733194549, "J_1KI": 7.589447487321568, "W_1KI": 0.6836159144330536, "W_D": 53.18673319454901, "J_D": 590.4747243003845, "W_D_1KI": 0.504914970803974, "J_D_1KI": 0.004793284197573278}

View File

@ -0,0 +1,14 @@
/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, 2, ..., 87232, 87232, 87232]),
col_indices=tensor([ 1809, 21783, 106, ..., 1101, 16290, 882]),
values=tensor([1., 1., 1., ..., 2., 1., 1.]), size=(31379, 31379),
nnz=87232, layout=torch.sparse_csr)
tensor([0.6837, 0.2004, 0.3302, ..., 0.8179, 0.3608, 0.2300])
Matrix: as-caida_G_031
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 87232
Density: 8.859263753197291e-05
Time: 10.614361763000488 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 128646, "MATRIX_FILE": "as-caida_G_032", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 87938, "MATRIX_DENSITY": 8.930964966166813e-05, "TIME_S": 12.817463397979736, "TIME_S_1KI": 0.09963359449947715, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 975.9403296279908, "W": 81.64731509315205, "J_1KI": 7.586246984966426, "W_1KI": 0.634666566338262, "W_D": 62.757315093152044, "J_D": 750.1458524227144, "W_D_1KI": 0.4878295096089427, "J_D_1KI": 0.0037920301416984807}

View File

@ -0,0 +1,14 @@
/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, 2, ..., 87938, 87938, 87938]),
col_indices=tensor([ 1809, 21783, 106, ..., 160, 16290, 882]),
values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=87938, layout=torch.sparse_csr)
tensor([0.8667, 0.8794, 0.6030, ..., 0.3085, 0.3900, 0.2285])
Matrix: as-caida_G_032
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 87938
Density: 8.930964966166813e-05
Time: 12.817463397979736 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 108236, "MATRIX_FILE": "as-caida_G_033", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 87784, "MATRIX_DENSITY": 8.915324758238617e-05, "TIME_S": 11.615520715713501, "TIME_S_1KI": 0.10731661106945473, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 814.9971894931793, "W": 73.68127735229483, "J_1KI": 7.529816230211568, "W_1KI": 0.6807464924082083, "W_D": 54.987277352294825, "J_D": 608.2206784462929, "W_D_1KI": 0.5080313144637165, "J_D_1KI": 0.004693736967956285}

View File

@ -0,0 +1,14 @@
/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, 2, ..., 87784, 87784, 87784]),
col_indices=tensor([ 1809, 21783, 106, ..., 160, 16290, 882]),
values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=87784, layout=torch.sparse_csr)
tensor([0.1413, 0.0383, 0.6907, ..., 0.9449, 0.3132, 0.3976])
Matrix: as-caida_G_033
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 87784
Density: 8.915324758238617e-05
Time: 11.615520715713501 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 118393, "MATRIX_FILE": "as-caida_G_034", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 88490, "MATRIX_DENSITY": 8.987025971208138e-05, "TIME_S": 10.513860940933228, "TIME_S_1KI": 0.0888047514712291, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 818.8505859184265, "W": 73.52922556303459, "J_1KI": 6.916376693879085, "W_1KI": 0.6210605826614292, "W_D": 54.69022556303459, "J_D": 609.0520184781552, "W_D_1KI": 0.4619379994005945, "J_D_1KI": 0.0039017340501600137}

View File

@ -0,0 +1,14 @@
/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, 2, ..., 88489, 88489, 88490]),
col_indices=tensor([ 1809, 21783, 106, ..., 10144, 882, 16085]),
values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=88490, layout=torch.sparse_csr)
tensor([0.8060, 0.0837, 0.1341, ..., 0.3387, 0.8994, 0.6830])
Matrix: as-caida_G_034
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 88490
Density: 8.987025971208138e-05
Time: 10.513860940933228 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 139029, "MATRIX_FILE": "as-caida_G_035", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 87560, "MATRIX_DENSITY": 8.892575364888514e-05, "TIME_S": 13.630029678344727, "TIME_S_1KI": 0.0980373136420799, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 946.0831095790862, "W": 76.8712422214838, "J_1KI": 6.804933571982005, "W_1KI": 0.5529151631780693, "W_D": 57.994242221483795, "J_D": 713.7568150713442, "W_D_1KI": 0.4171377354471642, "J_D_1KI": 0.003000364927081143}

