pytorch-test

AMD Ryzen 7 5700X 8-Core testing with a MSI PRO B550M-P GEN3 (MS-7D95) v1.0 (1.60 BIOS) and NVIDIA GeForce RTX 4070 SUPER 12GB on Debian via the Phoronix Test Suite.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 2403306-NE-PYTORCHTE32
Jump To Table - Results

Statistics

Remove Outliers Before Calculating Averages

Graph Settings

Prefer Vertical Bar Graphs

Multi-Way Comparison

Condense Multi-Option Tests Into Single Result Graphs

Table

Show Detailed System Result Table

Run Management

Result
Identifier
Performance Per
Dollar
Date
Run
  Test
  Duration
All options
March 30
  3 Hours, 45 Minutes
Only show results matching title/arguments (delimit multiple options with a comma):
Do not show results matching title/arguments (delimit multiple options with a comma):

All options

Processor: AMD Ryzen 7 5700X 8-Core @ 3.40GHz (8 Cores / 16 Threads), Motherboard: MSI PRO B550M-P GEN3 (MS-7D95) v1.0 (1.60 BIOS), Chipset: AMD Starship/Matisse, Memory: 128GB, Disk: 256GB ARDOR GAMING m.2 NVME 256Gb AL1282 + 4001GB Seagate ST4000VX016-3CV1 + 512GB Apacer AS350 512, Graphics: NVIDIA GeForce RTX 4070 SUPER 12GB, Audio: NVIDIA Device 22bc, Monitor: PHL 242V8, Network: Realtek RTL8111/8168/8211/8411

OS: Debian, Kernel: 6.6.15-amd64 (x86_64), Desktop: Xfce 4.18, Display Server: X Server 1.21.1.11, Display Driver: NVIDIA 550.54.15, OpenGL: 4.6.0, OpenCL: OpenCL 3.0 CUDA 12.4.89 + OpenCL 3.0 PoCL 5.0+debian Linux +Asserts RELOC SPIR LLVM 16.0.6 SLEEF DISTRO POCL_DEBUG, Compiler: GCC 13.2.0 + CUDA 12.0, File-System: ext4, Screen Resolution: 4480x1440

Kernel Notes: Transparent Huge Pages: always
Processor Notes: Scaling Governor: acpi-cpufreq schedutil (Boost: Enabled) - CPU Microcode: 0xa20120e
Python Notes: Python 3.11.8
Security Notes: 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: Vulnerable + spec_store_bypass: Vulnerable + spectre_v1: Vulnerable: __user pointer sanitization and usercopy barriers only; no swapgs barriers + spectre_v2: Vulnerable IBPB: disabled STIBP: disabled PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

pytorch-testOpenBenchmarking.orgPhoronix Test SuiteAMD Ryzen 7 5700X 8-Core @ 3.40GHz (8 Cores / 16 Threads)MSI PRO B550M-P GEN3 (MS-7D95) v1.0 (1.60 BIOS)AMD Starship/Matisse128GB256GB ARDOR GAMING m.2 NVME 256Gb AL1282 + 4001GB Seagate ST4000VX016-3CV1 + 512GB Apacer AS350 512NVIDIA GeForce RTX 4070 SUPER 12GBNVIDIA Device 22bcPHL 242V8Realtek RTL8111/8168/8211/8411Debian6.6.15-amd64 (x86_64)Xfce 4.18X Server 1.21.1.11NVIDIA 550.54.154.6.0OpenCL 3.0 CUDA 12.4.89 + OpenCL 3.0 PoCL 5.0+debian Linux +Asserts RELOC SPIR LLVM 16.0.6 SLEEF DISTRO POCL_DEBUGGCC 13.2.0 + CUDA 12.0ext44480x1440ProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerDisplay DriverOpenGLOpenCLCompilerFile-SystemScreen ResolutionPytorch-test BenchmarksSystem Logs- Transparent Huge Pages: always- Scaling Governor: acpi-cpufreq schedutil (Boost: Enabled) - CPU Microcode: 0xa20120e- Python 3.11.8- 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: Vulnerable + spec_store_bypass: Vulnerable + spectre_v1: Vulnerable: __user pointer sanitization and usercopy barriers only; no swapgs barriers + spectre_v2: Vulnerable IBPB: disabled STIBP: disabled PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

