2024-12-04-1617

AMD Ryzen 9 7950X 16-Core testing with a MSI MPG X670E CARBON WIFI (MS-7D70) v1.0 (1.M1 BIOS) and NVIDIA GeForce RTX 4090 24GB on Ubuntu 24.04 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 2412063-NOKI-202412076
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
AMD Ryzen 9 7950X 16-Core
December 04
  2 Days, 8 Hours, 52 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):

AMD Ryzen 9 7950X 16-Core

Processor: AMD Ryzen 9 7950X 16-Core @ 5.88GHz (16 Cores / 32 Threads), Motherboard: MSI MPG X670E CARBON WIFI (MS-7D70) v1.0 (1.M1 BIOS), Chipset: AMD Device 14d8, Memory: 4 x 48 GB DDR5-3600MT/s Kingston KF560C32-48, Disk: 4 x 2000GB Samsung SSD 980 PRO 2TB + 2 x 4001GB Samsung SSD 870 + 62GB USB DISK 2.0, Graphics: NVIDIA GeForce RTX 4090 24GB, Audio: NVIDIA AD102 HD Audio, Monitor: BenQ GW2270, Network: Realtek RTL8125 2.5GbE + MEDIATEK MT7922 802.11ax PCI

OS: Ubuntu 24.04, Kernel: 6.8.0-41-generic (x86_64), Desktop: GNOME Shell 46.0, Display Server: X Server 1.21.1.11, Display Driver: nouveau, Compiler: GCC 13.2.0, File-System: tmpfs, Screen Resolution: 1920x1080

Kernel Notes: Transparent Huge Pages: madvise
Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-uJ7kn6/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-uJ7kn6/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v
Processor Notes: Scaling Governor: amd-pstate-epp powersave (EPP: balance_performance) - CPU Microcode: 0xa601209
Python Notes: Python 3.12.3
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 + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS; IBPB: conditional; STIBP: always-on; RSB filling; PBRSB-eIBRS: Not affected; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected

2024-12-04-1617OpenBenchmarking.orgPhoronix Test SuiteAMD Ryzen 9 7950X 16-Core @ 5.88GHz (16 Cores / 32 Threads)MSI MPG X670E CARBON WIFI (MS-7D70) v1.0 (1.M1 BIOS)AMD Device 14d84 x 48 GB DDR5-3600MT/s Kingston KF560C32-484 x 2000GB Samsung SSD 980 PRO 2TB + 2 x 4001GB Samsung SSD 870 + 62GB USB DISK 2.0NVIDIA GeForce RTX 4090 24GBNVIDIA AD102 HD AudioBenQ GW2270Realtek RTL8125 2.5GbE + MEDIATEK MT7922 802.11ax PCIUbuntu 24.046.8.0-41-generic (x86_64)GNOME Shell 46.0X Server 1.21.1.11nouveauGCC 13.2.0tmpfs1920x1080ProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerDisplay DriverCompilerFile-SystemScreen Resolution2024-12-04-1617 BenchmarksSystem Logs- Transparent Huge Pages: madvise- --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-uJ7kn6/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-uJ7kn6/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v - Scaling Governor: amd-pstate-epp powersave (EPP: balance_performance) - CPU Microcode: 0xa601209- Python 3.12.3- gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS; IBPB: conditional; STIBP: always-on; RSB filling; PBRSB-eIBRS: Not affected; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected

