April 2023 AMD Zen 4 Linux Benchmarks
Benchmarks by Michael Larabel for a future article.
HTML result view exported from: https://openbenchmarking.org/result/2303304-PTS-EARLYUPP87&sro&grr.
OpenFOAM
Input: drivaerFastback, Medium Mesh Size - Execution Time
OpenFOAM
Input: drivaerFastback, Medium Mesh Size - Mesh Time
Timed Linux Kernel Compilation
Build: allmodconfig
Timed LLVM Compilation
Build System: Unix Makefiles
Timed LLVM Compilation
Build System: Ninja
ONNX Runtime
Model: bertsquad-12 - Device: CPU - Executor: Standard
ONNX Runtime
Model: bertsquad-12 - Device: CPU - Executor: Standard
Timed Node.js Compilation
Time To Compile
OpenVKL
Benchmark: vklBenchmark ISPC
ONNX Runtime
Model: ResNet50 v1-12-int8 - Device: CPU - Executor: Standard
ONNX Runtime
Model: ResNet50 v1-12-int8 - Device: CPU - Executor: Standard
OSPRay
Benchmark: particle_volume/pathtracer/real_time
LeelaChessZero
Backend: Eigen
LeelaChessZero
Backend: BLAS
ONNX Runtime
Model: yolov4 - Device: CPU - Executor: Standard
ONNX Runtime
Model: yolov4 - Device: CPU - Executor: Standard
ONNX Runtime
Model: ArcFace ResNet-100 - Device: CPU - Executor: Standard
ONNX Runtime
Model: ArcFace ResNet-100 - Device: CPU - Executor: Standard
ONNX Runtime
Model: Faster R-CNN R-50-FPN-int8 - Device: CPU - Executor: Standard
ONNX Runtime
Model: Faster R-CNN R-50-FPN-int8 - Device: CPU - Executor: Standard
OSPRay
Benchmark: particle_volume/scivis/real_time
ONNX Runtime
Model: super-resolution-10 - Device: CPU - Executor: Standard
ONNX Runtime
Model: super-resolution-10 - Device: CPU - Executor: Standard
ONNX Runtime
Model: CaffeNet 12-int8 - Device: CPU - Executor: Standard
ONNX Runtime
Model: CaffeNet 12-int8 - Device: CPU - Executor: Standard
ONNX Runtime
Model: fcn-resnet101-11 - Device: CPU - Executor: Standard
ONNX Runtime
Model: fcn-resnet101-11 - Device: CPU - Executor: Standard
Blender
Blend File: Pabellon Barcelona - Compute: CPU-Only
Timed Godot Game Engine Compilation
Time To Compile
ClickHouse
100M Rows Hits Dataset, Third Run
ClickHouse
100M Rows Hits Dataset, Second Run
ClickHouse
100M Rows Hits Dataset, First Run / Cold Cache
NCNN
Target: CPU - Model: FastestDet
NCNN
Target: CPU - Model: vision_transformer
NCNN
Target: CPU - Model: regnety_400m
NCNN
Target: CPU - Model: squeezenet_ssd
NCNN
Target: CPU - Model: yolov4-tiny
NCNN
Target: CPU - Model: resnet50
NCNN
Target: CPU - Model: alexnet
NCNN
Target: CPU - Model: resnet18
NCNN
Target: CPU - Model: vgg16
NCNN
Target: CPU - Model: googlenet
NCNN
Target: CPU - Model: blazeface
NCNN
Target: CPU - Model: efficientnet-b0
NCNN
Target: CPU - Model: mnasnet
NCNN
Target: CPU - Model: shufflenet-v2
NCNN
Target: CPU-v3-v3 - Model: mobilenet-v3
NCNN
Target: CPU-v2-v2 - Model: mobilenet-v2
NCNN
Target: CPU - Model: mobilenet
OSPRay
Benchmark: particle_volume/ao/real_time
BRL-CAD
VGR Performance Metric
OpenCV
Test: Image Processing
FFmpeg
Encoder: libx264 - Scenario: Upload
FFmpeg
Encoder: libx264 - Scenario: Upload
OpenSSL
Algorithm: SHA256
OpenSSL
Algorithm: AES-256-GCM
OpenSSL
Algorithm: AES-128-GCM
OpenSSL
Algorithm: ChaCha20-Poly1305
OpenSSL
Algorithm: ChaCha20
OpenSSL
Algorithm: SHA512
Zstd Compression
Compression Level: 19, Long Mode - Decompression Speed
Zstd Compression
Compression Level: 19, Long Mode - Compression Speed
GPAW
Input: Carbon Nanotube
