AMD Preferred Core Linux - Ryzen 9 7950X3D

Benchmarks by Michael Larabel for a future article.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 2310092-NE-AMDPREFER06
Jump To Table - Results

View

Do Not Show Noisy Results
Do Not Show Results With Incomplete Data
Do Not Show Results With Little Change/Spread
List Notable Results
Show Result Confidence Charts
Allow Limiting Results To Certain Suite(s)

Statistics

Show Overall Harmonic Mean(s)
Show Overall Geometric Mean
Show Wins / Losses Counts (Pie Chart)
Normalize Results
Remove Outliers Before Calculating Averages

Graph Settings

Force Line Graphs Where Applicable
Convert To Scalar Where Applicable
Disable Color Branding
Prefer Vertical Bar Graphs
No Box Plots
On Line Graphs With Missing Data, Connect The Line Gaps

Multi-Way Comparison

Condense Multi-Option Tests Into Single Result Graphs
Condense Test Profiles With Multiple Version Results Into Single Result Graphs

Table

Show Detailed System Result Table

Run Management

Highlight
Result
Toggle/Hide
Result
Result
Identifier
View Logs
Performance Per
Dollar
Date
Run
  Test
  Duration
AMD Preferred Core Patched
October 07 2023
  18 Hours, 2 Minutes
AMD Preferred Core Disabled
October 08 2023
  17 Hours, 42 Minutes
Invert Behavior (Only Show Selected Data)
  17 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 Preferred Core Patched

Kernel Notes: Transparent Huge Pages: madvise
Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-XeT9lY/gcc-11-11.4.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-XeT9lY/gcc-11-11.4.0/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-build-config=bootstrap-lto-lean --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
Disk Notes: NONE / errors=remount-ro,relatime,rw,stripe=64 / Block Size: 4096
Processor Notes: Scaling Governor: amd-pstate-epp powersave (EPP: balance_performance) - CPU Microcode: 0xa601203
Java Notes: OpenJDK Runtime Environment (build 11.0.20+8-post-Ubuntu-1ubuntu122.04)
Python Notes: Python 3.10.12
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 + retbleed: Not affected + spec_rstack_overflow: Mitigation of safe RET no microcode + 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 + srbds: Not affected + tsx_async_abort: Not affected

AMD Preferred Core Disabled

Processor: AMD Ryzen 9 7950X3D 16-Core @ 5.76GHz (16 Cores / 32 Threads), Motherboard: ASRockRack B650D4U-2L2T/BCM (2.09 BIOS), Chipset: AMD Device 14d8, Memory: 2 x 32 GB DDR5-4800MT/s MTC20C2085S1EC48BA1, Disk: 3201GB Micron_7450_MTFDKCC3T2TFS + 0GB Virtual HDisk0 + 0GB Virtual HDisk1 + 0GB Virtual HDisk2 + 0GB Virtual HDisk3, Graphics: ASPEED 512MB, Audio: AMD Device 1640, Monitor: VA2431, Network: 2 x Intel I210 + 2 x Broadcom BCM57416 NetXtreme-E Dual-Media 10G RDMA

OS: Ubuntu 22.04, Kernel: 6.6.0-rc4-phx-amd-pref-core (x86_64), Desktop: GNOME Shell 42.9, Display Server: X Server, Vulkan: 1.3.238, Compiler: GCC 11.4.0, File-System: ext4, Screen Resolution: 1920x1200

AMD Preferred Core Linux - Ryzen 9 7950X3DOpenBenchmarking.orgPhoronix Test SuiteAMD Ryzen 9 7950X3D 16-Core @ 5.76GHz (16 Cores / 32 Threads)ASRockRack B650D4U-2L2T/BCM (2.09 BIOS)AMD Device 14d82 x 32 GB DDR5-4800MT/s MTC20C2085S1EC48BA13201GB Micron_7450_MTFDKCC3T2TFS + 0GB Virtual HDisk0 + 0GB Virtual HDisk1 + 0GB Virtual HDisk2 + 0GB Virtual HDisk3ASPEED 512MBAMD Device 1640VA24312 x Intel I210 + 2 x Broadcom BCM57416 NetXtreme-E Dual-Media 10G RDMAUbuntu 22.046.6.0-rc4-phx-amd-pref-core (x86_64)GNOME Shell 42.9X Server1.3.238GCC 11.4.0ext41920x1200ProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerVulkanCompilerFile-SystemScreen ResolutionAMD Preferred Core Linux - Ryzen 9 7950X3D BenchmarksSystem Logs- Transparent Huge Pages: madvise- --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-XeT9lY/gcc-11-11.4.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-XeT9lY/gcc-11-11.4.0/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-build-config=bootstrap-lto-lean --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 - NONE / errors=remount-ro,relatime,rw,stripe=64 / Block Size: 4096- Scaling Governor: amd-pstate-epp powersave (EPP: balance_performance) - CPU Microcode: 0xa601203- OpenJDK Runtime Environment (build 11.0.20+8-post-Ubuntu-1ubuntu122.04)- Python 3.10.12- gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of safe RET no microcode + 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 + srbds: Not affected + tsx_async_abort: Not affected

AMD Preferred Core Patched vs. AMD Preferred Core Disabled ComparisonPhoronix Test SuiteBaseline+3.6%+3.6%+7.2%+7.2%+10.8%+10.8%14.4%10.2%10%9.2%6.9%5.6%5.2%5.2%5.1%4.8%4.4%4.2%3.8%3.2%3.2%3.1%3.1%3%3%2.7%2.4%2.1%192000 - 1024Speed 10 Realtime - Bosphorus 4K192000 - 512Speed 11 Realtime - Bosphorus 1080pSpeed 9 Realtime - Bosphorus 1080pSpeed 11 Realtime - Bosphorus 4Koltp_update_non_index - 32Scala Dottyoltp_update_index - 16oltp_update_non_index - 16Speed 8 Realtime - Bosphorus 1080poltp_read_write - 16Speed 6 Realtime - Bosphorus 1080p4.1%Update Randcrypto_pyaes3.8%All3.6%Delete - 20 - 1000003.5%Open - 50 - 1000003.4%A.U.C.T3.4%Speed 6 Realtime - Bosphorus 4KSpeed 6 Two-Pass - Bosphorus 1080pjson_loads3.2%Tradesoap2Bosphorus 1080p3%F.H.R32T.F.A.T.T2.9%Open - 20 - 10000002.7%resize2.7%Server Rack - CPU-onlychaos2.6%Jython2.6%Bosphorus 1080p - Fast2.4%Apache Spark ALS2.4%A.G.R.R.0.F - CPUP.B.S2.3%6, Lossless2.3%H22.3%Bosphorus 1080p - Medium2.3%82.2%Bosphorus 1080p - Slow2.1%JPEG - 902.1%Boat - CPU-onlyBosphorus 1080p - Medium2%nbody2%pathlib2%Stargate Digital Audio WorkstationAOM AV1Stargate Digital Audio WorkstationAOM AV1AOM AV1AOM AV1TiDB Community ServerRenaissanceTiDB Community ServerTiDB Community ServerAOM AV1TiDB Community ServerAOM AV1RocksDBPyPerformanceJPEG XL Decoding libjxlApache HadoopApache HadoopRenaissanceAOM AV1AOM AV1PyPerformanceDaCapo BenchmarkSQLitex265RenaissanceSQLitePyBenchApache HadoopGIMPDarktablePyPerformanceDaCapo BenchmarkVVenCRenaissanceOpenVINOPHPBenchlibavif avifencDaCapo BenchmarkKvazaarSQLiteuvg266JPEG XL libjxlDarktableuvg266PyPerformancePyPerformanceAMD Preferred Core PatchedAMD Preferred Core Disabled

