AMD EPYC 7F52
AMD EPYC 7F52 16-Core testing with a Supermicro H11DSi-NT v2.00 (2.1 BIOS) and llvmpipe on Ubuntu 20.04 via the Phoronix Test Suite.
HTML result view exported from: https://openbenchmarking.org/result/2012294-HA-AMDEPYC7F75&grs.
Polyhedron Fortran Benchmarks
Benchmark: tfft2
oneDNN
Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPU
Redis
Test: LPOP
oneDNN
Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPU
oneDNN
Harness: Matrix Multiply Batch Shapes Transformer - Data Type: f32 - Engine: CPU
oneDNN
Harness: IP Shapes 3D - Data Type: f32 - Engine: CPU
Stress-NG
Test: Forking
Stress-NG
Test: System V Message Passing
oneDNN
Harness: IP Shapes 1D - Data Type: f32 - Engine: CPU
oneDNN
Harness: Deconvolution Batch shapes_1d - Data Type: f32 - Engine: CPU
oneDNN
Harness: Deconvolution Batch shapes_3d - Data Type: f32 - Engine: CPU
oneDNN
Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPU
PyPerformance
Benchmark: python_startup
oneDNN
Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPU
oneDNN
Harness: Recurrent Neural Network Training - Data Type: f32 - Engine: CPU
oneDNN
Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPU
oneDNN
Harness: Recurrent Neural Network Inference - Data Type: f32 - Engine: CPU
oneDNN
Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPU
oneDNN
Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPU
Stress-NG
Test: MMAP
Redis
Test: GET
GraphicsMagick
Operation: HWB Color Space
OpenVKL
Benchmark: vklBenchmarkVdbVolume
Stress-NG
Test: SENDFILE
Kvazaar
Video Input: Bosphorus 1080p - Video Preset: Ultra Fast
OpenVKL
Benchmark: vklBenchmarkStructuredVolume
Stress-NG
Test: MEMFD
SVT-VP9
Tuning: PSNR/SSIM Optimized - Input: Bosphorus 1080p
Stress-NG
Test: Socket Activity
Redis
Test: SADD
Open Porous Media
OPM Benchmark: Flow MPI Norne - Threads: 8
PostgreSQL pgbench
Scaling Factor: 1 - Clients: 250 - Mode: Read Only - Average Latency
Open Porous Media
OPM Benchmark: Flow MPI Norne - Threads: 16
SVT-VP9
Tuning: Visual Quality Optimized - Input: Bosphorus 1080p
NCNN
Target: CPU - Model: squeezenet_ssd
Kvazaar
Video Input: Bosphorus 1080p - Video Preset: Very Fast
PostgreSQL pgbench
Scaling Factor: 1 - Clients: 250 - Mode: Read Only
Darmstadt Automotive Parallel Heterogeneous Suite
Backend: OpenMP - Kernel: Points2Image
oneDNN
Harness: Deconvolution Batch shapes_3d - Data Type: u8s8f32 - Engine: CPU
librsvg
Operation: SVG Files To PNG
Redis
Test: LPUSH
PyPerformance
Benchmark: float
AOM AV1
Encoder Mode: Speed 0 Two-Pass
PlaidML
FP16: No - Mode: Inference - Network: ResNet 50 - Device: CPU
Timed HMMer Search
Pfam Database Search
Timed GDB GNU Debugger Compilation
Time To Compile
SVT-VP9
Tuning: VMAF Optimized - Input: Bosphorus 1080p
PostgreSQL pgbench
Scaling Factor: 1 - Clients: 1 - Mode: Read Only - Average Latency
GPAW
Input: Carbon Nanotube
Numenta Anomaly Benchmark
