zen 1 epyc
2 x AMD EPYC 7601 32-Core testing with a Dell 02MJ3T (1.2.5 BIOS) and Matrox G200eW3 on Ubuntu 22.04 via the Phoronix Test Suite.
HTML result view exported from: https://openbenchmarking.org/result/2211196-NE-ZEN1EPYC153&sro&grs.
NCNN
Target: CPU - Model: resnet50
NCNN
Target: CPU-v2-v2 - Model: mobilenet-v2
NCNN
Target: CPU - Model: FastestDet
NCNN
Target: CPU - Model: shufflenet-v2
NCNN
Target: CPU - Model: regnety_400m
NCNN
Target: CPU - Model: blazeface
Stress-NG
Test: CPU Cache
NCNN
Target: CPU-v3-v3 - Model: mobilenet-v3
NCNN
Target: CPU - Model: resnet18
GraphicsMagick
Operation: Resizing
NCNN
Target: CPU - Model: mnasnet
oneDNN
Harness: Matrix Multiply Batch Shapes Transformer - Data Type: f32 - Engine: CPU
NCNN
Target: CPU - Model: alexnet
Xmrig
Variant: Monero - Hash Count: 1M
NCNN
Target: CPU - Model: mobilenet
GraphicsMagick
Operation: HWB Color Space
AOM AV1
Encoder Mode: Speed 9 Realtime - Input: Bosphorus 1080p
SVT-AV1
Encoder Mode: Preset 10 - Input: Bosphorus 4K
NCNN
Target: CPU - Model: vgg16
SVT-AV1
Encoder Mode: Preset 12 - Input: Bosphorus 1080p
oneDNN
Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPU
GraphicsMagick
Operation: Noise-Gaussian
oneDNN
Harness: IP Shapes 1D - Data Type: u8s8f32 - Engine: CPU
EnCodec
Target Bandwidth: 1.5 kbps
WebP2 Image Encode
Encode Settings: Default
NCNN
Target: CPU - Model: googlenet
Stress-NG
Test: Atomic
oneDNN
Harness: Recurrent Neural Network Training - Data Type: f32 - Engine: CPU
AOM AV1
Encoder Mode: Speed 8 Realtime - Input: Bosphorus 1080p
AOM AV1
Encoder Mode: Speed 10 Realtime - Input: Bosphorus 1080p
AOM AV1
Encoder Mode: Speed 8 Realtime - Input: Bosphorus 4K
Neural Magic DeepSparse
Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Synchronous Single-Stream
Neural Magic DeepSparse
Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Synchronous Single-Stream
Mobile Neural Network
Model: resnet-v2-50
Facebook RocksDB
Test: Read While Writing
oneDNN
Harness: Deconvolution Batch shapes_3d - Data Type: f32 - Engine: CPU
NCNN
Target: CPU - Model: vision_transformer
Stress-NG
Test: Memory Copying
NCNN
Target: CPU - Model: squeezenet_ssd
Mobile Neural Network
Model: nasnet
ClickHouse
100M Rows Web Analytics Dataset, Third Run
ClickHouse
100M Rows Web Analytics Dataset, First Run / Cold Cache
oneDNN
Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPU
Mobile Neural Network
Model: squeezenetv1.1
Stress-NG
Test: Context Switching
Mobile Neural Network
Model: SqueezeNetV1.0
LAMMPS Molecular Dynamics Simulator
Model: 20k Atoms
Cpuminer-Opt
Algorithm: Garlicoin
ClickHouse
100M Rows Web Analytics Dataset, Second Run
Mobile Neural Network
Model: mobilenet-v1-1.0
NCNN
Target: CPU - Model: yolov4-tiny
SVT-AV1
Encoder Mode: Preset 12 - Input: Bosphorus 4K
EnCodec
Target Bandwidth: 24 kbps
NCNN
Target: CPU - Model: efficientnet-b0
AOM AV1
Encoder Mode: Speed 6 Realtime - Input: Bosphorus 4K
Stress-NG
Test: MMAP
oneDNN
Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPU
oneDNN
Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPU
Natron
Input: Spaceship
Stress-NG
Test: System V Message Passing
AOM AV1
Encoder Mode: Speed 6 Two-Pass - Input: Bosphorus 4K
EnCodec
Target Bandwidth: 3 kbps
WebP2 Image Encode
Encode Settings: Quality 75, Compression Effort 7
Stress-NG
Test: Glibc C String Functions
oneDNN
Harness: Recurrent Neural Network Inference - Data Type: f32 - Engine: CPU
OpenVINO