Zstd Compression
Compression Level: 19 - Decompression Speed
Zstd Compression
Compression Level: 19 - Compression Speed
SPECFEM3D
Model: Layered Halfspace
ASKAP
Test: tConvolve MT - Degridding
ASKAP
Test: tConvolve MT - Gridding
SPECFEM3D
Model: Water-layered Halfspace
simdjson
Throughput Test: Kostya
ONNX Runtime
Model: GPT-2 - Device: CPU - Executor: Standard
ONNX Runtime
Model: GPT-2 - Device: CPU - Executor: Standard
Zstd Compression
Compression Level: 8 - Decompression Speed
Zstd Compression
Compression Level: 8 - Compression Speed
OpenCV
Test: Graph API
OSPRay Studio
Camera: 3 - Resolution: 1080p - Samples Per Pixel: 32 - Renderer: Path Tracer
simdjson
Throughput Test: LargeRandom
FFmpeg
Encoder: libx265 - Scenario: Platform
FFmpeg
Encoder: libx265 - Scenario: Platform
FFmpeg
Encoder: libx265 - Scenario: Video On Demand
FFmpeg
Encoder: libx265 - Scenario: Video On Demand
FFmpeg
Encoder: libx265 - Scenario: Upload
FFmpeg
Encoder: libx265 - Scenario: Upload
FFmpeg
Encoder: libx264 - Scenario: Platform
FFmpeg
Encoder: libx264 - Scenario: Platform
FFmpeg
Encoder: libx264 - Scenario: Video On Demand
FFmpeg
Encoder: libx264 - Scenario: Video On Demand
SPECFEM3D
Model: Mount St. Helens
SPECFEM3D
Model: Homogeneous Halfspace
OpenCV
Test: Stitching
Apache HTTP Server
Concurrent Requests: 1000
OSPRay Studio
Camera: 3 - Resolution: 1080p - Samples Per Pixel: 1 - Renderer: Path Tracer
Selenium
Benchmark: Jetstream 2 - Browser: Firefox
OSPRay Studio
Camera: 3 - Resolution: 1080p - Samples Per Pixel: 16 - Renderer: Path Tracer
OSPRay Studio
Camera: 1 - Resolution: 1080p - Samples Per Pixel: 1 - Renderer: Path Tracer
OSPRay
Benchmark: gravity_spheres_volume/dim_512/scivis/real_time
OSPRay
Benchmark: gravity_spheres_volume/dim_512/ao/real_time
OSPRay Studio
Camera: 1 - Resolution: 1080p - Samples Per Pixel: 32 - Renderer: Path Tracer
OSPRay Studio
Camera: 1 - Resolution: 1080p - Samples Per Pixel: 16 - Renderer: Path Tracer
OSPRay
Benchmark: gravity_spheres_volume/dim_512/pathtracer/real_time
Build2
Time To Compile
VVenC
Video Input: Bosphorus 4K - Video Preset: Fast
Blender
Blend File: Fishy Cat - Compute: CPU-Only
OpenCV
Test: DNN - Deep Neural Network
GROMACS
Implementation: MPI CPU - Input: water_GMX50_bare
Numpy Benchmark
nginx
Connections: 1000
nginx
Connections: 500
Apache HTTP Server
Concurrent Requests: 500
nginx
Connections: 200
Apache HTTP Server
Concurrent Requests: 200
nginx
Connections: 100
Apache HTTP Server
Concurrent Requests: 100
Stockfish
Total Time
oneDNN
Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPU
OpenCV
Test: Object Detection
oneDNN
Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPU
Blender
Blend File: BMW27 - Compute: CPU-Only
ACES DGEMM
Sustained Floating-Point Rate
OpenCV
Test: Core
OpenFOAM
Input: drivaerFastback, Small Mesh Size - Execution Time
OpenFOAM
Input: drivaerFastback, Small Mesh Size - Mesh Time
Stress-NG
Test: CPU Cache
Chaos Group V-RAY
Mode: CPU
Stargate Digital Audio Workstation
Sample Rate: 192000 - Buffer Size: 512
Xcompact3d Incompact3d
Input: input.i3d 193 Cells Per Direction
Selenium
Benchmark: Kraken - Browser: Firefox
Memcached
Set To Get Ratio: 1:100
Memcached
Set To Get Ratio: 1:100
Memcached
Set To Get Ratio: 1:10
Memcached
Set To Get Ratio: 1:10
Memcached
Set To Get Ratio: 1:5
Stargate Digital Audio Workstation