Detector: EXPoSE
Zstd Compression
Compression Level: 19
WireGuard + Linux Networking Stack Stress Test
simdjson
Throughput Test: LargeRandom
Stress-NG
Test: Memory Copying
Polyhedron Fortran Benchmarks
Benchmark: test_fpu2
Zstd Compression
Compression Level: 3
PyPerformance
Benchmark: pathlib
ECP-CANDLE
Benchmark: P1B2
PlaidML
FP16: No - Mode: Inference - Network: VGG16 - Device: CPU
Numenta Anomaly Benchmark
Detector: Bayesian Changepoint
PlaidML
FP16: No - Mode: Inference - Network: Inception V3 - Device: CPU
GraphicsMagick
Operation: Noise-Gaussian
Kvazaar
Video Input: Bosphorus 4K - Video Preset: Ultra Fast
ECP-CANDLE
Benchmark: P3B1
PlaidML
FP16: No - Mode: Inference - Network: VGG19 - Device: CPU
Redis
Test: SET
PostgreSQL pgbench
Scaling Factor: 1 - Clients: 50 - Mode: Read Only - Average Latency
LZ4 Compression
Compression Level: 9 - Compression Speed
Stress-NG
Test: Context Switching
PostgreSQL pgbench
Scaling Factor: 1 - Clients: 50 - Mode: Read Write
simdjson
Throughput Test: Kostya
PostgreSQL pgbench
Scaling Factor: 1 - Clients: 50 - Mode: Read Write - Average Latency
NCNN
Target: CPU - Model: resnet50
Tachyon
Total Time
PostgreSQL pgbench
Scaling Factor: 1 - Clients: 50 - Mode: Read Only
Crafty
Elapsed Time
Stress-NG
Test: NUMA
KeyDB
Mobile Neural Network
Model: resnet-v2-50
Mobile Neural Network
Model: inception-v3
PyPerformance
Benchmark: django_template
Stress-NG
Test: Semaphores
PostgreSQL pgbench
Scaling Factor: 1 - Clients: 100 - Mode: Read Only - Average Latency
dav1d
Video Input: Summer Nature 1080p
PyPerformance
Benchmark: pickle_pure_python
Mlpack Benchmark
Benchmark: scikit_svm
NCNN
Target: CPU - Model: mobilenet
PostgreSQL pgbench
Scaling Factor: 1 - Clients: 100 - Mode: Read Only
VP9 libvpx Encoding
Speed: Speed 5
x265
Video Input: Bosphorus 4K
Embree
Binary: Pathtracer ISPC - Model: Asian Dragon
PostgreSQL pgbench
Scaling Factor: 1 - Clients: 1 - Mode: Read Only
Mobile Neural Network
Model: MobileNetV2_224
Open Porous Media
OPM Benchmark: Flow MPI Norne - Threads: 4
Stress-NG
Test: Matrix Math
BRL-CAD
VGR Performance Metric
7-Zip Compression
Compress Speed Test
Caffe
Model: AlexNet - Acceleration: CPU - Iterations: 100
Embree
Binary: Pathtracer - Model: Crown
Timed Apache Compilation
Time To Compile
OpenVINO
Model: Age Gender Recognition Retail 0013 FP32 - Device: CPU
AOM AV1
Encoder Mode: Speed 6 Realtime
Timed Eigen Compilation
Time To Compile
SVT-AV1
Encoder Mode: Enc Mode 8 - Input: 1080p
GNU Octave Benchmark
Mlpack Benchmark
Benchmark: scikit_ica
PostgreSQL pgbench
Scaling Factor: 1 - Clients: 250 - Mode: Read Write
XZ Compression
Compressing ubuntu-16.04.3-server-i386.img, Compression Level 9
Numenta Anomaly Benchmark
Detector: Windowed Gaussian
dav1d
Video Input: Chimera 1080p
PHPBench
PHP Benchmark Suite
InfluxDB
Concurrent Streams: 4 - Batch Size: 10000 - Tags: 2,5000,1 - Points Per Series: 10000