AMD Preferred Core Linux - Ryzen 9 7950X3Dstargate: 192000 - 1024stargate: 192000 - 512aom-av1: Speed 11 Realtime - Bosphorus 4Ktidb: oltp_update_non_index - 32tidb: oltp_update_index - 16tidb: oltp_update_non_index - 16tidb: oltp_read_write - 16rocksdb: Update Randpyperformance: crypto_pyaesjpegxl-decode: Allhadoop: Delete - 20 - 100000renaissance: Akka Unbalanced Cobwebbed Treeaom-av1: Speed 6 Two-Pass - Bosphorus 1080ppyperformance: json_loadsdacapobench: Tradesoapsqlite: 2x265: Bosphorus 1080prenaissance: Finagle HTTP Requestssqlite: 32pybench: Total For Average Test Timeshadoop: Open - 20 - 1000000gimp: resizepyperformance: chaosdacapobench: Jythonvvenc: Bosphorus 1080p - Fastrenaissance: Apache Spark ALSopenvino: Age Gender Recognition Retail 0013 FP16 - CPUphpbench: PHP Benchmark Suiteavifenc: 6, Losslessdacapobench: H2kvazaar: Bosphorus 1080p - Mediumsqlite: 8uvg266: Bosphorus 1080p - Slowjpegxl: JPEG - 90darktable: Boat - CPU-onlyuvg266: Bosphorus 1080p - Mediumpyperformance: nbodypyperformance: pathlibopenvino: Face Detection Retail FP16-INT8 - CPUrocksdb: Seq Fillpyperformance: raytraceapache: 100liquid-dsp: 1 - 256 - 512apache-iotdb: 100 - 100 - 200 - 100kvazaar: Bosphorus 1080p - Slowliquid-dsp: 8 - 256 - 512darktable: Server Room - CPU-onlypyperformance: pickle_pure_pythoncouchdb: 500 - 1000 - 30nginx: 1000avifenc: 10, Losslessaom-av1: Speed 0 Two-Pass - Bosphorus 1080ppyperformance: django_templateliquid-dsp: 2 - 256 - 512pyperformance: gorenaissance: Genetic Algorithm Using Jenetics + Futuresrenaissance: Savina Reactors.IOhadoop: Create - 20 - 100000openvino: Person Vehicle Bike Detection FP16 - CPUcouchdb: 100 - 3000 - 30pyperformance: regex_compilecouchdb: 300 - 3000 - 30vvenc: Bosphorus 4K - Fasthadoop: Delete - 50 - 100000build-ffmpeg: Time To Compileopenvino: Person Vehicle Bike Detection FP16 - CPUkvazaar: Bosphorus 1080p - Very Fastjpegxl-decode: 1avifenc: 6john-the-ripper: MD5couchdb: 300 - 1000 - 30jpegxl: JPEG - 100gimp: rotatesqlite: 16gcrypt: renaissance: Rand Forestopenradioss: Cell Phone Drop Testjpegxl: PNG - 100hadoop: Create - 50 - 100000vvenc: Bosphorus 1080p - Fasteraom-av1: Speed 4 Two-Pass - Bosphorus 4Kpyperformance: 2to3rocksdb: Rand Filltidb: oltp_update_non_index - 1apache-iotdb: 100 - 100 - 200 - 100avifenc: 0avifenc: 2sqlite-speedtest: Timed Time - Size 1,000svt-av1: Preset 4 - Bosphorus 4Krenaissance: Apache Spark PageRankredis: SET - 50gimp: unsharp-maskpyperformance: floatdeepsparse: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Synchronous Single-Streamdeepsparse: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Synchronous Single-Streamjpegxl: PNG - 90john-the-ripper: HMAC-SHA512openvino: Vehicle Detection FP16 - CPUtidb: oltp_point_select - 1vvenc: Bosphorus 4K - Fasterapache: 200openvino: Vehicle Detection FP16 - CPUopenradioss: Rubber O-Ring Seal Installationdeepsparse: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Asynchronous Multi-Streamkvazaar: Bosphorus 4K - Mediumdeepsparse: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Asynchronous Multi-Streamopenvino: Face Detection Retail FP16 - CPUtensorflow: CPU - 16 - ResNet-50apache: 500rocksdb: Read Rand Write Randdeepsparse: NLP Text Classification, BERT base uncased SST2 - Synchronous Single-Streamdeepsparse: NLP Text Classification, BERT base uncased SST2 - Synchronous Single-Streamnginx: 200kvazaar: Bosphorus 4K - Slowopenvino: Age Gender Recognition Retail 0013 FP16 - CPUdeepsparse: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Synchronous Single-Streamdeepsparse: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Synchronous Single-Streamaom-av1: Speed 4 Two-Pass - Bosphorus 1080pcouchdb: 100 - 1000 - 30sqlite: 4gimp: auto-levelsopenvino: Face Detection Retail FP16-INT8 - CPUopenvino: Age Gender Recognition Retail 0013 FP16-INT8 - CPUdeepsparse: ResNet-50, Sparse INT8 - Synchronous Single-Streamdeepsparse: ResNet-50, Sparse INT8 - Synchronous Single-Streamffmpeg: libx265 - Liveuvg266: Bosphorus 1080p - Very Fastuvg266: Bosphorus 1080p - Super Fastffmpeg: libx265 - Livenginx: 100darktable: Masskrug - CPU-onlyopenvino: Person Detection FP32 - CPUrocksdb: Read While Writingopenvino: Person Detection FP32 - CPUkvazaar: Bosphorus 1080p - Super Fastapache: 1000kvazaar: Bosphorus 4K - Ultra Fastopenvino: Face Detection FP16-INT8 - CPUkvazaar: Bosphorus 4K - Very Fastdeepsparse: CV Detection, YOLOv5s COCO, Sparse INT8 - Asynchronous Multi-Streamtensorflow: CPU - 32 - ResNet-50deepsparse: CV Detection, YOLOv5s COCO, Sparse INT8 - Asynchronous Multi-Streamuvg266: Bosphorus 4K - Slowencode-mp3: WAV To MP3openvino: Face Detection FP16-INT8 - CPUuvg266: Bosphorus 1080p - Ultra Fastdeepsparse: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Asynchronous Multi-Streamdeepsparse: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Asynchronous Multi-Streamopenradioss: Bird Strike on Windshieldrenaissance: ALS Movie Lensdacapobench: Tradebeansuvg266: Bosphorus 4K - Mediumkvazaar: Bosphorus 4K - Super Fastopenssl: SHA256openvino: Road Segmentation ADAS FP16 - CPUopenradioss: Bumper Beammemcached: 1:10openvino: Face Detection FP16 - CPUrocksdb: Rand Readdeepsparse: CV Detection, YOLOv5s COCO - Asynchronous Multi-Streamtidb: oltp_read_write - 1openvino: Road Segmentation ADAS FP16 - CPUopenvino: Road Segmentation ADAS FP16-INT8 - CPUdeepsparse: CV Detection, YOLOv5s COCO - Asynchronous Multi-Streamopenvino: Road Segmentation ADAS FP16-INT8 - CPUopenvino: Face Detection FP16 - CPUdeepsparse: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Asynchronous Multi-Streamliquid-dsp: 32 - 256 - 512openssl: ChaCha20kvazaar: Bosphorus 1080p - Ultra Fastrenaissance: Apache Spark Bayesdeepsparse: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Asynchronous Multi-Streamopenvino: Handwritten English Recognition FP16 - CPUopenvino: Vehicle Detection FP16-INT8 - CPUblender: Pabellon Barcelona - CPU-Onlyffmpeg: libx265 - Video On Demandffmpeg: libx265 - Video On Demanduvg266: Bosphorus 4K - Very Fastopenssl: ChaCha20-Poly1305svt-av1: Preset 12 - Bosphorus 4Kopenvino: Weld Porosity Detection FP16-INT8 - CPUopenssl: AES-256-GCMjohn-the-ripper: WPA PSKx265: Bosphorus 4Kopenvino: Handwritten English Recognition FP16 - CPUdeepsparse: BERT-Large, NLP Question Answering - Synchronous Single-Streamdeepsparse: BERT-Large, NLP Question Answering - Synchronous Single-Streamdeepsparse: BERT-Large, NLP Question Answering, Sparse INT8 - Synchronous Single-Streamdeepsparse: BERT-Large, NLP Question Answering, Sparse INT8 - Synchronous Single-Streamblender: Fishy Cat - CPU-Onlyhadoop: Create - 20 - 1000000openssl: RSA4096liquid-dsp: 4 - 256 - 512renaissance: In-Memory Database Shootoutbuild-linux-kernel: defconfigdeepsparse: BERT-Large, NLP Question Answering, Sparse INT8 - Asynchronous Multi-Streamopenvino: Weld Porosity Detection FP16-INT8 - CPUhadoop: Open - 20 - 100000deepsparse: BERT-Large, NLP Question Answering, Sparse INT8 - Asynchronous Multi-Streampyperformance: python_startupdeepsparse: NLP Token Classification, BERT base uncased conll2003 - Synchronous Single-Streamdeepsparse: NLP Token Classification, BERT base uncased conll2003 - Synchronous Single-Streamopenradioss: INIVOL and Fluid Structure Interaction Drop Containerdeepsparse: CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Streamdeepsparse: NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Streamhadoop: Delete - 20 - 1000000openvino: Handwritten English Recognition FP16-INT8 - CPUdeepsparse: CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Streamdeepsparse: NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Streamdeepsparse: ResNet-50, Baseline - Synchronous Single-Streamdeepsparse: ResNet-50, Baseline - Synchronous Single-Streamsvt-av1: Preset 8 - Bosphorus 4Kunpack-firefox: firefox-84.0.source.tar.xzjohn-the-ripper: bcryptdeepsparse: NLP Text Classification, DistilBERT mnli - Synchronous Single-Streamuvg266: Bosphorus 4K - Ultra Fastdeepsparse: NLP Text Classification, DistilBERT mnli - Synchronous Single-Streamopenvino: Handwritten English Recognition FP16-INT8 - CPUnginx: 500uvg266: Bosphorus 4K - Super Fastdeepsparse: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Synchronous Single-Streamdeepsparse: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Synchronous Single-Streamdeepsparse: NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Streamdeepsparse: CV Detection, YOLOv5s COCO - Synchronous Single-Streamdeepsparse: CV Detection, YOLOv5s COCO - Synchronous Single-Streamdeepsparse: ResNet-50, Sparse INT8 - Asynchronous Multi-Streamdeepsparse: BERT-Large, NLP Question Answering - Asynchronous Multi-Streamhadoop: Delete - 50 - 1000000deepsparse: ResNet-50, Sparse INT8 - Asynchronous Multi-Streamblender: BMW27 - CPU-Onlydeepsparse: BERT-Large, NLP Question Answering - Asynchronous Multi-Streamdeepsparse: ResNet-50, Baseline - Asynchronous Multi-Streamdeepsparse: ResNet-50, Baseline - Asynchronous Multi-Streamdeepsparse: NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Streamdeepsparse: NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Streamopenradioss: Chrysler Neon 1Mdeepsparse: CV Detection, YOLOv5s COCO, Sparse INT8 - Synchronous Single-Streamopenvino: Weld Porosity Detection FP16 - CPUdeepsparse: CV Detection, YOLOv5s COCO, Sparse INT8 - Synchronous Single-Streamopenvino: Machine Translation EN To DE FP16 - CPUdeepsparse: CV Classification, ResNet-50 ImageNet - Synchronous Single-Streamopenssl: AES-128-GCMopenssl: RSA4096john-the-ripper: Blowfishdeepsparse: CV Classification, ResNet-50 ImageNet - Synchronous Single-Streammemcached: 1:100openvino: Machine Translation EN To DE FP16 - CPUdeepsparse: NLP Text Classification, BERT base uncased SST2 - Asynchronous Multi-Streamdeepsparse: NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Streamopenssl: SHA512ffmpeg: libx265 - Platformblender: Classroom - CPU-Onlydeepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Asynchronous Multi-Streamffmpeg: libx265 - Platformliquid-dsp: 16 - 256 - 512deepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Synchronous Single-Streamdeepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Synchronous Single-Streamopenvino: Person Detection FP16 - CPUdeepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Asynchronous Multi-Streamopenvino: Person Detection FP16 - CPUdeepsparse: NLP Document Classification, oBERT base uncased on IMDB - Synchronous Single-Streamdeepsparse: NLP Document Classification, oBERT base uncased on IMDB - Synchronous Single-Streamdeepsparse: NLP Text Classification, BERT base uncased SST2 - Asynchronous Multi-Streamffmpeg: libx265 - Uploadffmpeg: libx265 - Uploadaom-av1: Speed 0 Two-Pass - Bosphorus 4Kaom-av1: Speed 6 Two-Pass - Bosphorus 4Kopenvino: Face Detection Retail FP16 - CPUopenvino: Vehicle Detection FP16-INT8 - CPUopenvino: Weld Porosity Detection FP16 - CPUopenvino: Age Gender Recognition Retail 0013 FP16-INT8 - CPUrenaissance: Scala Dottyaom-av1: Speed 6 Realtime - Bosphorus 4Kaom-av1: Speed 6 Realtime - Bosphorus 1080paom-av1: Speed 8 Realtime - Bosphorus 4Kaom-av1: Speed 8 Realtime - Bosphorus 1080paom-av1: Speed 9 Realtime - Bosphorus 4Kaom-av1: Speed 9 Realtime - Bosphorus 1080paom-av1: Speed 10 Realtime - Bosphorus 4Kaom-av1: Speed 10 Realtime - Bosphorus 1080paom-av1: Speed 11 Realtime - Bosphorus 1080phadoop: Open - 50 - 100000svt-av1: Preset 13 - Bosphorus 4Kdarktable: Server Rack - CPU-onlyopenfoam: drivaerFastback, Small Mesh SizeAMD Preferred Core PatchedAMD Preferred Core Disabled3.3754733.064180104.543142714845217864219582572365.9247.321206109494.172.6312.518712.164112.242759.210.922659173518516.82857.0257422.4052133.50.4311064066.241194484.793.58155.549.702.72661.9584.815.34397.121349300272392066.9516513000121.3681.781305300002.673245165.306141227.153.9791.2926.9326673331391247.34721.0682755.06259.85487.0402.7637.58511769322.1071573.81175.0755.613.6134450000119.1770.8112.7955.326164.883437.252.300.876847441.23711.30189127067719721327239175.83338.31147.7265.1332159.83633880.2517.15665.16.0486165.213610.111860940001022.22694014.586398392.997.8168.588.016921.52996.33673018.3139.43330962.46311279118.516053.9978144792.620.9933562.19276.05353.620524.3377.6042.53414.0803.4947818.160.82681206.2497175.22160.60174.4328.82143486.293.11794.82412358684.28231.30328237.0777.59313.4645.6465.282840.39122.451013.895.35225.48194.21395.901620.1924150.509094.9185015.6059.9233294033313436.5491.033765364.8313.3113053017467.8363400418.29522.57117.849815.28598.44112.1971371413333125463772453278.79903.271.240321.421605.40169.6768.08111.2742.1389227109407169.5792581.0721026698781715973132.26746.0518.858753.0203112.38968.895068.4685764357968.1653130002531.243.16717.31176.12820810461.71686.2057.544817.3763273.77266.8739180.3396137039588.5429.955944.33465.1765193.111379.64914.853419849.056253.60110.373927.14146317.3145.8616.589260.2586372.418589.406611.18272.650126.92451304093008.218755.10296.4755266.862229.9570372.357721.4631808.4811.04701344.2690.5116129.975.183524615376313314041.142158192.85643746607.0561.4487.103421.461210581245807111.76140.79209.931567.7825230000028.968934.505984.1338.093195.0117.372457.557391.817776.0633.200.4124.852.524.9311.860.29565.895.02219.5291.53216.90100.66226.64102.48238.89221.27757585164.4350.1163.8624343.370842110.353306615603228274398485733468.4238.771165219812.574.9612.918142.099108.962678.910.608678168934217.28358.5264021.8792184.30.4210810186.383198882.923.65854.399.502.67160.7286.515.64480.761374435277399046.1816227333123.4980.381283600002.629249162.667139003.774.0421.2727.3331503331371229.74788.4673234.99256.29088.2397.4467.48611615921.8221594.25172.8454.913.6594506333117.7050.8012.9525.264166.821442.351.700.886769740.79611.18191128357219921314273876.58138.67948.1825.1812179.93667648.0017.31565.76.1030163.741310.021844690001013.34699714.469401516.087.8769.108.077321.36988.96993040.7039.72333352.80313507618.386754.3721143802.1120.8533780.37277.83453.597424.1877.1432.54914.1633.4748090.160.82231212.7974176.16159.75173.5128.67144206.533.10295.24410594383.93230.35329562.5877.29312.2645.4765.525640.54122.009013.845.37125.57193.53394.549220.2615150.009125.2184415.5559.7333399261110435.1890.753776936.5113.3513090042168.0284399318.34523.99117.532015.24596.92111.9156372340000125775143277278.10905.471.410221.371609.06170.0567.93111.5142.0489411023847169.2332586.2020985752835316003132.20747.4018.892752.9252112.58858.879368.5885615358591.0652006672526.943.09417.28296.13822149462.46986.2157.455917.4031274.18267.2616180.5948136850589.3029.917744.27855.1828192.877779.55614.836420329.045953.54110.496227.11146476.9845.9116.572660.3182372.052389.320311.19342.647626.94961302893010.961655.05296.2199267.075229.9339372.634021.4480809.0311.05441345.1690.4514129.895.180424630029162014049.142182192.96413748498.2161.4787.066021.469110577375097111.72140.74209.998067.8025223000028.976834.496684.1138.084895.0317.375957.545991.830376.0533.200.4124.852.524.9311.860.29538.098.09210.8392.15226.48102.43242.35112.90236.30241.52732532164.9990.113OpenBenchmarking.org