2024-12-04-1617tensorflow: GPU - 512 - VGG-16tensorflow: GPU - 256 - VGG-16tensorflow: GPU - 512 - ResNet-50tensorflow: GPU - 256 - ResNet-50tensorflow: CPU - 512 - VGG-16tensorflow: GPU - 64 - VGG-16tensorflow: GPU - 512 - GoogLeNettensorflow: CPU - 512 - ResNet-50tensorflow: CPU - 256 - VGG-16tensorflow: GPU - 32 - VGG-16tensorflow: GPU - 512 - AlexNettensorflow: GPU - 256 - GoogLeNettensorflow: CPU - 256 - ResNet-50tensorflow: GPU - 64 - ResNet-50whisper-cpp: ggml-medium.en - 2016 State of the Uniontensorflow: GPU - 16 - VGG-16tensorflow: GPU - 256 - AlexNettensorflow: CPU - 512 - GoogLeNetbuild-llvm: Unix Makefilestensorflow: GPU - 32 - ResNet-50build-llvm: Ninjatensorflow: CPU - 64 - VGG-16whisper-cpp: ggml-small.en - 2016 State of the Uniontensorflow: CPU - 256 - GoogLeNettensorflow: GPU - 64 - GoogLeNettensorflow: CPU - 64 - ResNet-50tensorflow: CPU - 32 - VGG-16tensorflow: GPU - 16 - ResNet-50ffmpeg: libx264 - Uploadpytorch: CPU - 64 - Efficientnet_v2_lpytorch: CPU - 32 - Efficientnet_v2_lpytorch: CPU - 256 - Efficientnet_v2_lpytorch: CPU - 16 - Efficientnet_v2_lpytorch: CPU - 512 - Efficientnet_v2_ltensorflow: GPU - 64 - AlexNetopenvino-genai: Gemma-7b-int4-ov - CPU - Time Per Output Tokenopenvino-genai: Gemma-7b-int4-ov - CPU - Time To First Tokenopenvino-genai: Gemma-7b-int4-ov - CPUtensorflow: CPU - 512 - AlexNetffmpeg: libx264 - Platformffmpeg: libx264 - Video On Demandtensorflow: GPU - 32 - GoogLeNetffmpeg: libx265 - Platformffmpeg: libx265 - Video On Demandffmpeg: libx265 - Uploadopenvino-genai: Falcon-7b-instruct-int4-ov - CPU - Time Per Output Tokenopenvino-genai: Falcon-7b-instruct-int4-ov - CPU - Time To First Tokenopenvino-genai: Falcon-7b-instruct-int4-ov - CPUtensorflow: CPU - 16 - VGG-16pytorch: CPU - 64 - ResNet-152tensorflow: CPU - 32 - ResNet-50pytorch: CPU - 32 - ResNet-152pytorch: CPU - 256 - ResNet-152pytorch: CPU - 512 - ResNet-152pytorch: CPU - 16 - ResNet-152whisper-cpp: ggml-base.en - 2016 State of the Unionpytorch: CPU - 1 - Efficientnet_v2_ltensorflow: GPU - 32 - AlexNetpytorch: CPU - 16 - ResNet-50tensorflow: CPU - 256 - AlexNettensorflow: GPU - 16 - GoogLeNettensorflow: CPU - 64 - GoogLeNetjohn-the-ripper: MD5john-the-ripper: HMAC-SHA512openvino-genai: TinyLlama-1.1B-Chat-v1.0 - CPU - Time Per Output Tokenopenvino-genai: TinyLlama-1.1B-Chat-v1.0 - CPU - Time To First Tokenopenvino-genai: TinyLlama-1.1B-Chat-v1.0 - CPUtensorflow: GPU - 1 - VGG-16tensorflow: CPU - 16 - ResNet-50pytorch: CPU - 32 - ResNet-50tensorflow: GPU - 16 - AlexNetpytorch: CPU - 1 - ResNet-152pytorch: CPU - 256 - ResNet-50pytorch: CPU - 512 - ResNet-50pytorch: CPU - 64 - ResNet-50ffmpeg: libx265 - Liveopenvino-genai: Phi-3-mini-128k-instruct-int4-ov - CPU - Time Per Output Tokenopenvino-genai: Phi-3-mini-128k-instruct-int4-ov - CPU - Time To First Tokenopenvino-genai: Phi-3-mini-128k-instruct-int4-ov - CPUtensorflow: CPU - 32 - GoogLeNetffmpeg: libx264 - Livejohn-the-ripper: WPA PSKjohn-the-ripper: bcryptjohn-the-ripper: Blowfishtensorflow: CPU - 1 - VGG-16tensorflow: CPU - 64 - AlexNetcompress-7zip: Decompression Ratingcompress-7zip: Compression Ratingcompress-7zip: Decompression Ratingcompress-7zip: Compression Ratingpytorch: CPU - 1 - ResNet-50tensorflow: GPU - 1 - ResNet-50tensorflow: CPU - 32 - AlexNetsvt-av1: Preset 4 - Bosphorus 4Ktensorflow: CPU - 16 - GoogLeNettensorflow: CPU - 16 - AlexNettensorflow: GPU - 1 - AlexNettensorflow: CPU - 1 - ResNet-50tensorflow: CPU - 1 - AlexNetsvt-av1: Preset 8 - Bosphorus 4Ktensorflow: GPU - 1 - GoogLeNetsvt-av1: Preset 4 - Bosphorus 1080ptensorflow: CPU - 1 - GoogLeNetsvt-av1: Preset 8 - Bosphorus 1080pAMD Ryzen 9 7950X 16-Core2.652.649.019.0119.252.6128.5033.5519.192.5850.3628.4333.439.04753.885132.5350.07110.87414.7529.01386.66418.61259.85652110.7428.2233.6617.768.9717.4510.7210.6810.7110.8010.9048.15126.70135.947.89403.5065.3665.6328.2269.2369.3234.3695.68106.1810.4516.4517.8033.8318.0918.2218.2618.2492.6086713.9245.4243.41384.1027.83113.64425233318272800053.0654.6318.852.0933.2642.8940.4325.8943.1143.3343.16180.1763.7369.3315.69116.32286.8816357144266441744.24290.1216278114784217093415023862.936.70210.4510.805118.82134.4911.8812.8911.5481.82424.2528.94651.20257.824OpenBenchmarking.org

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: GPU - Batch Size: 512 - Model: VGG-16AMD Ryzen 9 7950X 16-Core0.59631.19261.78892.38522.9815SE +/- 0.00, N = 32.65