pytorch-testpytorch: CPU - 1 - ResNet-50pytorch: CPU - 1 - ResNet-152pytorch: CPU - 16 - ResNet-50pytorch: CPU - 32 - ResNet-50pytorch: CPU - 64 - ResNet-50pytorch: CPU - 16 - ResNet-152pytorch: CPU - 256 - ResNet-50pytorch: CPU - 32 - ResNet-152pytorch: CPU - 512 - ResNet-50pytorch: CPU - 64 - ResNet-152pytorch: CPU - 256 - ResNet-152pytorch: CPU - 512 - ResNet-152pytorch: CPU - 1 - Efficientnet_v2_lpytorch: CPU - 16 - Efficientnet_v2_lpytorch: CPU - 32 - Efficientnet_v2_lpytorch: CPU - 64 - Efficientnet_v2_lpytorch: CPU - 256 - Efficientnet_v2_lpytorch: CPU - 512 - Efficientnet_v2_lpytorch: NVIDIA CUDA GPU - 1 - ResNet-50pytorch: NVIDIA CUDA GPU - 1 - ResNet-152pytorch: NVIDIA CUDA GPU - 16 - ResNet-50pytorch: NVIDIA CUDA GPU - 32 - ResNet-50pytorch: NVIDIA CUDA GPU - 64 - ResNet-50pytorch: NVIDIA CUDA GPU - 16 - ResNet-152pytorch: NVIDIA CUDA GPU - 256 - ResNet-50pytorch: NVIDIA CUDA GPU - 32 - ResNet-152pytorch: NVIDIA CUDA GPU - 512 - ResNet-50pytorch: NVIDIA CUDA GPU - 64 - ResNet-152pytorch: NVIDIA CUDA GPU - 256 - ResNet-152pytorch: NVIDIA CUDA GPU - 512 - ResNet-152pytorch: NVIDIA CUDA GPU - 1 - Efficientnet_v2_lpytorch: NVIDIA CUDA GPU - 16 - Efficientnet_v2_lpytorch: NVIDIA CUDA GPU - 32 - Efficientnet_v2_lpytorch: NVIDIA CUDA GPU - 64 - Efficientnet_v2_lpytorch: NVIDIA CUDA GPU - 256 - Efficientnet_v2_lpytorch: NVIDIA CUDA GPU - 512 - Efficientnet_v2_lAll options44.6018.1628.4728.3327.7711.4727.8111.5227.9411.5011.4111.3210.987.747.617.607.507.44212.3176.01214.58217.50217.2273.74212.8976.28215.1075.0376.7677.1240.7338.8139.0339.0039.6537.97OpenBenchmarking.org

PyTorch

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 1 - Model: ResNet-50All options1020304050SE +/- 0.50, N = 344.60MIN: 31.6 / MAX: 46.25

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 1 - Model: ResNet-152All options48121620SE +/- 0.12, N = 318.16MIN: 14.49 / MAX: 18.57

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 16 - Model: ResNet-50All options714212835SE +/- 0.32, N = 528.47MIN: 19.33 / MAX: 30.1

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 32 - Model: ResNet-50All options714212835SE +/- 0.29, N = 328.33MIN: 19.28 / MAX: 29.17

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 64 - Model: ResNet-50All options714212835SE +/- 0.21, N = 1027.77MIN: 16.01 / MAX: 29.49

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 16 - Model: ResNet-152All options3691215SE +/- 0.07, N = 311.47MIN: 8.19 / MAX: 11.89

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 256 - Model: ResNet-50All options714212835SE +/- 0.22, N = 1527.81MIN: 12.38 / MAX: 30.18

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 32 - Model: ResNet-152All options3691215SE +/- 0.02, N = 311.52MIN: 9.36 / MAX: 11.86

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 512 - Model: ResNet-50All options714212835SE +/- 0.32, N = 1527.94MIN: 16.81 / MAX: 29.62

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 64 - Model: ResNet-152All options3691215SE +/- 0.10, N = 311.50MIN: 8.81 / MAX: 11.87

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 256 - Model: ResNet-152All options3691215SE +/- 0.02, N = 311.41MIN: 8.11 / MAX: 11.89

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 512 - Model: ResNet-152All options3691215SE +/- 0.10, N = 311.32MIN: 7.82 / MAX: 12.05

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 1 - Model: Efficientnet_v2_lAll options3691215SE +/- 0.04, N = 310.98MIN: 8.22 / MAX: 11.42

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 16 - Model: Efficientnet_v2_lAll options246810SE +/- 0.01, N = 37.74MIN: 5.95 / MAX: 7.85

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 32 - Model: Efficientnet_v2_lAll options246810SE +/- 0.10, N = 37.61MIN: 5.99 / MAX: 7.8

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 64 - Model: Efficientnet_v2_lAll options246810SE +/- 0.05, N = 37.60MIN: 5.08 / MAX: 7.86

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 256 - Model: Efficientnet_v2_lAll options246810SE +/- 0.09, N = 37.50MIN: 5.5 / MAX: 7.79

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 512 - Model: Efficientnet_v2_lAll options246810SE +/- 0.07, N = 97.44MIN: 4.46 / MAX: 7.85

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: NVIDIA CUDA GPU - Batch Size: 1 - Model: ResNet-50All options50100150200250SE +/- 2.38, N = 3212.31MIN: 110.61 / MAX: 225.39

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: NVIDIA CUDA GPU - Batch Size: 1 - Model: ResNet-152All options20406080100SE +/- 0.73, N = 376.01MIN: 44.66 / MAX: 78.91