View File

@ -0,0 +1,14 @@
/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, 2, ..., 87559, 87559, 87560]),
col_indices=tensor([ 1809, 21783, 106, ..., 10144, 882, 16085]),
values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=87560, layout=torch.sparse_csr)
tensor([0.2258, 0.9079, 0.9180, ..., 0.7245, 0.2460, 0.3925])
Matrix: as-caida_G_035
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 87560
Density: 8.892575364888514e-05
Time: 13.630029678344727 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 130613, "MATRIX_FILE": "as-caida_G_036", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 88616, "MATRIX_DENSITY": 8.999822504967571e-05, "TIME_S": 12.590465784072876, "TIME_S_1KI": 0.09639519637457891, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 851.3253401374816, "W": 78.82016710712844, "J_1KI": 6.517921953691299, "W_1KI": 0.6034634156410804, "W_D": 60.08416710712844, "J_D": 648.9604865951537, "W_D_1KI": 0.4600167449421454, "J_D_1KI": 0.0035219828419999957}

View File

@ -0,0 +1,14 @@
/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, 2, ..., 88615, 88615, 88616]),
col_indices=tensor([ 1809, 21783, 106, ..., 10144, 882, 16085]),
values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=88616, layout=torch.sparse_csr)
tensor([0.1612, 0.9891, 0.1530, ..., 0.9514, 0.9568, 0.3072])
Matrix: as-caida_G_036
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 88616
Density: 8.999822504967571e-05
Time: 12.590465784072876 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 138339, "MATRIX_FILE": "as-caida_G_037", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 88654, "MATRIX_DENSITY": 9.003681777053749e-05, "TIME_S": 12.654316425323486, "TIME_S_1KI": 0.09147323911061585, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 888.2931414222718, "W": 75.51405609798103, "J_1KI": 6.421133168681802, "W_1KI": 0.5458623822492648, "W_D": 56.56405609798102, "J_D": 665.3789463734628, "W_D_1KI": 0.4088800417668266, "J_D_1KI": 0.0029556382637349306}

View File

@ -0,0 +1,14 @@
/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, 1, ..., 88653, 88653, 88654]),
col_indices=tensor([ 1809, 106, 329, ..., 10144, 882, 16085]),
values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=88654, layout=torch.sparse_csr)
tensor([0.8026, 0.1152, 0.4121, ..., 0.0272, 0.6179, 0.0032])
Matrix: as-caida_G_037
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 88654
Density: 9.003681777053749e-05
Time: 12.654316425323486 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 127019, "MATRIX_FILE": "as-caida_G_038", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 88644, "MATRIX_DENSITY": 9.002666179136334e-05, "TIME_S": 12.376951217651367, "TIME_S_1KI": 0.0974417309036551, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1144.6182090759278, "W": 77.91927208125107, "J_1KI": 9.01139364249386, "W_1KI": 0.6134458000870032, "W_D": 59.13827208125106, "J_D": 868.7291509461402, "W_D_1KI": 0.46558603107606783, "J_D_1KI": 0.003665483361355922}

View File

@ -0,0 +1,14 @@
/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, 2, ..., 88643, 88643, 88644]),
col_indices=tensor([ 1809, 21783, 106, ..., 10144, 882, 16085]),
values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=88644, layout=torch.sparse_csr)
tensor([0.2941, 0.4961, 0.4501, ..., 0.6791, 0.6484, 0.8779])
Matrix: as-caida_G_038
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 88644
Density: 9.002666179136334e-05
Time: 12.376951217651367 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 141833, "MATRIX_FILE": "as-caida_G_039", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 87804, "MATRIX_DENSITY": 8.917355954073447e-05, "TIME_S": 12.296685218811035, "TIME_S_1KI": 0.0866983369089777, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1033.3566898822785, "W": 80.89532252127742, "J_1KI": 7.285728214747474, "W_1KI": 0.5703561408224984, "W_D": 61.89832252127742, "J_D": 790.6890494568347, "W_D_1KI": 0.43641693062458964, "J_D_1KI": 0.0030769773651025474}

View File

@ -0,0 +1,14 @@
/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, 1, ..., 87803, 87803, 87804]),
col_indices=tensor([ 1809, 106, 329, ..., 10144, 882, 16085]),
values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=87804, layout=torch.sparse_csr)
tensor([0.1788, 0.5284, 0.8814, ..., 0.5648, 0.9765, 0.7499])
Matrix: as-caida_G_039
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 87804
Density: 8.917355954073447e-05
Time: 12.296685218811035 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 126480, "MATRIX_FILE": "as-caida_G_040", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 89658, "MATRIX_DENSITY": 9.105647807962247e-05, "TIME_S": 10.94791579246521, "TIME_S_1KI": 0.0865584740074732, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1060.2959427833557, "W": 82.28098111370639, "J_1KI": 8.383111502082192, "W_1KI": 0.650545391474592, "W_D": 63.29598111370639, "J_D": 815.6498751103878, "W_D_1KI": 0.5004426084258886, "J_D_1KI": 0.003956693615005445}