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: GPU - Batch Size: 256 - Model: VGG-16AMD Ryzen 9 7950X 16-Core0.5941.1881.7822.3762.97SE +/- 0.01, N = 32.64

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: GPU - Batch Size: 512 - Model: ResNet-50AMD Ryzen 9 7950X 16-Core3691215SE +/- 0.04, N = 39.01

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: GPU - Batch Size: 256 - Model: ResNet-50AMD Ryzen 9 7950X 16-Core3691215SE +/- 0.01, N = 39.01

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 512 - Model: VGG-16AMD Ryzen 9 7950X 16-Core510152025SE +/- 0.02, N = 319.25

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: GPU - Batch Size: 64 - Model: VGG-16AMD Ryzen 9 7950X 16-Core0.58731.17461.76192.34922.9365SE +/- 0.00, N = 32.61

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: GPU - Batch Size: 512 - Model: GoogLeNetAMD Ryzen 9 7950X 16-Core714212835SE +/- 0.01, N = 328.50

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 512 - Model: ResNet-50AMD Ryzen 9 7950X 16-Core816243240SE +/- 0.02, N = 333.55

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 256 - Model: VGG-16AMD Ryzen 9 7950X 16-Core510152025SE +/- 0.01, N = 319.19

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: GPU - Batch Size: 32 - Model: VGG-16AMD Ryzen 9 7950X 16-Core0.58051.1611.74152.3222.9025SE +/- 0.01, N = 32.58

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: GPU - Batch Size: 512 - Model: AlexNetAMD Ryzen 9 7950X 16-Core1122334455SE +/- 0.05, N = 350.36

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: GPU - Batch Size: 256 - Model: GoogLeNetAMD Ryzen 9 7950X 16-Core714212835SE +/- 0.05, N = 328.43

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 256 - Model: ResNet-50AMD Ryzen 9 7950X 16-Core816243240SE +/- 0.00, N = 333.43

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: GPU - Batch Size: 64 - Model: ResNet-50AMD Ryzen 9 7950X 16-Core3691215SE +/- 0.01, N = 39.04

Whisper.cpp

OpenBenchmarking.orgSeconds, Fewer Is BetterWhisper.cpp 1.6.2Model: ggml-medium.en - Input: 2016 State of the UnionAMD Ryzen 9 7950X 16-Core160320480640800SE +/- 1.36, N = 3753.891. (CXX) g++ options: -O3 -std=c++11 -fPIC -pthread -msse3 -mssse3 -mavx -mf16c -mfma -mavx2 -mavx512f -mavx512cd -mavx512vl -mavx512dq -mavx512bw -mavx512vbmi -mavx512vnni

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: GPU - Batch Size: 16 - Model: VGG-16AMD Ryzen 9 7950X 16-Core0.56931.13861.70792.27722.8465SE +/- 0.01, N = 32.53

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: GPU - Batch Size: 256 - Model: AlexNetAMD Ryzen 9 7950X 16-Core1122334455SE +/- 0.09, N = 350.07

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 512 - Model: GoogLeNetAMD Ryzen 9 7950X 16-Core20406080100SE +/- 0.02, N = 3110.87

Timed LLVM Compilation

This test times how long it takes to compile/build the LLVM compiler stack. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed LLVM Compilation 16.0Build System: Unix MakefilesAMD Ryzen 9 7950X 16-Core90180270360450SE +/- 0.08, N = 3414.75

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: GPU - Batch Size: 32 - Model: ResNet-50AMD Ryzen 9 7950X 16-Core3691215SE +/- 0.00, N = 39.01

Timed LLVM Compilation

This test times how long it takes to compile/build the LLVM compiler stack. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed LLVM Compilation 16.0Build System: NinjaAMD Ryzen 9 7950X 16-Core80160240320400SE +/- 0.30, N = 3386.66

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 64 - Model: VGG-16AMD Ryzen 9 7950X 16-Core510152025SE +/- 0.00, N = 318.61

Whisper.cpp

OpenBenchmarking.orgSeconds, Fewer Is BetterWhisper.cpp 1.6.2Model: ggml-small.en - Input: 2016 State of the UnionAMD Ryzen 9 7950X 16-Core60120180240300SE +/- 0.32, N = 3259.861. (CXX) g++ options: -O3 -std=c++11 -fPIC -pthread -msse3 -mssse3 -mavx -mf16c -mfma -mavx2 -mavx512f -mavx512cd -mavx512vl -mavx512dq -mavx512bw -mavx512vbmi -mavx512vnni

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 256 - Model: GoogLeNetAMD Ryzen 9 7950X 16-Core20406080100SE +/- 0.05, N = 3110.74

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: GPU - Batch Size: 64 - Model: GoogLeNetAMD Ryzen 9 7950X 16-Core714212835SE +/- 0.04, N = 328.22

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 64 - Model: ResNet-50AMD Ryzen 9 7950X 16-Core816243240SE +/- 0.01, N = 333.66