PostgreSQL pgbench
Scaling Factor: 1 - Clients: 250 - Mode: Read Write - Average Latency
PostgreSQL pgbench
Scaling Factor: 1 - Clients: 100 - Mode: Read Write - Average Latency
LibRaw
Post-Processing Benchmark
PostgreSQL pgbench
Scaling Factor: 1 - Clients: 100 - Mode: Read Write
oneDNN
Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPU
PyPerformance
Benchmark: 2to3
PyPerformance
Benchmark: crypto_pyaes
AI Benchmark Alpha
Device Training Score
Coremark
CoreMark Size 666 - Iterations Per Second
PyPerformance
Benchmark: chaos
eSpeak-NG Speech Engine
Text-To-Speech Synthesis
Node.js V8 Web Tooling Benchmark
SVT-AV1
Encoder Mode: Enc Mode 0 - Input: 1080p
Kvazaar
Video Input: Bosphorus 1080p - Video Preset: Slow
Darmstadt Automotive Parallel Heterogeneous Suite
Backend: OpenMP - Kernel: NDT Mapping
Kvazaar
Video Input: Bosphorus 1080p - Video Preset: Medium
Chaos Group V-RAY
Mode: CPU
x265
Video Input: Bosphorus 1080p
GraphicsMagick
Operation: Rotate
PyPerformance
Benchmark: json_loads
Embree
Binary: Pathtracer ISPC - Model: Crown
Blender
Blend File: BMW27 - Compute: CPU-Only
oneDNN
Harness: IP Shapes 1D - Data Type: u8s8f32 - Engine: CPU
Build2
Time To Compile
NCNN
Target: CPU - Model: efficientnet-b0
dav1d
Video Input: Chimera 1080p 10-bit
Timed FFmpeg Compilation
Time To Compile
LibreOffice
Test: 20 Documents To PDF
Kvazaar
Video Input: Bosphorus 4K - Video Preset: Medium
Polyhedron Fortran Benchmarks
Benchmark: aermod
NCNN
Target: CPU-v3-v3 - Model: mobilenet-v3
NCNN
Target: CPU - Model: regnety_400m
PlaidML
FP16: No - Mode: Inference - Network: DenseNet 201 - Device: CPU
SQLite Speedtest
Timed Time - Size 1,000
YafaRay
Total Time For Sample Scene
LZ4 Compression
Compression Level: 1 - Compression Speed
NCNN
Target: CPU-v2-v2 - Model: mobilenet-v2
Mlpack Benchmark
Benchmark: scikit_linearridgeregression
AI Benchmark Alpha
Device AI Score
FLAC Audio Encoding
WAV To FLAC
VP9 libvpx Encoding
Speed: Speed 0
PostgreSQL pgbench
Scaling Factor: 1 - Clients: 1 - Mode: Read Write
BYTE Unix Benchmark
Computational Test: Dhrystone 2
NCNN
Target: CPU - Model: blazeface
x264
H.264 Video Encoding
SVT-AV1
Encoder Mode: Enc Mode 4 - Input: 1080p
Polyhedron Fortran Benchmarks
Benchmark: channel2
OpenVKL
Benchmark: vklBenchmark
TensorFlow Lite
Model: NASNet Mobile
NAMD
ATPase Simulation - 327,506 Atoms
TNN
Target: CPU - Model: SqueezeNet v1.1
NCNN
Target: CPU - Model: vgg16
Kvazaar
Video Input: Bosphorus 4K - Video Preset: Very Fast
Darmstadt Automotive Parallel Heterogeneous Suite
Backend: OpenMP - Kernel: Euclidean Cluster
Ogg Audio Encoding
WAV To Ogg
LZ4 Compression
Compression Level: 3 - Decompression Speed
InfluxDB
Concurrent Streams: 64 - Batch Size: 10000 - Tags: 2,5000,1 - Points Per Series: 10000
asmFish
1024 Hash Memory, 26 Depth
LZ4 Compression
Compression Level: 9 - Decompression Speed