View File

@ -0,0 +1,14 @@
/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, ..., 89657, 89657, 89658]),
col_indices=tensor([ 106, 329, 1040, ..., 10144, 882, 16085]),
values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=89658, layout=torch.sparse_csr)
tensor([0.5503, 0.9498, 0.7790, ..., 0.0188, 0.3119, 0.2062])
Matrix: as-caida_G_040
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 89658
Density: 9.105647807962247e-05
Time: 10.94791579246521 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 105131, "MATRIX_FILE": "as-caida_G_041", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 88944, "MATRIX_DENSITY": 9.033134116658794e-05, "TIME_S": 13.39965295791626, "TIME_S_1KI": 0.12745672501846517, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 743.1411743736267, "W": 70.18952496371918, "J_1KI": 7.068715929398814, "W_1KI": 0.6676387075526646, "W_D": 51.187524963719184, "J_D": 541.9549061555863, "W_D_1KI": 0.4868927810419304, "J_D_1KI": 0.004631296012041456}

View File

@ -0,0 +1,14 @@
/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, ..., 88943, 88943, 88944]),
col_indices=tensor([ 106, 329, 1040, ..., 10144, 882, 16085]),
values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=88944, layout=torch.sparse_csr)
tensor([0.4738, 0.4885, 0.1707, ..., 0.1175, 0.9603, 0.6544])
Matrix: as-caida_G_041
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 88944
Density: 9.033134116658794e-05
Time: 13.39965295791626 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 115909, "MATRIX_FILE": "as-caida_G_042", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 89176, "MATRIX_DENSITY": 9.056695988342829e-05, "TIME_S": 11.162479877471924, "TIME_S_1KI": 0.09630382349491345, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 797.8189125061035, "W": 71.99675341775847, "J_1KI": 6.883148957424389, "W_1KI": 0.6211489480347383, "W_D": 53.25075341775847, "J_D": 590.0885271224976, "W_D_1KI": 0.45941862510899467, "J_D_1KI": 0.003963614776324485}

View File

@ -0,0 +1,14 @@
/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, ..., 89173, 89173, 89176]),
col_indices=tensor([ 106, 329, 1040, ..., 2232, 5295, 16085]),
values=tensor([1., 1., 1., ..., 1., 1., 3.]), size=(31379, 31379),
nnz=89176, layout=torch.sparse_csr)
tensor([0.8462, 0.1475, 0.5030, ..., 0.4659, 0.2643, 0.5823])
Matrix: as-caida_G_042
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 89176
Density: 9.056695988342829e-05
Time: 11.162479877471924 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 130909, "MATRIX_FILE": "as-caida_G_043", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 89678, "MATRIX_DENSITY": 9.107679003797077e-05, "TIME_S": 12.607298135757446, "TIME_S_1KI": 0.09630581652718642, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 920.6237554645538, "W": 74.85525467413538, "J_1KI": 7.032547460178856, "W_1KI": 0.5718113702964303, "W_D": 56.03525467413538, "J_D": 689.1618607282638, "W_D_1KI": 0.42804738157143807, "J_D_1KI": 0.0032698086577044976}

View File

@ -0,0 +1,14 @@
/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, ..., 89676, 89676, 89678]),
col_indices=tensor([ 106, 329, 1040, ..., 882, 2232, 16085]),
values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=89678, layout=torch.sparse_csr)
tensor([0.4155, 0.5214, 0.2516, ..., 0.2688, 0.8252, 0.1018])
Matrix: as-caida_G_043
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 89678
Density: 9.107679003797077e-05
Time: 12.607298135757446 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 106563, "MATRIX_FILE": "as-caida_G_044", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 89560, "MATRIX_DENSITY": 9.095694948371577e-05, "TIME_S": 12.652342557907104, "TIME_S_1KI": 0.11873110327137097, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 621.244135837555, "W": 66.15231158833835, "J_1KI": 5.829829639157634, "W_1KI": 0.6207812429111262, "W_D": 47.27031158833835, "J_D": 443.9210538282395, "W_D_1KI": 0.44359028544934315, "J_D_1KI": 0.004162704554576571}

View File

@ -0,0 +1,14 @@
/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, 1, ..., 89559, 89559, 89560]),
col_indices=tensor([ 5326, 106, 329, ..., 10144, 882, 16085]),
values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=89560, layout=torch.sparse_csr)
tensor([0.4404, 0.2775, 0.5687, ..., 0.1114, 0.2519, 0.8574])
Matrix: as-caida_G_044
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 89560
Density: 9.095694948371577e-05
Time: 12.652342557907104 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 122630, "MATRIX_FILE": "as-caida_G_045", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 89152, "MATRIX_DENSITY": 9.054258553341032e-05, "TIME_S": 12.735206127166748, "TIME_S_1KI": 0.10385065748321576, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1027.8703737640383, "W": 77.18248362128287, "J_1KI": 8.381883501296896, "W_1KI": 0.6293931633473283, "W_D": 58.48148362128287, "J_D": 778.8215875902177, "W_D_1KI": 0.4768937749431858, "J_D_1KI": 0.0038888834293662705}