Sample Rate: 192000 - Buffer Size: 1024
simdjson
Throughput Test: DistinctUserID
simdjson
Throughput Test: TopTweet
simdjson
Throughput Test: PartialTweets
Timed Linux Kernel Compilation
Build: defconfig
OpenVINO
Model: Person Detection FP16 - Device: CPU
OpenVINO
Model: Person Detection FP16 - Device: CPU
OpenVINO
Model: Person Detection FP32 - Device: CPU
OpenVINO
Model: Person Detection FP32 - Device: CPU
OpenVINO
Model: Face Detection FP16 - Device: CPU
OpenVINO
Model: Face Detection FP16 - Device: CPU
Zstd Compression
Compression Level: 12 - Decompression Speed
Zstd Compression
Compression Level: 12 - Compression Speed
Selenium
Benchmark: Jetstream 2 - Browser: Google Chrome
Zstd Compression
Compression Level: 8, Long Mode - Decompression Speed
Zstd Compression
Compression Level: 8, Long Mode - Compression Speed
OpenVINO
Model: Face Detection FP16-INT8 - Device: CPU
OpenVINO
Model: Face Detection FP16-INT8 - Device: CPU
OpenVINO
Model: Machine Translation EN To DE FP16 - Device: CPU
OpenVINO
Model: Machine Translation EN To DE FP16 - Device: CPU
SPECFEM3D
Model: Tomographic Model
OpenVINO
Model: Person Vehicle Bike Detection FP16 - Device: CPU
OpenVINO
Model: Person Vehicle Bike Detection FP16 - Device: CPU
OpenVINO
Model: Vehicle Detection FP16-INT8 - Device: CPU
OpenVINO
Model: Vehicle Detection FP16-INT8 - Device: CPU
OpenVINO
Model: Vehicle Detection FP16 - Device: CPU
OpenVINO
Model: Vehicle Detection FP16 - Device: CPU
OpenVINO
Model: Age Gender Recognition Retail 0013 FP16 - Device: CPU
OpenVINO
Model: Age Gender Recognition Retail 0013 FP16 - Device: CPU
OpenVINO
Model: Weld Porosity Detection FP16-INT8 - Device: CPU
OpenVINO
Model: Weld Porosity Detection FP16-INT8 - Device: CPU
OpenVINO
Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPU
OpenVINO
Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPU
OpenVINO
Model: Weld Porosity Detection FP16 - Device: CPU
OpenVINO
Model: Weld Porosity Detection FP16 - Device: CPU
John The Ripper
Test: MD5
VP9 libvpx Encoding
Speed: Speed 0 - Input: Bosphorus 4K
RocksDB
Test: Update Random
RocksDB
Test: Read While Writing
RocksDB
Test: Read Random Write Random
John The Ripper
Test: HMAC-SHA512
RocksDB
Test: Random Read
OpenSSL
Algorithm: RSA4096
OpenSSL
Algorithm: RSA4096
NAMD
ATPase Simulation - 327,506 Atoms
Stress-NG
Test: CPU Stress
VVenC
Video Input: Bosphorus 4K - Video Preset: Faster
CloverLeaf
Lagrangian-Eulerian Hydrodynamics
VP9 libvpx Encoding
Speed: Speed 0 - Input: Bosphorus 1080p
Darmstadt Automotive Parallel Heterogeneous Suite
Backend: OpenMP - Kernel: Points2Image
Stress-NG
Test: Crypto
OpenCV
Test: Video
Stargate Digital Audio Workstation
Sample Rate: 96000 - Buffer Size: 512
x265
Video Input: Bosphorus 4K
Selenium
Benchmark: Speedometer - Browser: Firefox
Stargate Digital Audio Workstation
Sample Rate: 96000 - Buffer Size: 1024
FFmpeg
Encoder: libx265 - Scenario: Live
FFmpeg
Encoder: libx265 - Scenario: Live
Selenium
Benchmark: PSPDFKit WASM - Browser: Firefox
Pennant
Test: sedovbig
Embree
Binary: Pathtracer - Model: Asian Dragon Obj
SVT-AV1
Encoder Mode: Preset 4 - Input: Bosphorus 4K
Stress-NG
Test: Context Switching
Embree
Binary: Pathtracer ISPC - Model: Asian Dragon Obj
Crafty
Elapsed Time
OpenCV
Test: Features 2D
Selenium
Benchmark: PSPDFKit WASM - Browser: Google Chrome
Selenium
