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