View File

@ -0,0 +1,14 @@
/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, ..., 89150, 89150, 89152]),
col_indices=tensor([ 106, 329, 1040, ..., 160, 2232, 16085]),
values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=89152, layout=torch.sparse_csr)
tensor([0.4846, 0.9604, 0.9616, ..., 0.8396, 0.3019, 0.8305])
Matrix: as-caida_G_045
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 89152
Density: 9.054258553341032e-05
Time: 12.735206127166748 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 113074, "MATRIX_FILE": "as-caida_G_046", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 90172, "MATRIX_DENSITY": 9.157849540917394e-05, "TIME_S": 10.500521898269653, "TIME_S_1KI": 0.09286415885411017, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 758.198433933258, "W": 73.8415920062998, "J_1KI": 6.705329553507066, "W_1KI": 0.6530377629366593, "W_D": 55.064592006299804, "J_D": 565.3979862837791, "W_D_1KI": 0.4869783682040063, "J_D_1KI": 0.004306722749739165}

View File

@ -0,0 +1,14 @@
/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, ..., 90170, 90170, 90172]),
col_indices=tensor([ 106, 329, 1040, ..., 882, 2232, 16085]),
values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=90172, layout=torch.sparse_csr)
tensor([0.3790, 0.3032, 0.0405, ..., 0.0113, 0.1281, 0.2860])
Matrix: as-caida_G_046
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 90172
Density: 9.157849540917394e-05
Time: 10.500521898269653 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 126496, "MATRIX_FILE": "as-caida_G_047", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 90002, "MATRIX_DENSITY": 9.140584376321335e-05, "TIME_S": 10.600306034088135, "TIME_S_1KI": 0.08379953543264715, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1235.8763800144193, "W": 78.10613100977226, "J_1KI": 9.770082690475741, "W_1KI": 0.6174592952328316, "W_D": 59.286131009772255, "J_D": 938.0867805147168, "W_D_1KI": 0.4686798871883084, "J_D_1KI": 0.0037050965025637837}

View File

@ -0,0 +1,14 @@
/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, 1, ..., 90000, 90000, 90002]),
col_indices=tensor([ 5326, 106, 329, ..., 882, 2232, 16085]),
values=tensor([1., 1., 1., ..., 1., 1., 1.]), size=(31379, 31379),
nnz=90002, layout=torch.sparse_csr)
tensor([0.5848, 0.1698, 0.4450, ..., 0.6169, 0.4741, 0.5274])
Matrix: as-caida_G_047
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 90002
Density: 9.140584376321335e-05
Time: 10.600306034088135 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 123716, "MATRIX_FILE": "as-caida_G_048", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 90820, "MATRIX_DENSITY": 9.223660285965907e-05, "TIME_S": 10.587847232818604, "TIME_S_1KI": 0.08558187488132984, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 737.9514087200165, "W": 72.62112727115486, "J_1KI": 5.964882543244338, "W_1KI": 0.5869986684919886, "W_D": 53.34112727115486, "J_D": 542.0345496082307, "W_D_1KI": 0.4311578718286629, "J_D_1KI": 0.003485061526630855}

View File

@ -0,0 +1,14 @@
/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, ..., 90816, 90816, 90820]),
col_indices=tensor([ 106, 329, 1040, ..., 3562, 13498, 16085]),
values=tensor([1., 1., 1., ..., 3., 3., 3.]), size=(31379, 31379),
nnz=90820, layout=torch.sparse_csr)
tensor([0.6448, 0.2976, 0.8562, ..., 0.0284, 0.6709, 0.2386])
Matrix: as-caida_G_048
Shape: torch.Size([31379, 31379])
Size: 984641641
NNZ: 90820
Density: 9.223660285965907e-05
Time: 10.587847232818604 seconds

View File

@ -0,0 +1 @@
{"CPU": "ALTRA", "ITERATIONS": 132716, "MATRIX_FILE": "as-caida_G_049", "MATRIX_SHAPE": [31379, 31379], "MATRIX_SIZE": 984641641, "MATRIX_NNZ": 90784, "MATRIX_DENSITY": 9.220004133463212e-05, "TIME_S": 12.987859964370728, "TIME_S_1KI": 0.09786205102904492, "BASELINE_TIME_S": 10, "BASELINE_DELAY_S": 10, "J": 1069.818884792328, "W": 76.0638705536987, "J_1KI": 8.060963898793876, "W_1KI": 0.573132633244663, "W_D": 57.34087055369871, "J_D": 806.484678497553, "W_D_1KI": 0.4320569528444099, "J_D_1KI": 0.003255500111850944}

Some files were not shown because too many files have changed in this diff Show More