Meta Performance Per Watts

OpenBenchmarking.orgPerformance Per Watts, More Is BetterMeta Performance Per WattsPerformance Per WattsAMD Preferred Core DisabledAMD Preferred Core Patched150030004500600075001025.346952.94

Stargate Digital Audio Workstation

Stargate is an open-source, cross-platform digital audio workstation (DAW) software package with "a unique and carefully curated experience" with scalability from old systems up through modern multi-core systems. Stargate is GPLv3 licensed and makes use of Qt5 (PyQt5) for its user-interface. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRender Ratio, More Is BetterStargate Digital Audio Workstation 22.11.5Sample Rate: 192000 - Buffer Size: 1024AMD Preferred Core DisabledAMD Preferred Core Patched0.8691.7382.6073.4764.345SE +/- 0.036937, N = 3SE +/- 0.016608, N = 33.8624343.3754731. (CXX) g++ options: -lpthread -lsndfile -lm -O3 -march=native -ffast-math -funroll-loops -fstrength-reduce -fstrict-aliasing -finline-functions
OpenBenchmarking.orgRender Ratio Per Watt, More Is BetterStargate Digital Audio Workstation 22.11.5Sample Rate: 192000 - Buffer Size: 1024AMD Preferred Core DisabledAMD Preferred Core Patched0.01060.02120.03180.04240.0530.0470.045
MinAvgMaxAMD Preferred Core Disabled12.982.199.0AMD Preferred Core Patched12.875.197.5OpenBenchmarking.orgWatts, Fewer Is BetterStargate Digital Audio Workstation 22.11.5CPU Power Consumption Monitor20406080100

OpenBenchmarking.orgRender Ratio, More Is BetterStargate Digital Audio Workstation 22.11.5Sample Rate: 192000 - Buffer Size: 512AMD Preferred Core DisabledAMD Preferred Core Patched0.75841.51682.27523.03363.792SE +/- 0.044014, N = 3SE +/- 0.031164, N = 153.3708423.0641801. (CXX) g++ options: -lpthread -lsndfile -lm -O3 -march=native -ffast-math -funroll-loops -fstrength-reduce -fstrict-aliasing -finline-functions
OpenBenchmarking.orgRender Ratio Per Watt, More Is BetterStargate Digital Audio Workstation 22.11.5Sample Rate: 192000 - Buffer Size: 512AMD Preferred Core DisabledAMD Preferred Core Patched0.00970.01940.02910.03880.04850.0430.041
MinAvgMaxAMD Preferred Core Disabled11.179.199.0AMD Preferred Core Patched10.274.595.5OpenBenchmarking.orgWatts, Fewer Is BetterStargate Digital Audio Workstation 22.11.5CPU Power Consumption Monitor20406080100

AOM AV1

This is a test of the AOMedia AV1 encoder (libaom) developed by AOMedia and Google as the AV1 Codec Library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.7Encoder Mode: Speed 11 Realtime - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched20406080100SE +/- 1.64, N = 15SE +/- 1.52, N = 15110.35104.541. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterAOM AV1 3.7Encoder Mode: Speed 11 Realtime - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched0.66421.32841.99262.65683.3212.9522.802
MinAvgMaxAMD Preferred Core Disabled12.037.456.6AMD Preferred Core Patched11.437.356.7OpenBenchmarking.orgWatts, Fewer Is BetterAOM AV1 3.7CPU Power Consumption Monitor1632486480

TiDB Community Server

This is a PingCAP TiDB Community Server benchmark facilitated using the sysbench OLTP database benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Second, More Is BetterTiDB Community Server 7.3Test: oltp_update_non_index - Threads: 32AMD Preferred Core DisabledAMD Preferred Core Patched7K14K21K28K35KSE +/- 44.79, N = 3SE +/- 121.34, N = 33306631427
OpenBenchmarking.orgQueries Per Second Per Watt, More Is BetterTiDB Community Server 7.3Test: oltp_update_non_index - Threads: 32AMD Preferred Core DisabledAMD Preferred Core Patched110220330440550499.52467.76
OpenBenchmarking.orgWatts, Fewer Is BetterTiDB Community Server 7.3CPU Power Consumption MonitorAMD Preferred Core DisabledAMD Preferred Core Patched20406080100Min: 13.44 / Avg: 66.2 / Max: 90.27Min: 12.12 / Avg: 67.19 / Max: 92.06