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 32 - Model: VGG-16AMD Ryzen 9 7950X 16-Core48121620SE +/- 0.01, N = 317.76

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: GPU - Batch Size: 16 - Model: ResNet-50AMD Ryzen 9 7950X 16-Core3691215SE +/- 0.01, N = 38.97

FFmpeg

This is a benchmark of the FFmpeg multimedia framework. The FFmpeg test profile is making use of a modified version of vbench from Columbia University's Architecture and Design Lab (ARCADE) [http://arcade.cs.columbia.edu/vbench/] that is a benchmark for video-as-a-service workloads. The test profile offers the options of a range of vbench scenarios based on freely distributable video content and offers the options of using the x264 or x265 video encoders for transcoding. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterFFmpeg 7.0Encoder: libx264 - Scenario: UploadAMD Ryzen 9 7950X 16-Core48121620SE +/- 0.07, N = 317.451. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl

PyTorch

This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 64 - Model: Efficientnet_v2_lAMD Ryzen 9 7950X 16-Core3691215SE +/- 0.09, N = 310.72MIN: 8.64 / MAX: 11.15

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 32 - Model: Efficientnet_v2_lAMD Ryzen 9 7950X 16-Core3691215SE +/- 0.01, N = 310.68MIN: 8.73 / MAX: 11.11

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 256 - Model: Efficientnet_v2_lAMD Ryzen 9 7950X 16-Core3691215SE +/- 0.01, N = 310.71MIN: 8.83 / MAX: 11.09

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 16 - Model: Efficientnet_v2_lAMD Ryzen 9 7950X 16-Core3691215SE +/- 0.15, N = 310.80MIN: 8.55 / MAX: 11.28

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 512 - Model: Efficientnet_v2_lAMD Ryzen 9 7950X 16-Core3691215SE +/- 0.03, N = 310.90MIN: 9.08 / MAX: 11.2

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: GPU - Batch Size: 64 - Model: AlexNetAMD Ryzen 9 7950X 16-Core1122334455SE +/- 0.14, N = 348.15

OpenVINO GenAI

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: Gemma-7b-int4-ov - Device: CPU - Time Per Output TokenAMD Ryzen 9 7950X 16-Core306090120150SE +/- 0.17, N = 3126.70

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: Gemma-7b-int4-ov - Device: CPU - Time To First TokenAMD Ryzen 9 7950X 16-Core306090120150SE +/- 0.25, N = 3135.94

OpenBenchmarking.orgtokens/s, More Is BetterOpenVINO GenAI 2024.5Model: Gemma-7b-int4-ov - Device: CPUAMD Ryzen 9 7950X 16-Core246810SE +/- 0.01, N = 37.89
OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: Gemma-7b-int4-ov - Device: CPU - Time To First TokenAMD Ryzen 9 7950X 16-Core306090120150SE +/- 0.25, N = 3135.94
OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: Gemma-7b-int4-ov - Device: CPU - Time Per Output TokenAMD Ryzen 9 7950X 16-Core306090120150SE +/- 0.17, N = 3126.70

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 512 - Model: AlexNetAMD Ryzen 9 7950X 16-Core90180270360450SE +/- 0.08, N = 3403.50

FFmpeg

This is a benchmark of the FFmpeg multimedia framework. The FFmpeg test profile is making use of a modified version of vbench from Columbia University's Architecture and Design Lab (ARCADE) [http://arcade.cs.columbia.edu/vbench/] that is a benchmark for video-as-a-service workloads. The test profile offers the options of a range of vbench scenarios based on freely distributable video content and offers the options of using the x264 or x265 video encoders for transcoding. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterFFmpeg 7.0Encoder: libx264 - Scenario: PlatformAMD Ryzen 9 7950X 16-Core1530456075SE +/- 0.11, N = 365.361. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl

OpenBenchmarking.orgFPS, More Is BetterFFmpeg 7.0Encoder: libx264 - Scenario: Video On DemandAMD Ryzen 9 7950X 16-Core1530456075SE +/- 0.39, N = 365.631. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: GPU - Batch Size: 32 - Model: GoogLeNetAMD Ryzen 9 7950X 16-Core714212835SE +/- 0.08, N = 328.22

FFmpeg

This is a benchmark of the FFmpeg multimedia framework. The FFmpeg test profile is making use of a modified version of vbench from Columbia University's Architecture and Design Lab (ARCADE) [http://arcade.cs.columbia.edu/vbench/] that is a benchmark for video-as-a-service workloads. The test profile offers the options of a range of vbench scenarios based on freely distributable video content and offers the options of using the x264 or x265 video encoders for transcoding. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterFFmpeg 7.0Encoder: libx265 - Scenario: PlatformAMD Ryzen 9 7950X 16-Core1530456075SE +/- 0.10, N = 369.231. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl

OpenBenchmarking.orgFPS, More Is BetterFFmpeg 7.0Encoder: libx265 - Scenario: Video On DemandAMD Ryzen 9 7950X 16-Core1530456075SE +/- 0.23, N = 369.321. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl

OpenBenchmarking.orgFPS, More Is BetterFFmpeg 7.0Encoder: libx265 - Scenario: UploadAMD Ryzen 9 7950X 16-Core816243240SE +/- 0.12, N = 334.361. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl

OpenVINO GenAI

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: Falcon-7b-instruct-int4-ov - Device: CPU - Time Per Output TokenAMD Ryzen 9 7950X 16-Core20406080100SE +/- 0.06, N = 395.68

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: Falcon-7b-instruct-int4-ov - Device: CPU - Time To First TokenAMD Ryzen 9 7950X 16-Core20406080100SE +/- 0.05, N = 3106.18

OpenBenchmarking.orgtokens/s, More Is BetterOpenVINO GenAI 2024.5Model: Falcon-7b-instruct-int4-ov - Device: CPUAMD Ryzen 9 7950X 16-Core3691215SE +/- 0.01, N = 310.45
OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: Falcon-7b-instruct-int4-ov - Device: CPU - Time To First TokenAMD Ryzen 9 7950X 16-Core20406080100SE +/- 0.05, N = 3106.18
OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: Falcon-7b-instruct-int4-ov - Device: CPU - Time Per Output TokenAMD Ryzen 9 7950X 16-Core20406080100SE +/- 0.06, N = 395.68

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 16 - Model: VGG-16AMD Ryzen 9 7950X 16-Core48121620SE +/- 0.01, N = 316.45

PyTorch

This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 64 - Model: ResNet-152AMD Ryzen 9 7950X 16-Core48121620SE +/- 0.05, N = 317.80MIN: 17.31 / MAX: 18.07

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 32 - Model: ResNet-50AMD Ryzen 9 7950X 16-Core816243240SE +/- 0.01, N = 333.83

PyTorch

This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 32 - Model: ResNet-152AMD Ryzen 9 7950X 16-Core48121620SE +/- 0.15, N = 318.09MIN: 17.52 / MAX: 18.48

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 256 - Model: ResNet-152AMD Ryzen 9 7950X 16-Core48121620SE +/- 0.14, N = 318.22MIN: 17.5 / MAX: 18.6

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 512 - Model: ResNet-152AMD Ryzen 9 7950X 16-Core48121620SE +/- 0.11, N = 318.26MIN: 17.68 / MAX: 18.59

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 16 - Model: ResNet-152AMD Ryzen 9 7950X 16-Core48121620SE +/- 0.04, N = 318.24MIN: 17.72 / MAX: 18.43

Whisper.cpp

OpenBenchmarking.orgSeconds, Fewer Is BetterWhisper.cpp 1.6.2Model: ggml-base.en - Input: 2016 State of the UnionAMD Ryzen 9 7950X 16-Core20406080100SE +/- 0.23, N = 392.611. (CXX) g++ options: -O3 -std=c++11 -fPIC -pthread -msse3 -mssse3 -mavx -mf16c -mfma -mavx2 -mavx512f -mavx512cd -mavx512vl -mavx512dq -mavx512bw -mavx512vbmi -mavx512vnni

PyTorch

This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 1 - Model: Efficientnet_v2_lAMD Ryzen 9 7950X 16-Core48121620SE +/- 0.11, N = 313.92MIN: 13.59 / MAX: 14.24

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: GPU - Batch Size: 32 - Model: AlexNetAMD Ryzen 9 7950X 16-Core1020304050SE +/- 0.05, N = 345.42

PyTorch

This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 16 - Model: ResNet-50AMD Ryzen 9 7950X 16-Core1020304050SE +/- 0.48, N = 543.41MIN: 39.89 / MAX: 44.61

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 256 - Model: AlexNetAMD Ryzen 9 7950X 16-Core80160240320400SE +/- 0.17, N = 3384.10

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: GPU - Batch Size: 16 - Model: GoogLeNetAMD Ryzen 9 7950X 16-Core714212835SE +/- 0.09, N = 327.83

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 64 - Model: GoogLeNetAMD Ryzen 9 7950X 16-Core306090120150SE +/- 0.02, N = 3113.64

John The Ripper

This is a benchmark of John The Ripper, which is a password cracker. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: MD5AMD Ryzen 9 7950X 16-Core900K1800K2700K3600K4500KSE +/- 7218.80, N = 342523331. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lm -lrt -lz -ldl -lcrypt

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: HMAC-SHA512AMD Ryzen 9 7950X 16-Core40M80M120M160M200MSE +/- 430001.55, N = 31827280001. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lm -lrt -lz -ldl -lcrypt

OpenVINO GenAI

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: TinyLlama-1.1B-Chat-v1.0 - Device: CPU - Time Per Output TokenAMD Ryzen 9 7950X 16-Core1224364860SE +/- 0.02, N = 353.06

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: TinyLlama-1.1B-Chat-v1.0 - Device: CPU - Time To First TokenAMD Ryzen 9 7950X 16-Core1224364860SE +/- 0.07, N = 354.63