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: NVIDIA CUDA GPU - Batch Size: 16 - Model: ResNet-50All options50100150200250SE +/- 2.06, N = 3214.58MIN: 172.29 / MAX: 221

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: NVIDIA CUDA GPU - Batch Size: 32 - Model: ResNet-50All options50100150200250SE +/- 0.92, N = 3217.50MIN: 172.82 / MAX: 221.5

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: NVIDIA CUDA GPU - Batch Size: 64 - Model: ResNet-50All options50100150200250SE +/- 0.25, N = 3217.22MIN: 199.94 / MAX: 220.82

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: NVIDIA CUDA GPU - Batch Size: 16 - Model: ResNet-152All options1632486480SE +/- 0.08, N = 373.74MIN: 43.03 / MAX: 79.24

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: NVIDIA CUDA GPU - Batch Size: 256 - Model: ResNet-50All options50100150200250SE +/- 2.38, N = 4212.89MIN: 127.22 / MAX: 220.8

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: NVIDIA CUDA GPU - Batch Size: 32 - Model: ResNet-152All options20406080100SE +/- 0.91, N = 376.28MIN: 39.08 / MAX: 78.37

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: NVIDIA CUDA GPU - Batch Size: 512 - Model: ResNet-50All options50100150200250SE +/- 2.52, N = 3215.10MIN: 122.92 / MAX: 222.21

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: NVIDIA CUDA GPU - Batch Size: 64 - Model: ResNet-152All options20406080100SE +/- 0.79, N = 575.03MIN: 50.37 / MAX: 78.56

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: NVIDIA CUDA GPU - Batch Size: 256 - Model: ResNet-152All options20406080100SE +/- 0.39, N = 376.76MIN: 58.4 / MAX: 78.75

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: NVIDIA CUDA GPU - Batch Size: 512 - Model: ResNet-152All options20406080100SE +/- 0.08, N = 377.12MIN: 69.63 / MAX: 78.12

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: NVIDIA CUDA GPU - Batch Size: 1 - Model: Efficientnet_v2_lAll options918273645SE +/- 0.08, N = 340.73MIN: 37.27 / MAX: 41.27

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: NVIDIA CUDA GPU - Batch Size: 16 - Model: Efficientnet_v2_lAll options918273645SE +/- 0.12, N = 338.81MIN: 33.4 / MAX: 40.1

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: NVIDIA CUDA GPU - Batch Size: 32 - Model: Efficientnet_v2_lAll options918273645SE +/- 0.56, N = 339.03MIN: 35.91 / MAX: 40.16

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: NVIDIA CUDA GPU - Batch Size: 64 - Model: Efficientnet_v2_lAll options918273645SE +/- 0.29, N = 339.00MIN: 36.15 / MAX: 40

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: NVIDIA CUDA GPU - Batch Size: 256 - Model: Efficientnet_v2_lAll options918273645SE +/- 0.05, N = 339.65MIN: 36.18 / MAX: 40.35

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: NVIDIA CUDA GPU - Batch Size: 512 - Model: Efficientnet_v2_lAll options918273645SE +/- 0.41, N = 537.97MIN: 24.02 / MAX: 39.94

All options

Processor: AMD Ryzen 7 5700X 8-Core @ 3.40GHz (8 Cores / 16 Threads), Motherboard: MSI PRO B550M-P GEN3 (MS-7D95) v1.0 (1.60 BIOS), Chipset: AMD Starship/Matisse, Memory: 128GB, Disk: 256GB ARDOR GAMING m.2 NVME 256Gb AL1282 + 4001GB Seagate ST4000VX016-3CV1 + 512GB Apacer AS350 512, Graphics: NVIDIA GeForce RTX 4070 SUPER 12GB, Audio: NVIDIA Device 22bc, Monitor: PHL 242V8, Network: Realtek RTL8111/8168/8211/8411

OS: Debian, Kernel: 6.6.15-amd64 (x86_64), Desktop: Xfce 4.18, Display Server: X Server 1.21.1.11, Display Driver: NVIDIA 550.54.15, OpenGL: 4.6.0, OpenCL: OpenCL 3.0 CUDA 12.4.89 + OpenCL 3.0 PoCL 5.0+debian Linux +Asserts RELOC SPIR LLVM 16.0.6 SLEEF DISTRO POCL_DEBUG, Compiler: GCC 13.2.0 + CUDA 12.0, File-System: ext4, Screen Resolution: 4480x1440

Kernel Notes: Transparent Huge Pages: always
Processor Notes: Scaling Governor: acpi-cpufreq schedutil (Boost: Enabled) - CPU Microcode: 0xa20120e
Python Notes: Python 3.11.8
Security Notes: 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: Vulnerable + spec_store_bypass: Vulnerable + spectre_v1: Vulnerable: __user pointer sanitization and usercopy barriers only; no swapgs barriers + spectre_v2: Vulnerable IBPB: disabled STIBP: disabled PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected

Testing initiated at 30 March 2024 10:45 by user serge.