OpenBenchmarking.orgQueries Per Second, More Is BetterTiDB Community Server 7.3Test: oltp_update_index - Threads: 16AMD Preferred Core DisabledAMD Preferred Core Patched3K6K9K12K15KSE +/- 216.96, N = 3SE +/- 31.97, N = 31560314845
OpenBenchmarking.orgQueries Per Second Per Watt, More Is BetterTiDB Community Server 7.3Test: oltp_update_index - Threads: 16AMD Preferred Core DisabledAMD Preferred Core Patched50100150200250246.27229.93
OpenBenchmarking.orgWatts, Fewer Is BetterTiDB Community Server 7.3CPU Power Consumption MonitorAMD Preferred Core DisabledAMD Preferred Core Patched20406080100Min: 12.41 / Avg: 63.36 / Max: 90.31Min: 13.37 / Avg: 64.56 / Max: 92.7

OpenBenchmarking.orgQueries Per Second, More Is BetterTiDB Community Server 7.3Test: oltp_update_non_index - Threads: 16AMD Preferred Core DisabledAMD Preferred Core Patched5K10K15K20K25KSE +/- 12.01, N = 3SE +/- 11.34, N = 32282721786
OpenBenchmarking.orgQueries Per Second Per Watt, More Is BetterTiDB Community Server 7.3Test: oltp_update_non_index - Threads: 16AMD Preferred Core DisabledAMD Preferred Core Patched80160240320400380.79361.77
OpenBenchmarking.orgWatts, Fewer Is BetterTiDB Community Server 7.3CPU Power Consumption MonitorAMD Preferred Core DisabledAMD Preferred Core Patched1530456075Min: 11.3 / Avg: 59.95 / Max: 79.26Min: 12.49 / Avg: 60.22 / Max: 80.9

OpenBenchmarking.orgQueries Per Second, More Is BetterTiDB Community Server 7.3Test: oltp_read_write - Threads: 16AMD Preferred Core DisabledAMD Preferred Core Patched9K18K27K36K45KSE +/- 33.56, N = 3SE +/- 51.44, N = 34398442195
OpenBenchmarking.orgQueries Per Second Per Watt, More Is BetterTiDB Community Server 7.3Test: oltp_read_write - Threads: 16AMD Preferred Core DisabledAMD Preferred Core Patched140280420560700663.27626.15
OpenBenchmarking.orgWatts, Fewer Is BetterTiDB Community Server 7.3CPU Power Consumption MonitorAMD Preferred Core DisabledAMD Preferred Core Patched20406080100Min: 8.28 / Avg: 66.31 / Max: 90.61Min: 12.3 / Avg: 67.39 / Max: 92.08

RocksDB

This is a benchmark of Meta/Facebook's RocksDB as an embeddable persistent key-value store for fast storage based on Google's LevelDB. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 8.0Test: Update RandomAMD Preferred Core DisabledAMD Preferred Core Patched200K400K600K800K1000KSE +/- 944.58, N = 3SE +/- 1790.51, N = 38573348257231. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread
OpenBenchmarking.orgOp/s Per Watt, More Is BetterRocksDB 8.0Test: Update RandomAMD Preferred Core DisabledAMD Preferred Core Patched2K4K6K8K10K9745.009412.03
MinAvgMaxAMD Preferred Core Disabled11.888.095.5AMD Preferred Core Patched12.787.795.7OpenBenchmarking.orgWatts, Fewer Is BetterRocksDB 8.0CPU Power Consumption Monitor20406080100

PyPerformance

PyPerformance is the reference Python performance benchmark suite. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: crypto_pyaesAMD Preferred Core DisabledAMD Preferred Core Patched1530456075SE +/- 0.32, N = 3SE +/- 0.17, N = 368.465.9
MinAvgMaxAMD Preferred Core Disabled10.826.128.6AMD Preferred Core Patched10.326.028.9OpenBenchmarking.orgWatts, Fewer Is BetterPyPerformance 1.0.0CPU Power Consumption Monitor918273645

JPEG XL Decoding libjxl

The JPEG XL Image Coding System is designed to provide next-generation JPEG image capabilities with JPEG XL offering better image quality and compression over legacy JPEG. This test profile is suited for JPEG XL decode performance testing to PNG output file, the pts/jpexl test is for encode performance. The JPEG XL encoding/decoding is done using the libjxl codebase. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterJPEG XL Decoding libjxl 0.7CPU Threads: AllAMD Preferred Core DisabledAMD Preferred Core Patched50100150200250SE +/- 2.62, N = 5SE +/- 2.67, N = 4238.77247.32
OpenBenchmarking.orgMP/s Per Watt, More Is BetterJPEG XL Decoding libjxl 0.7CPU Threads: AllAMD Preferred Core DisabledAMD Preferred Core Patched1.2042.4083.6124.8166.025.2305.351
MinAvgMaxAMD Preferred Core Disabled11.345.758.2AMD Preferred Core Patched11.446.259.5OpenBenchmarking.orgWatts, Fewer Is BetterJPEG XL Decoding libjxl 0.7CPU Power Consumption Monitor1632486480

Apache Hadoop

This is a benchmark of the Apache Hadoop making use of its built-in name-node throughput benchmark (NNThroughputBenchmark). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps per sec, More Is BetterApache Hadoop 3.3.6Operation: Delete - Threads: 20 - Files: 100000AMD Preferred Core DisabledAMD Preferred Core Patched30K60K90K120K150KSE +/- 454.85, N = 4SE +/- 867.17, N = 4116521120610
OpenBenchmarking.orgOps per sec Per Watt, More Is BetterApache Hadoop 3.3.6Operation: Delete - Threads: 20 - Files: 100000AMD Preferred Core DisabledAMD Preferred Core Patched90018002700360045003987.324137.27
MinAvgMaxAMD Preferred Core Disabled10.129.248.7AMD Preferred Core Patched10.729.251.7OpenBenchmarking.orgWatts, Fewer Is BetterApache Hadoop 3.3.6CPU Power Consumption Monitor1530456075

Renaissance

Renaissance is a suite of benchmarks designed to test the Java JVM from Apache Spark to a Twitter-like service to Scala and other features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.14Test: Akka Unbalanced Cobwebbed TreeAMD Preferred Core DisabledAMD Preferred Core Patched2K4K6K8K10KSE +/- 52.46, N = 3SE +/- 57.58, N = 39812.59494.1MIN: 7143.92 / MAX: 9885.13MIN: 7087.34 / MAX: 9603.99
MinAvgMaxAMD Preferred Core Disabled10.757.873.9AMD Preferred Core Patched12.559.180.0OpenBenchmarking.orgWatts, Fewer Is BetterRenaissance 0.14CPU Power Consumption Monitor20406080100

AOM AV1

This is a test of the AOMedia AV1 encoder (libaom) developed by AOMedia and Google as the AV1 Codec Library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.7Encoder Mode: Speed 6 Two-Pass - Input: Bosphorus 1080pAMD Preferred Core DisabledAMD Preferred Core Patched20406080100SE +/- 0.84, N = 4SE +/- 0.13, N = 474.9672.631. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterAOM AV1 3.7Encoder Mode: Speed 6 Two-Pass - Input: Bosphorus 1080pAMD Preferred Core DisabledAMD Preferred Core Patched0.32810.65620.98431.31241.64051.4581.394
MinAvgMaxAMD Preferred Core Disabled11.851.482.2AMD Preferred Core Patched11.552.180.7OpenBenchmarking.orgWatts, Fewer Is BetterAOM AV1 3.7CPU Power Consumption Monitor20406080100

PyPerformance

PyPerformance is the reference Python performance benchmark suite. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: json_loadsAMD Preferred Core DisabledAMD Preferred Core Patched3691215SE +/- 0.03, N = 3SE +/- 0.00, N = 312.912.5
MinAvgMaxAMD Preferred Core Disabled10.525.928.4AMD Preferred Core Patched12.025.928.4OpenBenchmarking.orgWatts, Fewer Is BetterPyPerformance 1.0.0CPU Power Consumption Monitor918273645

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 9.12-MR1Java Test: TradesoapAMD Preferred Core DisabledAMD Preferred Core Patched400800120016002000SE +/- 11.12, N = 7SE +/- 16.08, N = 618141871
MinAvgMaxAMD Preferred Core Disabled11.735.289.3AMD Preferred Core Patched10.434.989.7OpenBenchmarking.orgWatts, Fewer Is BetterDaCapo Benchmark 9.12-MR1CPU Power Consumption Monitor20406080100

SQLite

This is a simple benchmark of SQLite. At present this test profile just measures the time to perform a pre-defined number of insertions on an indexed database with a variable number of concurrent repetitions -- up to the maximum number of CPU threads available. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterSQLite 3.41.2Threads / Copies: 2AMD Preferred Core DisabledAMD Preferred Core Patched0.48690.97381.46071.94762.4345SE +/- 0.009, N = 11SE +/- 0.008, N = 32.0992.1641. (CC) gcc options: -O2 -lz -lm
MinAvgMaxAMD Preferred Core Disabled11.218.529.1AMD Preferred Core Patched10.718.530.8OpenBenchmarking.orgWatts, Fewer Is BetterSQLite 3.41.2CPU Power Consumption Monitor918273645

x265

This is a simple test of the x265 encoder run on the CPU with 1080p and 4K options for H.265 video encode performance with x265. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betterx265 3.4Video Input: Bosphorus 1080pAMD Preferred Core DisabledAMD Preferred Core Patched306090120150SE +/- 0.25, N = 7SE +/- 0.20, N = 7108.96112.241. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
OpenBenchmarking.orgFrames Per Second Per Watt, More Is Betterx265 3.4Video Input: Bosphorus 1080pAMD Preferred Core DisabledAMD Preferred Core Patched0.52041.04081.56122.08162.6022.2632.313
MinAvgMaxAMD Preferred Core Disabled12.848.176.7AMD Preferred Core Patched11.248.579.3OpenBenchmarking.orgWatts, Fewer Is Betterx265 3.4CPU Power Consumption Monitor20406080100