Benchmark: Speedometer - Browser: Google Chrome
oneDNN
Harness: Deconvolution Batch shapes_1d - Data Type: bf16bf16bf16 - Engine: CPU
VVenC
Video Input: Bosphorus 1080p - Video Preset: Fast
Embree
Binary: Pathtracer - Model: Crown
Selenium
Benchmark: WASM collisionDetection - Browser: Google Chrome
Darmstadt Automotive Parallel Heterogeneous Suite
Backend: OpenMP - Kernel: NDT Mapping
John The Ripper
Test: WPA PSK
oneDNN
Harness: IP Shapes 1D - Data Type: bf16bf16bf16 - Engine: CPU
John The Ripper
Test: bcrypt
John The Ripper
Test: Blowfish
Stress-NG
Test: Memory Copying
Stress-NG
Test: Matrix Math
Stress-NG
Test: Vector Math
Timed FFmpeg Compilation
Time To Compile
x264
Video Input: Bosphorus 4K
Selenium
Benchmark: WASM imageConvolute - Browser: Firefox
Embree
Binary: Pathtracer ISPC - Model: Crown
Embree
Binary: Pathtracer - Model: Asian Dragon
FFmpeg
Encoder: libx264 - Scenario: Live
FFmpeg
Encoder: libx264 - Scenario: Live
VP9 libvpx Encoding
Speed: Speed 5 - Input: Bosphorus 4K
Pennant
Test: leblancbig
Embree
Binary: Pathtracer ISPC - Model: Asian Dragon
PyBench
Total For Average Test Times
Selenium
Benchmark: WASM imageConvolute - Browser: Google Chrome
ASKAP
Test: tConvolve MPI - Gridding
ASKAP
Test: tConvolve MPI - Degridding
PHPBench
PHP Benchmark Suite
Selenium
Benchmark: WASM collisionDetection - Browser: Firefox
uvg266
Video Input: Bosphorus 4K - Video Preset: Very Fast
uvg266
Video Input: Bosphorus 4K - Video Preset: Super Fast
oneDNN
Harness: IP Shapes 3D - Data Type: bf16bf16bf16 - Engine: CPU
VVenC
Video Input: Bosphorus 1080p - Video Preset: Faster
uvg266
Video Input: Bosphorus 4K - Video Preset: Ultra Fast
Xcompact3d Incompact3d
Input: input.i3d 129 Cells Per Direction
7-Zip Compression
Test: Decompression Rating
7-Zip Compression
Test: Compression Rating
Selenium
Benchmark: Kraken - Browser: Google Chrome
VP9 libvpx Encoding
Speed: Speed 5 - Input: Bosphorus 1080p
dav1d
Video Input: Chimera 1080p 10-bit
SVT-AV1
Encoder Mode: Preset 8 - Input: Bosphorus 4K
GIMP
Test: unsharp-mask
GIMP
Test: resize
ASKAP
Test: Hogbom Clean OpenMP
Darmstadt Automotive Parallel Heterogeneous Suite
Backend: OpenMP - Kernel: Euclidean Cluster
dav1d
Video Input: Summer Nature 4K
oneDNN
Harness: Matrix Multiply Batch Shapes Transformer - Data Type: bf16bf16bf16 - Engine: CPU
dav1d
Video Input: Chimera 1080p
GIMP
Test: auto-levels
GIMP
Test: rotate
oneDNN
Harness: Convolution Batch Shapes Auto - Data Type: bf16bf16bf16 - Engine: CPU
WebP Image Encode
Encode Settings: Quality 100, Highest Compression
ASKAP
Test: tConvolve OpenMP - Degridding
ASKAP
Test: tConvolve OpenMP - Gridding
LULESH
x265
Video Input: Bosphorus 1080p
GNU Octave Benchmark
Google Draco
Model: Church Facade
Google Draco
Model: Lion
SVT-AV1
Encoder Mode: Preset 12 - Input: Bosphorus 4K
Darktable
Test: Masskrug - Acceleration: CPU-only
SVT-AV1
Encoder Mode: Preset 13 - Input: Bosphorus 4K
Darktable
Test: Boat - Acceleration: CPU-only
Darktable
Test: Server Room - Acceleration: CPU-only
x264
Video Input: Bosphorus 1080p
oneDNN
Harness: Deconvolution Batch shapes_3d - Data Type: bf16bf16bf16 - Engine: CPU
dav1d
Video Input: Summer Nature 1080p
WebP Image Encode
Encode Settings: Quality 100
WebP Image Encode
Encode Settings: Default
Darktable
Test: Server Rack - Acceleration: CPU-only
Phoronix Test Suite v10.8.5