Model: Machine Translation EN To DE FP16 - Device: CPU
OpenFOAM
Input: drivaerFastback, Small Mesh Size - Execution Time
OpenVINO
Model: Machine Translation EN To DE FP16 - Device: CPU
AOM AV1
Encoder Mode: Speed 6 Two-Pass - Input: Bosphorus 1080p
Stress-NG
Test: Socket Activity
WebP Image Encode
Encode Settings: Quality 100, Lossless, Highest Compression
Mobile Neural Network
Model: mobilenetV3
spaCy
Model: en_core_web_trf
Aircrack-ng
TensorFlow
Device: CPU - Batch Size: 32 - Model: GoogLeNet
FFmpeg
Encoder: libx265 - Scenario: Live
FFmpeg
Encoder: libx265 - Scenario: Live
AOM AV1
Encoder Mode: Speed 4 Two-Pass - Input: Bosphorus 1080p
AOM AV1
Encoder Mode: Speed 9 Realtime - Input: Bosphorus 4K
Y-Cruncher
Pi Digits To Calculate: 500M
oneDNN
Harness: IP Shapes 1D - Data Type: f32 - Engine: CPU
AOM AV1
Encoder Mode: Speed 0 Two-Pass - Input: Bosphorus 1080p
BRL-CAD
VGR Performance Metric
nginx
Connections: 200
Node.js V8 Web Tooling Benchmark
OpenVINO
Model: Person Detection FP32 - Device: CPU
ASTC Encoder
Preset: Fast
miniBUDE
Implementation: OpenMP - Input Deck: BM2
miniBUDE
Implementation: OpenMP - Input Deck: BM2
oneDNN
Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPU
miniBUDE
Implementation: OpenMP - Input Deck: BM1
miniBUDE
Implementation: OpenMP - Input Deck: BM1
Stress-NG
Test: Mutex
Neural Magic DeepSparse
Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Synchronous Single-Stream
Neural Magic DeepSparse
Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Synchronous Single-Stream
TensorFlow
Device: CPU - Batch Size: 16 - Model: GoogLeNet
OpenRadioss
Model: INIVOL and Fluid Structure Interaction Drop Container
SVT-AV1
Encoder Mode: Preset 10 - Input: Bosphorus 1080p
Stress-NG
Test: Forking
GraphicsMagick
Operation: Rotate
Cpuminer-Opt
Algorithm: Myriad-Groestl
GraphicsMagick
Operation: Sharpen
Neural Magic DeepSparse
Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Synchronous Single-Stream
Neural Magic DeepSparse
Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Synchronous Single-Stream
AI Benchmark Alpha
Device Inference Score
OpenVINO
Model: Person Detection FP32 - Device: CPU
Facebook RocksDB
Test: Random Read
SMHasher
Hash: t1ha0_aes_avx2 x86_64
oneDNN
Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPU
AOM AV1
Encoder Mode: Speed 6 Realtime - Input: Bosphorus 1080p
nginx
Connections: 1000
Mobile Neural Network
Model: MobileNetV2_224
WebP2 Image Encode
Encode Settings: Quality 100, Compression Effort 5
WebP Image Encode
Encode Settings: Quality 100, Lossless
TensorFlow
Device: CPU - Batch Size: 32 - Model: ResNet-50
libavif avifenc
Encoder Speed: 6, Lossless
nginx
Connections: 500
OpenVINO
Model: Person Vehicle Bike Detection FP16 - Device: CPU
OpenVINO
Model: Person Vehicle Bike Detection FP16 - Device: CPU
SVT-AV1
Encoder Mode: Preset 8 - Input: Bosphorus 4K
AI Benchmark Alpha
Device AI Score
Neural Magic DeepSparse
Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Stream
Unpacking The Linux Kernel
linux-5.19.tar.xz
Neural Magic DeepSparse
Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Synchronous Single-Stream
SVT-AV1
Encoder Mode: Preset 4 - Input: Bosphorus 1080p
Neural Magic DeepSparse
Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Synchronous Single-Stream
Stress-NG
Test: Glibc Qsort Data Sorting
OSPRay Studio
Camera: 2 - Resolution: 1080p - Samples Per Pixel: 32 - Renderer: Path Tracer
TensorFlow
Device: CPU - Batch Size: 256 - Model: GoogLeNet
FFmpeg