Renaissance

Renaissance is a suite of benchmarks designed to test the Java JVM from Apache Spark to a Twitter-like service to Scala and other features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.14Test: Finagle HTTP RequestsAMD Preferred Core DisabledAMD Preferred Core Patched6001200180024003000SE +/- 13.46, N = 3SE +/- 19.94, N = 32678.92759.2MIN: 2512.79 / MAX: 2737.41MIN: 2595.33 / MAX: 2812.53
MinAvgMaxAMD Preferred Core Disabled11.088.2104.0AMD Preferred Core Patched12.287.6102.8OpenBenchmarking.orgWatts, Fewer Is BetterRenaissance 0.14CPU Power Consumption Monitor20406080100

SQLite

This is a simple benchmark of SQLite. At present this test profile just measures the time to perform a pre-defined number of insertions on an indexed database with a variable number of concurrent repetitions -- up to the maximum number of CPU threads available. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterSQLite 3.41.2Threads / Copies: 32AMD Preferred Core DisabledAMD Preferred Core Patched3691215SE +/- 0.06, N = 5SE +/- 0.03, N = 510.6110.921. (CC) gcc options: -O2 -lz -lm
MinAvgMaxAMD Preferred Core Disabled11.038.150.3AMD Preferred Core Patched12.438.550.1OpenBenchmarking.orgWatts, Fewer Is BetterSQLite 3.41.2CPU Power Consumption Monitor1428425670

PyBench

This test profile reports the total time of the different average timed test results from PyBench. PyBench reports average test times for different functions such as BuiltinFunctionCalls and NestedForLoops, with this total result providing a rough estimate as to Python's average performance on a given system. This test profile runs PyBench each time for 20 rounds. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyBench 2018-02-16Total For Average Test TimesAMD Preferred Core DisabledAMD Preferred Core Patched150300450600750SE +/- 4.74, N = 4SE +/- 3.46, N = 4678659
MinAvgMaxAMD Preferred Core Disabled12.125.328.9AMD Preferred Core Patched12.025.328.7OpenBenchmarking.orgWatts, Fewer Is BetterPyBench 2018-02-16CPU Power Consumption Monitor918273645

Apache Hadoop

This is a benchmark of the Apache Hadoop making use of its built-in name-node throughput benchmark (NNThroughputBenchmark). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps per sec, More Is BetterApache Hadoop 3.3.6Operation: Open - Threads: 20 - Files: 1000000AMD Preferred Core DisabledAMD Preferred Core Patched400K800K1200K1600K2000KSE +/- 11337.87, N = 3SE +/- 8246.05, N = 316893421735185
OpenBenchmarking.orgOps per sec Per Watt, More Is BetterApache Hadoop 3.3.6Operation: Open - Threads: 20 - Files: 1000000AMD Preferred Core DisabledAMD Preferred Core Patched10K20K30K40K50K46198.8047024.83
MinAvgMaxAMD Preferred Core Disabled12.736.651.0AMD Preferred Core Patched12.236.950.8OpenBenchmarking.orgWatts, Fewer Is BetterApache Hadoop 3.3.6CPU Power Consumption Monitor1428425670

GIMP

GIMP is an open-source image manipulaton program. This test profile will use the system-provided GIMP program otherwise on Windows relys upon a pre-packaged Windows binary from upstream GIMP.org. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterGIMP 2.10.30Test: resizeAMD Preferred Core DisabledAMD Preferred Core Patched48121620SE +/- 0.15, N = 3SE +/- 0.24, N = 317.2816.83
MinAvgMaxAMD Preferred Core Disabled12.431.543.5AMD Preferred Core Patched12.231.443.8OpenBenchmarking.orgWatts, Fewer Is BetterGIMP 2.10.30CPU Power Consumption Monitor1224364860

PyPerformance

PyPerformance is the reference Python performance benchmark suite. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: chaosAMD Preferred Core DisabledAMD Preferred Core Patched1326395265SE +/- 0.24, N = 3SE +/- 0.18, N = 358.557.0
MinAvgMaxAMD Preferred Core Disabled11.525.928.7AMD Preferred Core Patched11.226.028.7OpenBenchmarking.orgWatts, Fewer Is BetterPyPerformance 1.0.0CPU Power Consumption Monitor918273645

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 9.12-MR1Java Test: JythonAMD Preferred Core DisabledAMD Preferred Core Patched6001200180024003000SE +/- 9.47, N = 9SE +/- 8.96, N = 926402574
MinAvgMaxAMD Preferred Core Disabled12.526.060.4AMD Preferred Core Patched12.425.759.1OpenBenchmarking.orgWatts, Fewer Is BetterDaCapo Benchmark 9.12-MR1CPU Power Consumption Monitor20406080100

VVenC

VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterVVenC 1.9Video Input: Bosphorus 1080p - Video Preset: FastAMD Preferred Core DisabledAMD Preferred Core Patched510152025SE +/- 0.08, N = 3SE +/- 0.05, N = 321.8822.411. (CXX) g++ options: -O3 -flto -fno-fat-lto-objects -flto=auto
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterVVenC 1.9Video Input: Bosphorus 1080p - Video Preset: FastAMD Preferred Core DisabledAMD Preferred Core Patched0.06770.13540.20310.27080.33850.2980.301
MinAvgMaxAMD Preferred Core Disabled10.673.592.6AMD Preferred Core Patched11.774.593.9OpenBenchmarking.orgWatts, Fewer Is BetterVVenC 1.9CPU Power Consumption Monitor20406080100

Renaissance

Renaissance is a suite of benchmarks designed to test the Java JVM from Apache Spark to a Twitter-like service to Scala and other features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.14Test: Apache Spark ALSAMD Preferred Core DisabledAMD Preferred Core Patched5001000150020002500SE +/- 6.39, N = 3SE +/- 17.87, N = 32184.32133.5MIN: 2069.01 / MAX: 2413.1MIN: 2045.99 / MAX: 2277.83
MinAvgMaxAMD Preferred Core Disabled12.852.060.6AMD Preferred Core Patched10.651.561.2OpenBenchmarking.orgWatts, Fewer Is BetterRenaissance 0.14CPU Power Consumption Monitor20406080100

OpenVINO

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.1Model: Age Gender Recognition Retail 0013 FP16 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched0.09680.19360.29040.38720.484SE +/- 0.00, N = 3SE +/- 0.00, N = 30.420.43MIN: 0.22 / MAX: 7.75MIN: 0.22 / MAX: 3.91. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
MinAvgMaxAMD Preferred Core Disabled12.7112.4121.2AMD Preferred Core Patched11.8112.2121.1OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.1CPU Power Consumption Monitor4080120160200

PHPBench

PHPBench is a benchmark suite for PHP. It performs a large number of simple tests in order to bench various aspects of the PHP interpreter. PHPBench can be used to compare hardware, operating systems, PHP versions, PHP accelerators and caches, compiler options, etc. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgScore, More Is BetterPHPBench 0.8.1PHP Benchmark SuiteAMD Preferred Core DisabledAMD Preferred Core Patched200K400K600K800K1000KSE +/- 11820.78, N = 3SE +/- 10510.59, N = 310810181106406
OpenBenchmarking.orgScore Per Watt, More Is BetterPHPBench 0.8.1PHP Benchmark SuiteAMD Preferred Core DisabledAMD Preferred Core Patched9K18K27K36K45K42982.5744179.75
MinAvgMaxAMD Preferred Core Disabled11.525.228.4AMD Preferred Core Patched12.825.028.6OpenBenchmarking.orgWatts, Fewer Is BetterPHPBench 0.8.1CPU Power Consumption Monitor918273645

libavif avifenc

This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 1.0Encoder Speed: 6, LosslessAMD Preferred Core DisabledAMD Preferred Core Patched246810SE +/- 0.008, N = 6SE +/- 0.017, N = 76.3836.2411. (CXX) g++ options: -O3 -fPIC -lm
MinAvgMaxAMD Preferred Core Disabled9.347.6117.4AMD Preferred Core Patched11.747.6117.4OpenBenchmarking.orgWatts, Fewer Is Betterlibavif avifenc 1.0CPU Power Consumption Monitor4080120160200

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 9.12-MR1Java Test: H2AMD Preferred Core DisabledAMD Preferred Core Patched400800120016002000SE +/- 24.66, N = 20SE +/- 22.52, N = 2019881944
MinAvgMaxAMD Preferred Core Disabled6.525.544.9AMD Preferred Core Patched11.625.545.0OpenBenchmarking.orgWatts, Fewer Is BetterDaCapo Benchmark 9.12-MR1CPU Power Consumption Monitor1224364860

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 1080p - Video Preset: MediumAMD Preferred Core DisabledAMD Preferred Core Patched20406080100SE +/- 0.11, N = 6SE +/- 0.08, N = 682.9284.791. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterKvazaar 2.2Video Input: Bosphorus 1080p - Video Preset: MediumAMD Preferred Core DisabledAMD Preferred Core Patched0.31520.63040.94561.26081.5761.4011.396
MinAvgMaxAMD Preferred Core Disabled11.059.291.2AMD Preferred Core Patched11.660.790.9OpenBenchmarking.orgWatts, Fewer Is BetterKvazaar 2.2CPU Power Consumption Monitor20406080100

SQLite

This is a simple benchmark of SQLite. At present this test profile just measures the time to perform a pre-defined number of insertions on an indexed database with a variable number of concurrent repetitions -- up to the maximum number of CPU threads available. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterSQLite 3.41.2Threads / Copies: 8AMD Preferred Core DisabledAMD Preferred Core Patched0.82311.64622.46933.29244.1155SE +/- 0.029, N = 9SE +/- 0.026, N = 93.6583.5811. (CC) gcc options: -O2 -lz -lm
MinAvgMaxAMD Preferred Core Disabled11.725.240.5AMD Preferred Core Patched12.625.241.1OpenBenchmarking.orgWatts, Fewer Is BetterSQLite 3.41.2CPU Power Consumption Monitor1122334455

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: SlowAMD Preferred Core DisabledAMD Preferred Core Patched1224364860SE +/- 0.02, N = 5SE +/- 0.07, N = 554.3955.54
OpenBenchmarking.orgFrames Per Second Per Watt, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: SlowAMD Preferred Core DisabledAMD Preferred Core Patched0.17460.34920.52380.69840.8730.7690.776
MinAvgMaxAMD Preferred Core Disabled12.570.895.3AMD Preferred Core Patched12.971.698.1OpenBenchmarking.orgWatts, Fewer Is Betteruvg266 0.4.1CPU Power Consumption Monitor20406080100