OpenBenchmarking.orgtokens/s, More Is BetterOpenVINO GenAI 2024.5Model: TinyLlama-1.1B-Chat-v1.0 - Device: CPUAMD Ryzen 9 7950X 16-Core510152025SE +/- 0.01, N = 318.85
OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: TinyLlama-1.1B-Chat-v1.0 - Device: CPU - Time To First TokenAMD Ryzen 9 7950X 16-Core1224364860SE +/- 0.07, N = 354.63
OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: TinyLlama-1.1B-Chat-v1.0 - Device: CPU - Time Per Output TokenAMD Ryzen 9 7950X 16-Core1224364860SE +/- 0.02, N = 353.06

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: GPU - Batch Size: 1 - Model: VGG-16AMD Ryzen 9 7950X 16-Core0.47030.94061.41091.88122.3515SE +/- 0.01, N = 32.09

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 16 - Model: ResNet-50AMD Ryzen 9 7950X 16-Core816243240SE +/- 0.02, N = 333.26

PyTorch

This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 32 - Model: ResNet-50AMD Ryzen 9 7950X 16-Core1020304050SE +/- 0.25, N = 342.89MIN: 40.6 / MAX: 43.82

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: GPU - Batch Size: 16 - Model: AlexNetAMD Ryzen 9 7950X 16-Core918273645SE +/- 0.09, N = 340.43

PyTorch

This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 1 - Model: ResNet-152AMD Ryzen 9 7950X 16-Core612182430SE +/- 0.21, N = 325.89MIN: 22.3 / MAX: 26.4

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 256 - Model: ResNet-50AMD Ryzen 9 7950X 16-Core1020304050SE +/- 0.33, N = 343.11MIN: 40.78 / MAX: 44.21

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 512 - Model: ResNet-50AMD Ryzen 9 7950X 16-Core1020304050SE +/- 0.13, N = 343.33MIN: 40.95 / MAX: 43.88

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 64 - Model: ResNet-50AMD Ryzen 9 7950X 16-Core1020304050SE +/- 0.43, N = 343.16MIN: 40.82 / MAX: 44.23

FFmpeg

This is a benchmark of the FFmpeg multimedia framework. The FFmpeg test profile is making use of a modified version of vbench from Columbia University's Architecture and Design Lab (ARCADE) [http://arcade.cs.columbia.edu/vbench/] that is a benchmark for video-as-a-service workloads. The test profile offers the options of a range of vbench scenarios based on freely distributable video content and offers the options of using the x264 or x265 video encoders for transcoding. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterFFmpeg 7.0Encoder: libx265 - Scenario: LiveAMD Ryzen 9 7950X 16-Core4080120160200SE +/- 0.04, N = 3180.171. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl

OpenVINO GenAI

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: Phi-3-mini-128k-instruct-int4-ov - Device: CPU - Time Per Output TokenAMD Ryzen 9 7950X 16-Core1428425670SE +/- 0.12, N = 363.73

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: Phi-3-mini-128k-instruct-int4-ov - Device: CPU - Time To First TokenAMD Ryzen 9 7950X 16-Core1530456075SE +/- 0.33, N = 369.33

OpenBenchmarking.orgtokens/s, More Is BetterOpenVINO GenAI 2024.5Model: Phi-3-mini-128k-instruct-int4-ov - Device: CPUAMD Ryzen 9 7950X 16-Core48121620SE +/- 0.03, N = 315.69
OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: Phi-3-mini-128k-instruct-int4-ov - Device: CPU - Time To First TokenAMD Ryzen 9 7950X 16-Core1530456075SE +/- 0.33, N = 369.33
OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: Phi-3-mini-128k-instruct-int4-ov - Device: CPU - Time Per Output TokenAMD Ryzen 9 7950X 16-Core1428425670SE +/- 0.12, N = 363.73

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 32 - Model: GoogLeNetAMD Ryzen 9 7950X 16-Core306090120150SE +/- 0.13, N = 3116.32

FFmpeg

This is a benchmark of the FFmpeg multimedia framework. The FFmpeg test profile is making use of a modified version of vbench from Columbia University's Architecture and Design Lab (ARCADE) [http://arcade.cs.columbia.edu/vbench/] that is a benchmark for video-as-a-service workloads. The test profile offers the options of a range of vbench scenarios based on freely distributable video content and offers the options of using the x264 or x265 video encoders for transcoding. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterFFmpeg 7.0Encoder: libx264 - Scenario: LiveAMD Ryzen 9 7950X 16-Core60120180240300SE +/- 0.63, N = 3286.881. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl

John The Ripper

This is a benchmark of John The Ripper, which is a password cracker. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: WPA PSKAMD Ryzen 9 7950X 16-Core40K80K120K160K200KSE +/- 66.12, N = 31635711. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lm -lrt -lz -ldl -lcrypt