Embree
Binary: Pathtracer - Model: Asian Dragon
Stress-NG
Test: Atomic
AOM AV1
Encoder Mode: Speed 4 Two-Pass
AOM AV1
Encoder Mode: Speed 8 Realtime
WebP Image Encode
Encode Settings: Quality 100, Lossless
Kvazaar
Video Input: Bosphorus 4K - Video Preset: Slow
PyPerformance
Benchmark: go
IndigoBench
Acceleration: CPU - Scene: Supercar
NCNN
Target: CPU - Model: yolov4-tiny
GROMACS
Water Benchmark
PostgreSQL pgbench
Scaling Factor: 1 - Clients: 1 - Mode: Read Write - Average Latency
GraphicsMagick
Operation: Resizing
ECP-CANDLE
Benchmark: P3B2
Numpy Benchmark
Mobile Neural Network
Model: mobilenet-v1-1.0
Stress-NG
Test: CPU Stress
Caffe
Model: GoogleNet - Acceleration: CPU - Iterations: 100
FFTE
N=256, 3D Complex FFT Routine
Timed Linux Kernel Compilation
Time To Compile
OpenVINO
Model: Person Detection 0106 FP32 - Device: CPU
OpenVINO
Model: Person Detection 0106 FP16 - Device: CPU
Caffe
Model: GoogleNet - Acceleration: CPU - Iterations: 200
TensorFlow Lite
Model: Mobilenet Float
Polyhedron Fortran Benchmarks
Benchmark: ac
Blender
Blend File: Fishy Cat - Compute: CPU-Only
LZ4 Compression
Compression Level: 1 - Decompression Speed
TensorFlow Lite
Model: Inception V4
TensorFlow Lite
Model: Mobilenet Quant
Caffe
Model: AlexNet - Acceleration: CPU - Iterations: 200
ASTC Encoder
Preset: Medium
OpenVINO
Model: Person Detection 0106 FP16 - Device: CPU
LuxCoreRender
Scene: Rainbow Colors and Prism
NCNN
Target: CPU - Model: alexnet
Polyhedron Fortran Benchmarks
Benchmark: capacita
NCNN
Target: CPU - Model: googlenet
AI Benchmark Alpha
Device Inference Score
Blender
Blend File: Barbershop - Compute: CPU-Only
WebP Image Encode
Encode Settings: Quality 100
rav1e
Speed: 1
AOM AV1
Encoder Mode: Speed 6 Two-Pass
Mlpack Benchmark
Benchmark: scikit_qda
Timed MAFFT Alignment
Multiple Sequence Alignment - LSU RNA
Aircrack-ng
OpenVINO
Model: Face Detection 0106 FP16 - Device: CPU
Blender
Blend File: Classroom - Compute: CPU-Only
Timed MPlayer Compilation
Time To Compile
Unpacking Firefox
Extracting: firefox-84.0.source.tar.xz
Stress-NG
Test: Glibc Qsort Data Sorting
Stress-NG
Test: Crypto
Embree
Binary: Pathtracer ISPC - Model: Asian Dragon Obj
Polyhedron Fortran Benchmarks
Benchmark: gas_dyn2
PyPerformance
Benchmark: raytrace
WebP Image Encode
Encode Settings: Quality 100, Highest Compression
rav1e
Speed: 6
TensorFlow Lite
Model: SqueezeNet
OpenVINO
Model: Person Detection 0106 FP32 - Device: CPU
TNN
Target: CPU - Model: MobileNet v2
Polyhedron Fortran Benchmarks
Benchmark: fatigue2
rav1e
Speed: 10
NCNN
Target: CPU - Model: resnet18
DeepSpeech
Acceleration: CPU
OpenSSL
RSA 4096-bit Performance
OpenVINO
Model: Age Gender Recognition Retail 0013 FP32 - Device: CPU
TensorFlow Lite
Model: Inception ResNet V2
OCRMyPDF
Processing 60 Page PDF Document
RNNoise
OpenVINO
Model: Face Detection 0106 FP32 - Device: CPU