JPEG XL libjxl

The JPEG XL Image Coding System is designed to provide next-generation JPEG image capabilities with JPEG XL offering better image quality and compression over legacy JPEG. This test profile is currently focused on the multi-threaded JPEG XL image encode performance using the reference libjxl library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterJPEG XL libjxl 0.7Input: JPEG - Quality: 90AMD Preferred Core DisabledAMD Preferred Core Patched3691215SE +/- 0.03, N = 3SE +/- 0.06, N = 39.509.701. (CXX) g++ options: -fno-rtti -funwind-tables -O3 -O2 -fPIE -pie -lm -latomic
OpenBenchmarking.orgMP/s Per Watt, More Is BetterJPEG XL libjxl 0.7Input: JPEG - Quality: 90AMD Preferred Core DisabledAMD Preferred Core Patched0.06480.12960.19440.25920.3240.2840.288
MinAvgMaxAMD Preferred Core Disabled12.733.445.7AMD Preferred Core Patched10.933.745.9OpenBenchmarking.orgWatts, Fewer Is BetterJPEG XL libjxl 0.7CPU Power Consumption Monitor1224364860

Darktable

Darktable is an open-source photography / workflow application this will use any system-installed Darktable program or on Windows will automatically download the pre-built binary from the project. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 3.8.1Test: Boat - Acceleration: CPU-onlyAMD Preferred Core DisabledAMD Preferred Core Patched0.61341.22681.84022.45363.067SE +/- 0.005, N = 9SE +/- 0.008, N = 92.6712.726
MinAvgMaxAMD Preferred Core Disabled10.039.7110.7AMD Preferred Core Patched12.239.5108.1OpenBenchmarking.orgWatts, Fewer Is BetterDarktable 3.8.1CPU Power Consumption Monitor20406080100

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: MediumAMD Preferred Core DisabledAMD Preferred Core Patched1428425670SE +/- 0.06, N = 5SE +/- 0.05, N = 560.7261.95
OpenBenchmarking.orgFrames Per Second Per Watt, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: MediumAMD Preferred Core DisabledAMD Preferred Core Patched0.1980.3960.5940.7920.990.8710.880
MinAvgMaxAMD Preferred Core Disabled12.069.795.9AMD Preferred Core Patched12.570.498.5OpenBenchmarking.orgWatts, Fewer Is Betteruvg266 0.4.1CPU Power Consumption Monitor20406080100

PyPerformance

PyPerformance is the reference Python performance benchmark suite. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: nbodyAMD Preferred Core DisabledAMD Preferred Core Patched20406080100SE +/- 0.07, N = 3SE +/- 0.67, N = 386.584.8
MinAvgMaxAMD Preferred Core Disabled13.026.728.8AMD Preferred Core Patched11.926.528.6OpenBenchmarking.orgWatts, Fewer Is BetterPyPerformance 1.0.0CPU Power Consumption Monitor918273645

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: pathlibAMD Preferred Core DisabledAMD Preferred Core Patched48121620SE +/- 0.00, N = 3SE +/- 0.03, N = 315.615.3
MinAvgMaxAMD Preferred Core Disabled12.725.527.8AMD Preferred Core Patched11.925.427.7OpenBenchmarking.orgWatts, Fewer Is BetterPyPerformance 1.0.0CPU Power Consumption Monitor918273645

OpenVINO

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.1Model: Face Detection Retail FP16-INT8 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched10002000300040005000SE +/- 4.43, N = 3SE +/- 11.27, N = 34480.764397.121. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

RocksDB

This is a benchmark of Meta/Facebook's RocksDB as an embeddable persistent key-value store for fast storage based on Google's LevelDB. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 8.0Test: Sequential FillAMD Preferred Core DisabledAMD Preferred Core Patched300K600K900K1200K1500KSE +/- 2280.82, N = 3SE +/- 2158.35, N = 3137443513493001. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread
OpenBenchmarking.orgOp/s Per Watt, More Is BetterRocksDB 8.0Test: Sequential FillAMD Preferred Core DisabledAMD Preferred Core Patched4K8K12K16K20K18745.8618402.71
MinAvgMaxAMD Preferred Core Disabled11.673.385.4AMD Preferred Core Patched10.773.385.5OpenBenchmarking.orgWatts, Fewer Is BetterRocksDB 8.0CPU Power Consumption Monitor20406080100

PyPerformance

PyPerformance is the reference Python performance benchmark suite. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: raytraceAMD Preferred Core DisabledAMD Preferred Core Patched60120180240300SE +/- 0.00, N = 3SE +/- 1.00, N = 3277272
MinAvgMaxAMD Preferred Core Disabled12.327.328.9AMD Preferred Core Patched12.227.228.8OpenBenchmarking.orgWatts, Fewer Is BetterPyPerformance 1.0.0CPU Power Consumption Monitor918273645

Apache HTTP Server

This is a test of the Apache HTTPD web server. This Apache HTTPD web server benchmark test profile makes use of the wrk program for facilitating the HTTP requests over a fixed period time with a configurable number of concurrent clients. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRequests Per Second, More Is BetterApache HTTP Server 2.4.56Concurrent Requests: 100AMD Preferred Core DisabledAMD Preferred Core Patched90K180K270K360K450KSE +/- 471.72, N = 3SE +/- 957.10, N = 3399046.18392066.951. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2
OpenBenchmarking.orgRequests Per Second Per Watt, More Is BetterApache HTTP Server 2.4.56Concurrent Requests: 100AMD Preferred Core DisabledAMD Preferred Core Patched100020003000400050004741.604674.43
MinAvgMaxAMD Preferred Core Disabled12.484.291.0AMD Preferred Core Patched12.683.990.4OpenBenchmarking.orgWatts, Fewer Is BetterApache HTTP Server 2.4.56CPU Power Consumption Monitor20406080100

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 1 - Buffer Length: 256 - Filter Length: 512AMD Preferred Core DisabledAMD Preferred Core Patched4M8M12M16M20MSE +/- 113862.98, N = 3SE +/- 155897.40, N = 316227333165130001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
OpenBenchmarking.orgsamples/s Per Watt, More Is BetterLiquid-DSP 1.6Threads: 1 - Buffer Length: 256 - Filter Length: 512AMD Preferred Core DisabledAMD Preferred Core Patched140K280K420K560K700K640629.47655340.90
MinAvgMaxAMD Preferred Core Disabled11.825.327.4AMD Preferred Core Patched11.125.227.3OpenBenchmarking.orgWatts, Fewer Is BetterLiquid-DSP 1.6CPU Power Consumption Monitor816243240

Apache IoTDB

Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgAverage Latency, Fewer Is BetterApache IoTDB 1.2Device Count: 100 - Batch Size Per Write: 100 - Sensor Count: 200 - Client Number: 100AMD Preferred Core DisabledAMD Preferred Core Patched306090120150SE +/- 2.81, N = 3SE +/- 0.89, N = 3123.49121.36MAX: 26546.31MAX: 25666.36
MinAvgMaxAMD Preferred Core Disabled11.921.185.5AMD Preferred Core Patched13.421.284.2OpenBenchmarking.orgWatts, Fewer Is BetterApache IoTDB 1.2CPU Power Consumption Monitor20406080100

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 1080p - Video Preset: SlowAMD Preferred Core DisabledAMD Preferred Core Patched20406080100SE +/- 0.12, N = 6SE +/- 0.10, N = 680.3881.781. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterKvazaar 2.2Video Input: Bosphorus 1080p - Video Preset: SlowAMD Preferred Core DisabledAMD Preferred Core Patched0.30420.60840.91261.21681.5211.3431.352
MinAvgMaxAMD Preferred Core Disabled11.359.990.1AMD Preferred Core Patched11.560.590.4OpenBenchmarking.orgWatts, Fewer Is BetterKvazaar 2.2CPU Power Consumption Monitor20406080100

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 8 - Buffer Length: 256 - Filter Length: 512AMD Preferred Core DisabledAMD Preferred Core Patched30M60M90M120M150MSE +/- 191398.36, N = 3SE +/- 75055.53, N = 31283600001305300001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
OpenBenchmarking.orgsamples/s Per Watt, More Is BetterLiquid-DSP 1.6Threads: 8 - Buffer Length: 256 - Filter Length: 512AMD Preferred Core DisabledAMD Preferred Core Patched600K1200K1800K2400K3000K2561420.642602005.80
MinAvgMaxAMD Preferred Core Disabled10.850.156.3AMD Preferred Core Patched12.050.256.2OpenBenchmarking.orgWatts, Fewer Is BetterLiquid-DSP 1.6CPU Power Consumption Monitor1632486480

Darktable

Darktable is an open-source photography / workflow application this will use any system-installed Darktable program or on Windows will automatically download the pre-built binary from the project. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 3.8.1Test: Server Room - Acceleration: CPU-onlyAMD Preferred Core DisabledAMD Preferred Core Patched0.60141.20281.80422.40563.007SE +/- 0.004, N = 9SE +/- 0.005, N = 92.6292.673
MinAvgMaxAMD Preferred Core Disabled12.730.162.3AMD Preferred Core Patched9.830.361.5OpenBenchmarking.orgWatts, Fewer Is BetterDarktable 3.8.1CPU Power Consumption Monitor20406080100

PyPerformance

PyPerformance is the reference Python performance benchmark suite. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: pickle_pure_pythonAMD Preferred Core DisabledAMD Preferred Core Patched50100150200250SE +/- 0.00, N = 3SE +/- 0.33, N = 3249245
MinAvgMaxAMD Preferred Core Disabled12.426.528.8AMD Preferred Core Patched11.726.528.7OpenBenchmarking.orgWatts, Fewer Is BetterPyPerformance 1.0.0CPU Power Consumption Monitor918273645