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: bcryptAMD Ryzen 9 7950X 16-Core9K18K27K36K45KSE +/- 9.84, N = 3442661. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lm -lrt -lz -ldl -lcrypt

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 2023.03.14Test: BlowfishAMD Ryzen 9 7950X 16-Core9K18K27K36K45KSE +/- 22.75, N = 3441741. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lm -lrt -lz -ldl -lcrypt

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 1 - Model: VGG-16AMD Ryzen 9 7950X 16-Core0.9541.9082.8623.8164.77SE +/- 0.00, N = 34.24

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 64 - Model: AlexNetAMD Ryzen 9 7950X 16-Core60120180240300SE +/- 0.15, N = 3290.12

7-Zip Compression

OpenBenchmarking.orgMIPS, More Is Better7-Zip CompressionTest: Decompression RatingAMD Ryzen 9 7950X 16-Core30K60K90K120K150KSE +/- 168.14, N = 31627811. 7-Zip 23.01 (x64) : Copyright (c) 1999-2023 Igor Pavlov : 2023-06-20

OpenBenchmarking.orgMIPS, More Is Better7-Zip CompressionTest: Compression RatingAMD Ryzen 9 7950X 16-Core30K60K90K120K150KSE +/- 121.76, N = 31478421. 7-Zip 23.01 (x64) : Copyright (c) 1999-2023 Igor Pavlov : 2023-06-20

OpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 24.05Test: Decompression RatingAMD Ryzen 9 7950X 16-Core40K80K120K160K200KSE +/- 405.66, N = 31709341. (CXX) g++ options: -lpthread -ldl -O2 -fPIC

OpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 24.05Test: Compression RatingAMD Ryzen 9 7950X 16-Core30K60K90K120K150KSE +/- 188.93, N = 31502381. (CXX) g++ options: -lpthread -ldl -O2 -fPIC

PyTorch

This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 1 - Model: ResNet-50AMD Ryzen 9 7950X 16-Core1428425670SE +/- 0.20, N = 362.93MIN: 45.93 / MAX: 66.01

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: GPU - Batch Size: 1 - Model: ResNet-50AMD Ryzen 9 7950X 16-Core246810SE +/- 0.02, N = 36.70

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 32 - Model: AlexNetAMD Ryzen 9 7950X 16-Core50100150200250SE +/- 0.16, N = 3210.45

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 GitEncoder Mode: Preset 4 - Input: Bosphorus 4KAMD Ryzen 9 7950X 16-Core3691215SE +/- 0.04, N = 310.811. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 16 - Model: GoogLeNetAMD Ryzen 9 7950X 16-Core306090120150SE +/- 0.19, N = 3118.82

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 16 - Model: AlexNetAMD Ryzen 9 7950X 16-Core306090120150SE +/- 0.06, N = 3134.49

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: GPU - Batch Size: 1 - Model: AlexNetAMD Ryzen 9 7950X 16-Core3691215SE +/- 0.03, N = 311.88

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 1 - Model: ResNet-50AMD Ryzen 9 7950X 16-Core3691215SE +/- 0.01, N = 312.89

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 1 - Model: AlexNetAMD Ryzen 9 7950X 16-Core3691215SE +/- 0.01, N = 311.54

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 GitEncoder Mode: Preset 8 - Input: Bosphorus 4KAMD Ryzen 9 7950X 16-Core20406080100SE +/- 0.20, N = 381.821. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: GPU - Batch Size: 1 - Model: GoogLeNetAMD Ryzen 9 7950X 16-Core612182430SE +/- 0.07, N = 324.25

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 GitEncoder Mode: Preset 4 - Input: Bosphorus 1080pAMD Ryzen 9 7950X 16-Core714212835SE +/- 0.15, N = 328.951. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

TensorFlow

This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 1 - Model: GoogLeNetAMD Ryzen 9 7950X 16-Core1224364860SE +/- 0.11, N = 351.20

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 GitEncoder Mode: Preset 8 - Input: Bosphorus 1080pAMD Ryzen 9 7950X 16-Core60120180240300SE +/- 0.25, N = 3257.821. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