Encoder: libx264 - Scenario: Platform
FFmpeg
Encoder: libx264 - Scenario: Platform
OpenRadioss
Model: Bird Strike on Windshield
OpenVINO
Model: Person Detection FP16 - Device: CPU
OSPRay Studio
Camera: 3 - Resolution: 4K - Samples Per Pixel: 16 - Renderer: Path Tracer
Neural Magic DeepSparse
Model: NLP Text Classification, DistilBERT mnli - Scenario: Synchronous Single-Stream
Neural Magic DeepSparse
Model: NLP Text Classification, DistilBERT mnli - Scenario: Synchronous Single-Stream
oneDNN
Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPU
Stress-NG
Test: Futex
OpenVINO
Model: Face Detection FP16 - Device: CPU
TensorFlow
Device: CPU - Batch Size: 512 - Model: ResNet-50
Stress-NG
Test: Malloc
Cpuminer-Opt
Algorithm: scrypt
oneDNN
Harness: IP Shapes 3D - Data Type: f32 - Engine: CPU
JPEG XL libjxl
Input: JPEG - Quality: 80
Timed Node.js Compilation
Time To Compile
Neural Magic DeepSparse
Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Stream
TensorFlow
Device: CPU - Batch Size: 512 - Model: GoogLeNet
Neural Magic DeepSparse
Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Stream
OpenVINO
Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPU
srsRAN
Test: OFDM_Test
Neural Magic DeepSparse
Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-Stream
Facebook RocksDB
Test: Update Random
Blender
Blend File: Classroom - Compute: CPU-Only
SVT-AV1
Encoder Mode: Preset 8 - Input: Bosphorus 1080p
TensorFlow
Device: CPU - Batch Size: 64 - Model: ResNet-50
OpenVINO
Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPU
JPEG XL Decoding libjxl
CPU Threads: 1
TensorFlow
Device: CPU - Batch Size: 256 - Model: ResNet-50
AOM AV1
Encoder Mode: Speed 10 Realtime - Input: Bosphorus 4K
srsRAN
Test: 4G PHY_DL_Test 100 PRB MIMO 256-QAM
Xmrig
Variant: Wownero - Hash Count: 1M
AOM AV1
Encoder Mode: Speed 4 Two-Pass - Input: Bosphorus 4K
OSPRay Studio
Camera: 2 - Resolution: 4K - Samples Per Pixel: 16 - Renderer: Path Tracer
SVT-AV1
Encoder Mode: Preset 4 - Input: Bosphorus 4K
Cpuminer-Opt
Algorithm: Magi
AI Benchmark Alpha
Device Training Score
libavif avifenc
Encoder Speed: 10, Lossless
OpenVINO
Model: Face Detection FP16 - Device: CPU
TensorFlow
Device: CPU - Batch Size: 16 - Model: AlexNet
OpenVINO
Model: Age Gender Recognition Retail 0013 FP16 - Device: CPU
JPEG XL libjxl
Input: JPEG - Quality: 90
OpenVINO
Model: Age Gender Recognition Retail 0013 FP16 - Device: CPU
C-Blosc
Test: blosclz shuffle
Facebook RocksDB
Test: Read Random Write Random
OSPRay Studio
Camera: 3 - Resolution: 1080p - Samples Per Pixel: 32 - Renderer: Path Tracer
FFmpeg
Encoder: libx265 - Scenario: Upload
Primesieve
Length: 1e12
FFmpeg
Encoder: libx265 - Scenario: Upload
Neural Magic DeepSparse
Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-Stream
EnCodec
Target Bandwidth: 6 kbps
srsRAN
Test: 4G PHY_DL_Test 100 PRB MIMO 256-QAM
Neural Magic DeepSparse
Model: CV Classification, ResNet-50 ImageNet - Scenario: Synchronous Single-Stream
TensorFlow
Device: CPU - Batch Size: 32 - Model: AlexNet
FLAC Audio Encoding
WAV To FLAC
Stress-NG
Test: MEMFD
FFmpeg
Encoder: libx265 - Scenario: Video On Demand
FFmpeg
Encoder: libx265 - Scenario: Video On Demand
Cpuminer-Opt
Algorithm: x25x
GraphicsMagick
Operation: Enhanced
Neural Magic DeepSparse
Model: CV Classification, ResNet-50 ImageNet - Scenario: Synchronous Single-Stream
Timed Wasmer Compilation
Time To Compile
Stress-NG
Test: CPU Stress
oneDNN
Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPU
OSPRay Studio
Camera: 3 - Resolution: 1080p - Samples Per Pixel: 1 - Renderer: Path Tracer
OSPRay Studio
Camera: 3 - Resolution: 1080p - Samples Per Pixel: 16 - Renderer: Path Tracer
TensorFlow
Device: CPU - Batch Size: 256 - Model: AlexNet
Primesieve
Length: 1e13
Timed CPython Compilation
Build Configuration: Default
Blender
Blend File: Fishy Cat - Compute: CPU-Only
JPEG XL Decoding libjxl
CPU Threads: All
OSPRay Studio
Camera: 2 - Resolution: 1080p - Samples Per Pixel: 1 - Renderer: Path Tracer
Neural Magic DeepSparse
Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-Stream
TensorFlow
Device: CPU - Batch Size: 64 - Model: AlexNet
srsRAN
Test: 4G PHY_DL_Test 100 PRB MIMO 64-QAM
Neural Magic DeepSparse
Model: CV Detection,YOLOv5s COCO - Scenario: Synchronous Single-Stream
Blender
Blend File: BMW27 - Compute: CPU-Only
Neural Magic DeepSparse
Model: CV Detection,YOLOv5s COCO - Scenario: Synchronous Single-Stream
GraphicsMagick
Operation: Swirl
JPEG XL libjxl
Input: PNG - Quality: 90
OpenVINO
Model: Vehicle Detection FP16 - Device: CPU
OSPRay Studio
Camera: 2 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path Tracer
OpenVINO
Model: Vehicle Detection FP16 - Device: CPU
Stress-NG
Test: IO_uring
ASTC Encoder
Preset: Medium
OSPRay Studio
Camera: 1 - Resolution: 1080p - Samples Per Pixel: 16 - Renderer: Path Tracer
libavif avifenc
Encoder Speed: 2
srsRAN
Test: 5G PHY_DL_NR Test 52 PRB SISO 64-QAM
OpenFOAM
Input: drivaerFastback, Small Mesh Size - Mesh Time
WebP Image Encode
Encode Settings: Quality 100
OSPRay Studio
Camera: 2 - Resolution: 1080p - Samples Per Pixel: 16 - Renderer: Path Tracer
OpenVINO
Model: Person Detection FP16 - Device: CPU
FFmpeg
Encoder: libx264 - Scenario: Upload
Neural Magic DeepSparse
Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Stream
Mobile Neural Network
Model: inception-v3
Neural Magic DeepSparse
Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Stream
FFmpeg
Encoder: libx264 - Scenario: Live
srsRAN
Test: 4G PHY_DL_Test 100 PRB SISO 64-QAM
FFmpeg
Encoder: libx264 - Scenario: Live
FFmpeg
Encoder: libx264 - Scenario: Upload
SMHasher
Hash: Spooky32
Cpuminer-Opt
Algorithm: Ringcoin
Neural Magic DeepSparse
Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-Stream
OpenVINO
Model: Vehicle Detection FP16-INT8 - Device: CPU
OSPRay Studio
Camera: 1 - Resolution: 4K - Samples Per Pixel: 32 - Renderer: Path Tracer
OpenVINO
Model: Vehicle Detection FP16-INT8 - Device: CPU
srsRAN
Test: 4G PHY_DL_Test 100 PRB MIMO 64-QAM
Blender
Blend File: Barbershop - Compute: CPU-Only
Stress-NG
Test: Semaphores
OSPRay Studio
Camera: 1 - Resolution: 1080p - Samples Per Pixel: 1 - Renderer: Path Tracer
TensorFlow
Device: CPU - Batch Size: 512 - Model: AlexNet
OpenRadioss
Model: Bumper Beam
FFmpeg
Encoder: libx264 - Scenario: Video On Demand
Neural Magic DeepSparse
Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Stream
C-Blosc
Test: blosclz bitshuffle
OSPRay Studio
Camera: 1 - Resolution: 4K - Samples Per Pixel: 16 - Renderer: Path Tracer
Timed CPython Compilation
Build Configuration: Released Build, PGO + LTO Optimized
OpenRadioss
Model: Cell Phone Drop Test
FFmpeg
Encoder: libx264 - Scenario: Video On Demand
WebP Image Encode
Encode Settings: Default
JPEG XL libjxl
Input: PNG - Quality: 80
LAMMPS Molecular Dynamics Simulator
Model: Rhodopsin Protein
Neural Magic DeepSparse
Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Stream
srsRAN