Apache CouchDB

This is a bulk insertion benchmark of Apache CouchDB. CouchDB is a document-oriented NoSQL database implemented in Erlang. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterApache CouchDB 3.3.2Bulk Size: 500 - Inserts: 1000 - Rounds: 30AMD Preferred Core DisabledAMD Preferred Core Patched4080120160200SE +/- 1.78, N = 4SE +/- 2.10, N = 3162.67165.311. (CXX) g++ options: -std=c++17 -lmozjs-91 -lm -lei -fPIC -MMD
MinAvgMaxAMD Preferred Core Disabled13.052.657.7AMD Preferred Core Patched11.753.358.1OpenBenchmarking.orgWatts, Fewer Is BetterApache CouchDB 3.3.2CPU Power Consumption Monitor1632486480

nginx

This is a benchmark of the lightweight Nginx HTTP(S) web-server. This Nginx web server benchmark test profile makes use of the wrk program for facilitating the HTTP requests over a fixed period time with a configurable number of concurrent clients/connections. HTTPS with a self-signed OpenSSL certificate is used by this test for local benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.23.2Connections: 1000AMD Preferred Core DisabledAMD Preferred Core Patched30K60K90K120K150KSE +/- 67.36, N = 3SE +/- 323.53, N = 3139003.77141227.151. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2
OpenBenchmarking.orgRequests Per Second Per Watt, More Is Betternginx 1.23.2Connections: 1000AMD Preferred Core DisabledAMD Preferred Core Patched300600900120015001310.351323.09
MinAvgMaxAMD Preferred Core Disabled11.3106.1115.1AMD Preferred Core Patched11.8106.7115.9OpenBenchmarking.orgWatts, Fewer Is Betternginx 1.23.2CPU Power Consumption Monitor4080120160200

libavif avifenc

This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 1.0Encoder Speed: 10, LosslessAMD Preferred Core DisabledAMD Preferred Core Patched0.90951.8192.72853.6384.5475SE +/- 0.013, N = 8SE +/- 0.012, N = 84.0423.9791. (CXX) g++ options: -O3 -fPIC -lm
MinAvgMaxAMD Preferred Core Disabled10.430.495.4AMD Preferred Core Patched12.330.384.2OpenBenchmarking.orgWatts, Fewer Is Betterlibavif avifenc 1.0CPU Power Consumption Monitor20406080100

AOM AV1

This is a test of the AOMedia AV1 encoder (libaom) developed by AOMedia and Google as the AV1 Codec Library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.7Encoder Mode: Speed 0 Two-Pass - Input: Bosphorus 1080pAMD Preferred Core DisabledAMD Preferred Core Patched0.29030.58060.87091.16121.4515SE +/- 0.00, N = 4SE +/- 0.00, N = 41.271.291. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterAOM AV1 3.7Encoder Mode: Speed 0 Two-Pass - Input: Bosphorus 1080pAMD Preferred Core DisabledAMD Preferred Core Patched0.00360.00720.01080.01440.0180.0150.016
MinAvgMaxAMD Preferred Core Disabled12.483.3138.4AMD Preferred Core Patched12.283.0138.6OpenBenchmarking.orgWatts, Fewer Is BetterAOM AV1 3.7CPU Power Consumption Monitor4080120160200

PyPerformance

PyPerformance is the reference Python performance benchmark suite. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: django_templateAMD Preferred Core DisabledAMD Preferred Core Patched612182430SE +/- 0.03, N = 3SE +/- 0.03, N = 327.326.9
MinAvgMaxAMD Preferred Core Disabled11.626.128.6AMD Preferred Core Patched10.625.928.5OpenBenchmarking.orgWatts, Fewer Is BetterPyPerformance 1.0.0CPU Power Consumption Monitor918273645

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 2 - Buffer Length: 256 - Filter Length: 512AMD Preferred Core DisabledAMD Preferred Core Patched7M14M21M28M35MSE +/- 126366.31, N = 3SE +/- 320603.98, N = 633150333326673331. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
OpenBenchmarking.orgsamples/s Per Watt, More Is BetterLiquid-DSP 1.6Threads: 2 - Buffer Length: 256 - Filter Length: 512AMD Preferred Core DisabledAMD Preferred Core Patched200K400K600K800K1000K1145727.051121681.80
MinAvgMaxAMD Preferred Core Disabled11.928.931.7AMD Preferred Core Patched10.329.131.9OpenBenchmarking.orgWatts, Fewer Is BetterLiquid-DSP 1.6CPU Power Consumption Monitor918273645

PyPerformance

PyPerformance is the reference Python performance benchmark suite. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: goAMD Preferred Core DisabledAMD Preferred Core Patched306090120150SE +/- 0.33, N = 3SE +/- 0.33, N = 3137139
MinAvgMaxAMD Preferred Core Disabled10.126.028.5AMD Preferred Core Patched13.125.928.4OpenBenchmarking.orgWatts, Fewer Is BetterPyPerformance 1.0.0CPU Power Consumption Monitor918273645

Renaissance

Renaissance is a suite of benchmarks designed to test the Java JVM from Apache Spark to a Twitter-like service to Scala and other features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.14Test: Genetic Algorithm Using Jenetics + FuturesAMD Preferred Core DisabledAMD Preferred Core Patched30060090012001500SE +/- 12.32, N = 3SE +/- 10.41, N = 31229.71247.3MIN: 1139.55 / MAX: 1289.24MIN: 1186.95 / MAX: 1288.34
MinAvgMaxAMD Preferred Core Disabled12.938.643.4AMD Preferred Core Patched11.538.541.9OpenBenchmarking.orgWatts, Fewer Is BetterRenaissance 0.14CPU Power Consumption Monitor1224364860

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.14Test: Savina Reactors.IOAMD Preferred Core DisabledAMD Preferred Core Patched10002000300040005000SE +/- 5.48, N = 3SE +/- 40.42, N = 34788.44721.0MIN: 4777.52 / MAX: 6878.19MIN: 4661.99 / MAX: 6495.32
MinAvgMaxAMD Preferred Core Disabled10.634.956.8AMD Preferred Core Patched12.335.255.9OpenBenchmarking.orgWatts, Fewer Is BetterRenaissance 0.14CPU Power Consumption Monitor1632486480

Apache Hadoop

This is a benchmark of the Apache Hadoop making use of its built-in name-node throughput benchmark (NNThroughputBenchmark). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps per sec, More Is BetterApache Hadoop 3.3.6Operation: Create - Threads: 20 - Files: 100000AMD Preferred Core DisabledAMD Preferred Core Patched15K30K45K60K75KSE +/- 337.78, N = 4SE +/- 586.93, N = 46732368275
OpenBenchmarking.orgOps per sec Per Watt, More Is BetterApache Hadoop 3.3.6Operation: Create - Threads: 20 - Files: 100000AMD Preferred Core DisabledAMD Preferred Core Patched50010001500200025002523.242457.07
MinAvgMaxAMD Preferred Core Disabled10.926.752.9AMD Preferred Core Patched11.727.853.3OpenBenchmarking.orgWatts, Fewer Is BetterApache Hadoop 3.3.6CPU Power Consumption Monitor1530456075

OpenVINO

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.1Model: Person Vehicle Bike Detection FP16 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched1.13852.2773.41554.5545.6925SE +/- 0.02, N = 3SE +/- 0.02, N = 34.995.06MIN: 3.53 / MAX: 11.64MIN: 3.24 / MAX: 8.471. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
MinAvgMaxAMD Preferred Core Disabled12.1118.5130.9AMD Preferred Core Patched11.9117.4129.8OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.1CPU Power Consumption Monitor4080120160200

Apache CouchDB

This is a bulk insertion benchmark of Apache CouchDB. CouchDB is a document-oriented NoSQL database implemented in Erlang. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterApache CouchDB 3.3.2Bulk Size: 100 - Inserts: 3000 - Rounds: 30AMD Preferred Core DisabledAMD Preferred Core Patched60120180240300SE +/- 0.44, N = 3SE +/- 0.41, N = 3256.29259.851. (CXX) g++ options: -std=c++17 -lmozjs-91 -lm -lei -fPIC -MMD
MinAvgMaxAMD Preferred Core Disabled11.349.155.9AMD Preferred Core Patched12.950.256.3OpenBenchmarking.orgWatts, Fewer Is BetterApache CouchDB 3.3.2CPU Power Consumption Monitor1632486480

PyPerformance

PyPerformance is the reference Python performance benchmark suite. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: regex_compileAMD Preferred Core DisabledAMD Preferred Core Patched20406080100SE +/- 0.13, N = 3SE +/- 0.15, N = 388.287.0
MinAvgMaxAMD Preferred Core Disabled11.526.628.6AMD Preferred Core Patched12.626.628.5OpenBenchmarking.orgWatts, Fewer Is BetterPyPerformance 1.0.0CPU Power Consumption Monitor918273645

Apache CouchDB

This is a bulk insertion benchmark of Apache CouchDB. CouchDB is a document-oriented NoSQL database implemented in Erlang. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterApache CouchDB 3.3.2Bulk Size: 300 - Inserts: 3000 - Rounds: 30AMD Preferred Core DisabledAMD Preferred Core Patched90180270360450SE +/- 0.70, N = 3SE +/- 0.24, N = 3397.45402.761. (CXX) g++ options: -std=c++17 -lmozjs-91 -lm -lei -fPIC -MMD
OpenBenchmarking.orgWatts, Fewer Is BetterApache CouchDB 3.3.2CPU Power Consumption MonitorAMD Preferred Core DisabledAMD Preferred Core Patched1122334455Min: 12.15 / Avg: 50.19 / Max: 57.65Min: 11.84 / Avg: 51.35 / Max: 57.33

VVenC

VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterVVenC 1.9Video Input: Bosphorus 4K - Video Preset: FastAMD Preferred Core DisabledAMD Preferred Core Patched246810SE +/- 0.040, N = 3SE +/- 0.057, N = 37.4867.5851. (CXX) g++ options: -O3 -flto -fno-fat-lto-objects -flto=auto
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterVVenC 1.9Video Input: Bosphorus 4K - Video Preset: FastAMD Preferred Core DisabledAMD Preferred Core Patched0.02050.0410.06150.0820.10250.0910.091
MinAvgMaxAMD Preferred Core Disabled11.482.4119.6AMD Preferred Core Patched12.883.2121.3OpenBenchmarking.orgWatts, Fewer Is BetterVVenC 1.9CPU Power Consumption Monitor4080120160200