104 Results Shown

TensorFlow:
  GPU - 512 - VGG-16
  GPU - 256 - VGG-16
  GPU - 512 - ResNet-50
  GPU - 256 - ResNet-50
  CPU - 512 - VGG-16
  GPU - 64 - VGG-16
  GPU - 512 - GoogLeNet
  CPU - 512 - ResNet-50
  CPU - 256 - VGG-16
  GPU - 32 - VGG-16
  GPU - 512 - AlexNet
  GPU - 256 - GoogLeNet
  CPU - 256 - ResNet-50
  GPU - 64 - ResNet-50
Whisper.cpp
TensorFlow:
  GPU - 16 - VGG-16
  GPU - 256 - AlexNet
  CPU - 512 - GoogLeNet
Timed LLVM Compilation
TensorFlow
Timed LLVM Compilation
TensorFlow
Whisper.cpp
TensorFlow:
  CPU - 256 - GoogLeNet
  GPU - 64 - GoogLeNet
  CPU - 64 - ResNet-50
  CPU - 32 - VGG-16
  GPU - 16 - ResNet-50
FFmpeg
PyTorch:
  CPU - 64 - Efficientnet_v2_l
  CPU - 32 - Efficientnet_v2_l
  CPU - 256 - Efficientnet_v2_l
  CPU - 16 - Efficientnet_v2_l
  CPU - 512 - Efficientnet_v2_l
TensorFlow
OpenVINO GenAI:
  Gemma-7b-int4-ov - CPU - Time Per Output Token
  Gemma-7b-int4-ov - CPU - Time To First Token
  Gemma-7b-int4-ov - CPU
TensorFlow
FFmpeg:
  libx264 - Platform
  libx264 - Video On Demand
TensorFlow
FFmpeg:
  libx265 - Platform
  libx265 - Video On Demand
  libx265 - Upload
OpenVINO GenAI:
  Falcon-7b-instruct-int4-ov - CPU - Time Per Output Token
  Falcon-7b-instruct-int4-ov - CPU - Time To First Token
  Falcon-7b-instruct-int4-ov - CPU
TensorFlow
PyTorch
TensorFlow
PyTorch:
  CPU - 32 - ResNet-152
  CPU - 256 - ResNet-152
  CPU - 512 - ResNet-152
  CPU - 16 - ResNet-152
Whisper.cpp
PyTorch
TensorFlow
PyTorch
TensorFlow:
  CPU - 256 - AlexNet
  GPU - 16 - GoogLeNet
  CPU - 64 - GoogLeNet
John The Ripper:
  MD5
  HMAC-SHA512
OpenVINO GenAI:
  TinyLlama-1.1B-Chat-v1.0 - CPU - Time Per Output Token
  TinyLlama-1.1B-Chat-v1.0 - CPU - Time To First Token
  TinyLlama-1.1B-Chat-v1.0 - CPU
TensorFlow:
  GPU - 1 - VGG-16
  CPU - 16 - ResNet-50
PyTorch
TensorFlow
PyTorch:
  CPU - 1 - ResNet-152
  CPU - 256 - ResNet-50
  CPU - 512 - ResNet-50
  CPU - 64 - ResNet-50
FFmpeg
OpenVINO GenAI:
  Phi-3-mini-128k-instruct-int4-ov - CPU - Time Per Output Token
  Phi-3-mini-128k-instruct-int4-ov - CPU - Time To First Token
  Phi-3-mini-128k-instruct-int4-ov - CPU
TensorFlow
FFmpeg
John The Ripper:
  WPA PSK
  bcrypt
  Blowfish
TensorFlow:
  CPU - 1 - VGG-16
  CPU - 64 - AlexNet
7-Zip Compression:
  Decompression Rating
  Compression Rating
7-Zip Compression:
  Decompression Rating
  Compression Rating
PyTorch
TensorFlow:
  GPU - 1 - ResNet-50
  CPU - 32 - AlexNet
SVT-AV1
TensorFlow:
  CPU - 16 - GoogLeNet
  CPU - 16 - AlexNet
  GPU - 1 - AlexNet
  CPU - 1 - ResNet-50
  CPU - 1 - AlexNet
SVT-AV1
TensorFlow
SVT-AV1
TensorFlow
SVT-AV1

AMD Ryzen 9 7950X 16-Core

Processor: AMD Ryzen 9 7950X 16-Core @ 5.88GHz (16 Cores / 32 Threads), Motherboard: MSI MPG X670E CARBON WIFI (MS-7D70) v1.0 (1.M1 BIOS), Chipset: AMD Device 14d8, Memory: 4 x 48 GB DDR5-3600MT/s Kingston KF560C32-48, Disk: 4 x 2000GB Samsung SSD 980 PRO 2TB + 2 x 4001GB Samsung SSD 870 + 62GB USB DISK 2.0, Graphics: NVIDIA GeForce RTX 4090 24GB, Audio: NVIDIA AD102 HD Audio, Monitor: BenQ GW2270, Network: Realtek RTL8125 2.5GbE + MEDIATEK MT7922 802.11ax PCI

OS: Ubuntu 24.04, Kernel: 6.8.0-41-generic (x86_64), Desktop: GNOME Shell 46.0, Display Server: X Server 1.21.1.11, Display Driver: nouveau, Compiler: GCC 13.2.0, File-System: tmpfs, Screen Resolution: 1920x1080

Kernel Notes: Transparent Huge Pages: madvise
Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-uJ7kn6/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-uJ7kn6/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v
Processor Notes: Scaling Governor: amd-pstate-epp powersave (EPP: balance_performance) - CPU Microcode: 0xa601209
Python Notes: Python 3.12.3
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 + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS; IBPB: conditional; STIBP: always-on; RSB filling; PBRSB-eIBRS: Not affected; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected

Testing initiated at 4 December 2024 16:17 by user nokitakaze.