Test: 4G PHY_DL_Test 100 PRB SISO 256-QAM
OSPRay Studio
Camera: 1 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path Tracer
FFmpeg
Encoder: libx265 - Scenario: Platform
OSPRay Studio
Camera: 2 - Resolution: 4K - Samples Per Pixel: 32 - Renderer: Path Tracer
OpenRadioss
Model: Rubber O-Ring Seal Installation
ASTC Encoder
Preset: Thorough
Neural Magic DeepSparse
Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Stream
TensorFlow
Device: CPU - Batch Size: 16 - Model: ResNet-50
FFmpeg
Encoder: libx265 - Scenario: Platform
Timed Erlang/OTP Compilation
Time To Compile
SMHasher
Hash: FarmHash128
srsRAN
Test: 4G PHY_DL_Test 100 PRB SISO 64-QAM
Stress-NG
Test: NUMA
srsRAN
Test: 4G PHY_DL_Test 100 PRB SISO 256-QAM
oneDNN
Harness: Deconvolution Batch shapes_1d - Data Type: f32 - Engine: CPU
OpenFOAM
Input: drivaerFastback, Medium Mesh Size - Execution Time
Timed PHP Compilation
Time To Compile
oneDNN
Harness: Deconvolution Batch shapes_3d - Data Type: u8s8f32 - Engine: CPU
SMHasher
Hash: wyhash
Y-Cruncher
Pi Digits To Calculate: 1B
SMHasher
Hash: MeowHash x86_64 AES-NI
Cpuminer-Opt
Algorithm: Deepcoin
OpenVINO
Model: Weld Porosity Detection FP16-INT8 - Device: CPU
oneDNN
Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPU
OpenVINO
Model: Face Detection FP16-INT8 - Device: CPU
OSPRay Studio
Camera: 1 - Resolution: 1080p - Samples Per Pixel: 32 - Renderer: Path Tracer
OSPRay Studio
Camera: 3 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path Tracer
OSPRay Studio
Camera: 3 - Resolution: 4K - Samples Per Pixel: 32 - Renderer: Path Tracer
nekRS
Input: TurboPipe Periodic
Stress-NG
Test: SENDFILE
Stress-NG
Test: Matrix Math
Blender
Blend File: Pabellon Barcelona - Compute: CPU-Only
OpenVINO
Model: Weld Porosity Detection FP16-INT8 - Device: CPU
Stress-NG
Test: Vector Math
SMHasher
Hash: FarmHash32 x86_64 AVX
spaCy
Model: en_core_web_lg
Cpuminer-Opt
Algorithm: Blake-2 S
OpenFOAM
Input: drivaerFastback, Medium Mesh Size - Mesh Time
TensorFlow
Device: CPU - Batch Size: 64 - Model: GoogLeNet
SMHasher
Hash: t1ha2_atonce
OpenVINO
Model: Weld Porosity Detection FP16 - Device: CPU
ASTC Encoder
Preset: Exhaustive
Stress-NG
Test: Crypto
SMHasher
Hash: SHA3-256
OpenVINO
Model: Weld Porosity Detection FP16 - Device: CPU
Cpuminer-Opt
Algorithm: Triple SHA-256, Onecoin
Cpuminer-Opt
Algorithm: Skeincoin
libavif avifenc
Encoder Speed: 0
SMHasher
Hash: fasthash32
Neural Magic DeepSparse
Model: CV Detection,YOLOv5s COCO - Scenario: Asynchronous Multi-Stream
Cpuminer-Opt
Algorithm: Quad SHA-256, Pyrite
Neural Magic DeepSparse
Model: CV Detection,YOLOv5s COCO - Scenario: Asynchronous Multi-Stream
OpenVINO
Model: Face Detection FP16-INT8 - Device: CPU
Cpuminer-Opt
Algorithm: LBC, LBRY Credits
libavif avifenc
Encoder Speed: 6
AOM AV1
Encoder Mode: Speed 0 Two-Pass - Input: Bosphorus 4K
srsRAN
Test: 5G PHY_DL_NR Test 52 PRB SISO 64-QAM
WebP2 Image Encode
Encode Settings: Quality 100, Lossless Compression
WebP2 Image Encode
Encode Settings: Quality 95, Compression Effort 7
WebP Image Encode
Encode Settings: Quality 100, Highest Compression
JPEG XL libjxl
Input: JPEG - Quality: 100
JPEG XL libjxl
Input: PNG - Quality: 100
SMHasher
Hash: MeowHash x86_64 AES-NI
SMHasher
Hash: t1ha0_aes_avx2 x86_64
SMHasher
Hash: FarmHash32 x86_64 AVX
SMHasher
Hash: t1ha2_atonce
SMHasher
Hash: FarmHash128
SMHasher
Hash: fasthash32
SMHasher
Hash: Spooky32
SMHasher
Hash: SHA3-256
SMHasher
Hash: wyhash
Phoronix Test Suite v10.8.5