LAMMPS Molecular Dynamics Simulator
Model: 20k Atoms
dav1d
Video Input: Summer Nature 4K
PlaidML
FP16: No - Mode: Inference - Network: Mobilenet - Device: CPU
PlaidML
FP16: No - Mode: Inference - Network: IMDB LSTM - Device: CPU
John The Ripper
Test: MD5
Open Porous Media
OPM Benchmark: Flow MPI Norne - Threads: 1
GraphicsMagick
Operation: Swirl
NCNN
Target: CPU - Model: shufflenet-v2
Numenta Anomaly Benchmark
Detector: Earthgecko Skyline
Embree
Binary: Pathtracer - Model: Asian Dragon Obj
WebP Image Encode
Encode Settings: Quality 100, Lossless, Highest Compression
oneDNN
Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPU
Polyhedron Fortran Benchmarks
Benchmark: induct2
Numenta Anomaly Benchmark
Detector: Relative Entropy
WavPack Audio Encoding
WAV To WavPack
OpenVINO
Model: Age Gender Recognition Retail 0013 FP16 - Device: CPU
LAME MP3 Encoding
WAV To MP3
Stress-NG
Test: Malloc
Stress-NG
Test: Glibc C String Functions
Polyhedron Fortran Benchmarks
Benchmark: rnflow
Blender
Blend File: Pabellon Barcelona - Compute: CPU-Only
Stress-NG
Test: Vector Math
Hugin
Panorama Photo Assistant + Stitching Time
Open Porous Media
OPM Benchmark: Flow MPI Norne - Threads: 2
Monkey Audio Encoding
WAV To APE
John The Ripper
Test: Blowfish
Opus Codec Encoding
WAV To Opus Encode
OpenVKL
Benchmark: vklBenchmarkUnstructuredVolume
Timed Clash Compilation
Time To Compile
LZ4 Compression
Compression Level: 3 - Compression Speed
ASTC Encoder
Preset: Exhaustive
Polyhedron Fortran Benchmarks
Benchmark: mp_prop_design
Stockfish
Total Time
Hierarchical INTegration
Test: FLOAT
OpenVINO
Model: Face Detection 0106 FP16 - Device: CPU
PyPerformance
Benchmark: regex_compile
PyPerformance
Benchmark: nbody
IndigoBench
Acceleration: CPU - Scene: Bedroom
OpenVINO
Model: Age Gender Recognition Retail 0013 FP16 - Device: CPU
OpenVINO
Model: Face Detection 0106 FP32 - Device: CPU
PlaidML
FP16: No - Mode: Inference - Network: NASNer Large - Device: CPU
NCNN
Target: CPU - Model: mnasnet
ASTC Encoder
Preset: Thorough
ASTC Encoder
Preset: Fast
LuxCoreRender
Scene: DLSC
Intel Open Image Denoise
Scene: Memorial
rav1e
Speed: 5
GraphicsMagick
Operation: Enhanced
GraphicsMagick
Operation: Sharpen
simdjson
Throughput Test: DistinctUserID
simdjson
Throughput Test: PartialTweets
WebP Image Encode
Encode Settings: Default
Monte Carlo Simulations of Ionised Nebulae
Input: Dust 2D tau100.0
Polyhedron Fortran Benchmarks
Benchmark: protein
Polyhedron Fortran Benchmarks
Benchmark: linpk
Polyhedron Fortran Benchmarks
Benchmark: doduc
Polyhedron Fortran Benchmarks
Benchmark: mdbx
Polyhedron Fortran Benchmarks
Benchmark: air
CLOMP
Static OMP Speedup
Sunflow Rendering System
Global Illumination + Image Synthesis
Mobile Neural Network
Model: SqueezeNetV1.0
Stress-NG
Test: CPU Cache
LAMMPS Molecular Dynamics Simulator
Model: Rhodopsin Protein
Phoronix Test Suite v10.8.4