POWER - FIRST TRY LIGHT PYTORCH

POW OPENSTK

HTML result view exported from: https://openbenchmarking.org/result/2406286-NE-POWERFIRS01&grs.

POWER - FIRST TRY LIGHT PYTORCHProcessorMotherboardChipsetMemoryDiskGraphicsNetworkOSKernelDisplay ServerDisplay DriverVulkanCompilerFile-SystemScreen ResolutionSystem LayerLIGHT PYTORCHAMD EPYC 7543P 32-Core (4 Cores / 8 Threads)Blade Shadow ShadowM v2.0 (1.1.3 BIOS)Intel 82G33/G31/P35/P31 + ICH91 x 16GB RAM-2400MT/s Blade 2MEVPUH6DW9W57-PYL215GB QEMU HDDRed Hat QXL paravirtual graphic card 20GBRed Hat Virtio deviceUbuntu 22.045.15.0-113-generic (x86_64)X ServerNVIDIA1.3.277GCC 11.4.0ext41280x800KVMOpenBenchmarking.org- Transparent Huge Pages: madvise- CPU Microcode: 0xa0011d1- Python 3.10.12- gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines; IBPB: conditional; IBRS_FW; STIBP: always-on; RSB filling; PBRSB-eIBRS: Not affected; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected

POWER - FIRST TRY LIGHT PYTORCHpytorch: CPU - 16 - Efficientnet_v2_lpytorch: CPU - 16 - ResNet-152pytorch: CPU - 16 - ResNet-50LIGHT PYTORCH3.875.6814.53OpenBenchmarking.org

PyTorch

Device: CPU - Batch Size: 16 - Model: Efficientnet_v2_l

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 16 - Model: Efficientnet_v2_lLIGHT PYTORCH0.87081.74162.61243.48324.354SE +/- 0.04, N = 63.87MIN: 3.58 / MAX: 4.09

PyTorch

Device: CPU - Batch Size: 16 - Model: ResNet-152

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 16 - Model: ResNet-152LIGHT PYTORCH1.2782.5563.8345.1126.39SE +/- 0.02, N = 35.68MIN: 5.42 / MAX: 5.78

PyTorch

Device: CPU - Batch Size: 16 - Model: ResNet-50

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 16 - Model: ResNet-50LIGHT PYTORCH48121620SE +/- 0.13, N = 1214.53MIN: 11.72 / MAX: 15.38

Phoronix Test Suite v10.8.5