Apache Hadoop

This is a benchmark of the Apache Hadoop making use of its built-in name-node throughput benchmark (NNThroughputBenchmark). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps per sec, More Is BetterApache Hadoop 3.3.6Operation: Delete - Threads: 50 - Files: 100000AMD Preferred Core DisabledAMD Preferred Core Patched30K60K90K120K150KSE +/- 973.00, N = 15SE +/- 1345.47, N = 4116159117693
OpenBenchmarking.orgOps per sec Per Watt, More Is BetterApache Hadoop 3.3.6Operation: Delete - Threads: 50 - Files: 100000AMD Preferred Core DisabledAMD Preferred Core Patched90018002700360045003934.743997.35
MinAvgMaxAMD Preferred Core Disabled9.329.552.0AMD Preferred Core Patched12.029.453.8OpenBenchmarking.orgWatts, Fewer Is BetterApache Hadoop 3.3.6CPU Power Consumption Monitor1530456075

Timed FFmpeg Compilation

This test times how long it takes to build the FFmpeg multimedia library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed FFmpeg Compilation 6.0Time To CompileAMD Preferred Core DisabledAMD Preferred Core Patched510152025SE +/- 0.01, N = 3SE +/- 0.08, N = 321.8222.11
MinAvgMaxAMD Preferred Core Disabled11.479.1121.3AMD Preferred Core Patched13.078.7120.7OpenBenchmarking.orgWatts, Fewer Is BetterTimed FFmpeg Compilation 6.0CPU Power Consumption Monitor4080120160200

OpenVINO

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.1Model: Person Vehicle Bike Detection FP16 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched30060090012001500SE +/- 6.05, N = 3SE +/- 4.57, N = 31594.251573.811. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 1080p - Video Preset: Very FastAMD Preferred Core DisabledAMD Preferred Core Patched4080120160200SE +/- 0.18, N = 9SE +/- 0.09, N = 9172.84175.071. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterKvazaar 2.2Video Input: Bosphorus 1080p - Video Preset: Very FastAMD Preferred Core DisabledAMD Preferred Core Patched0.81251.6252.43753.254.06253.5953.611
MinAvgMaxAMD Preferred Core Disabled10.648.191.1AMD Preferred Core Patched11.548.592.3OpenBenchmarking.orgWatts, Fewer Is BetterKvazaar 2.2CPU Power Consumption Monitor20406080100

JPEG XL Decoding libjxl

The JPEG XL Image Coding System is designed to provide next-generation JPEG image capabilities with JPEG XL offering better image quality and compression over legacy JPEG. This test profile is suited for JPEG XL decode performance testing to PNG output file, the pts/jpexl test is for encode performance. The JPEG XL encoding/decoding is done using the libjxl codebase. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterJPEG XL Decoding libjxl 0.7CPU Threads: 1AMD Preferred Core DisabledAMD Preferred Core Patched1224364860SE +/- 0.18, N = 3SE +/- 0.38, N = 354.9155.61
OpenBenchmarking.orgMP/s Per Watt, More Is BetterJPEG XL Decoding libjxl 0.7CPU Threads: 1AMD Preferred Core DisabledAMD Preferred Core Patched0.46850.9371.40551.8742.34252.0612.082
MinAvgMaxAMD Preferred Core Disabled10.726.628.8AMD Preferred Core Patched11.626.728.3OpenBenchmarking.orgWatts, Fewer Is BetterJPEG XL Decoding libjxl 0.7CPU Power Consumption Monitor918273645

libavif avifenc

This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 1.0Encoder Speed: 6AMD Preferred Core DisabledAMD Preferred Core Patched0.82331.64662.46993.29324.1165SE +/- 0.007, N = 9SE +/- 0.007, N = 93.6593.6131. (CXX) g++ options: -O3 -fPIC -lm
MinAvgMaxAMD Preferred Core Disabled11.747.2120.0AMD Preferred Core Patched10.847.1119.7OpenBenchmarking.orgWatts, Fewer Is Betterlibavif avifenc 1.0CPU Power Consumption Monitor4080120160200

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 Preferred Core DisabledAMD Preferred Core Patched1000K2000K3000K4000K5000KSE +/- 2603.42, N = 3SE +/- 55518.77, N = 3450633344500001. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt
OpenBenchmarking.orgReal C/S Per Watt, More Is BetterJohn The Ripper 2023.03.14Test: MD5AMD Preferred Core DisabledAMD Preferred Core Patched8K16K24K32K40K37746.0537505.42
MinAvgMaxAMD Preferred Core Disabled12.5119.4130.4AMD Preferred Core Patched11.7118.6130.5OpenBenchmarking.orgWatts, Fewer Is BetterJohn The Ripper 2023.03.14CPU Power Consumption Monitor4080120160200

Apache CouchDB

This is a bulk insertion benchmark of Apache CouchDB. CouchDB is a document-oriented NoSQL database implemented in Erlang. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterApache CouchDB 3.3.2Bulk Size: 300 - Inserts: 1000 - Rounds: 30AMD Preferred Core DisabledAMD Preferred Core Patched306090120150SE +/- 0.22, N = 3SE +/- 0.20, N = 3117.71119.181. (CXX) g++ options: -std=c++17 -lmozjs-91 -lm -lei -fPIC -MMD
MinAvgMaxAMD Preferred Core Disabled11.451.158.2AMD Preferred Core Patched12.652.358.4OpenBenchmarking.orgWatts, Fewer Is BetterApache CouchDB 3.3.2CPU Power Consumption Monitor1632486480

JPEG XL libjxl

The JPEG XL Image Coding System is designed to provide next-generation JPEG image capabilities with JPEG XL offering better image quality and compression over legacy JPEG. This test profile is currently focused on the multi-threaded JPEG XL image encode performance using the reference libjxl library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterJPEG XL libjxl 0.7Input: JPEG - Quality: 100AMD Preferred Core DisabledAMD Preferred Core Patched0.18230.36460.54690.72920.9115SE +/- 0.01, N = 4SE +/- 0.01, N = 90.800.811. (CXX) g++ options: -fno-rtti -funwind-tables -O3 -O2 -fPIE -pie -lm -latomic
OpenBenchmarking.orgMP/s Per Watt, More Is BetterJPEG XL libjxl 0.7Input: JPEG - Quality: 100AMD Preferred Core DisabledAMD Preferred Core Patched0.00650.0130.01950.0260.03250.0280.029
OpenBenchmarking.orgWatts, Fewer Is BetterJPEG XL libjxl 0.7CPU Power Consumption MonitorAMD Preferred Core DisabledAMD Preferred Core Patched1020304050Min: 11.33 / Avg: 28.15 / Max: 47.62Min: 11.41 / Avg: 28.16 / Max: 47.55

GIMP

GIMP is an open-source image manipulaton program. This test profile will use the system-provided GIMP program otherwise on Windows relys upon a pre-packaged Windows binary from upstream GIMP.org. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterGIMP 2.10.30Test: rotateAMD Preferred Core DisabledAMD Preferred Core Patched3691215SE +/- 0.01, N = 4SE +/- 0.05, N = 412.9512.80
MinAvgMaxAMD Preferred Core Disabled12.924.931.8AMD Preferred Core Patched11.724.831.9OpenBenchmarking.orgWatts, Fewer Is BetterGIMP 2.10.30CPU Power Consumption Monitor918273645

SQLite

This is a simple benchmark of SQLite. At present this test profile just measures the time to perform a pre-defined number of insertions on an indexed database with a variable number of concurrent repetitions -- up to the maximum number of CPU threads available. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterSQLite 3.41.2Threads / Copies: 16AMD Preferred Core DisabledAMD Preferred Core Patched1.19842.39683.59524.79365.992SE +/- 0.024, N = 7SE +/- 0.032, N = 75.2645.3261. (CC) gcc options: -O2 -lz -lm
MinAvgMaxAMD Preferred Core Disabled12.331.346.7AMD Preferred Core Patched12.631.346.2OpenBenchmarking.orgWatts, Fewer Is BetterSQLite 3.41.2CPU Power Consumption Monitor1428425670

Gcrypt Library

Libgcrypt is a general purpose cryptographic library developed as part of the GnuPG project. This is a benchmark of libgcrypt's integrated benchmark and is measuring the time to run the benchmark command with a cipher/mac/hash repetition count set for 50 times as simple, high level look at the overall crypto performance of the system under test. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterGcrypt Library 1.9AMD Preferred Core DisabledAMD Preferred Core Patched4080120160200SE +/- 0.36, N = 3SE +/- 0.37, N = 3166.82164.881. (CC) gcc options: -O2 -fvisibility=hidden
MinAvgMaxAMD Preferred Core Disabled12.725.828.2AMD Preferred Core Patched10.925.828.1OpenBenchmarking.orgWatts, Fewer Is BetterGcrypt Library 1.9CPU Power Consumption Monitor918273645

Renaissance

Renaissance is a suite of benchmarks designed to test the Java JVM from Apache Spark to a Twitter-like service to Scala and other features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.14Test: Random ForestAMD Preferred Core DisabledAMD Preferred Core Patched100200300400500SE +/- 3.68, N = 3SE +/- 3.47, N = 3442.3437.2MIN: 404.28 / MAX: 533.26MIN: 398.35 / MAX: 523.08
MinAvgMaxAMD Preferred Core Disabled12.834.556.9AMD Preferred Core Patched10.134.755.1OpenBenchmarking.orgWatts, Fewer Is BetterRenaissance 0.14CPU Power Consumption Monitor1632486480

OpenRadioss

OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2023.09.15Model: Cell Phone Drop TestAMD Preferred Core DisabledAMD Preferred Core Patched1224364860SE +/- 0.36, N = 15SE +/- 0.45, N = 351.7052.30
MinAvgMaxAMD Preferred Core Disabled10.490.7104.2AMD Preferred Core Patched13.191.7104.6OpenBenchmarking.orgWatts, Fewer Is BetterOpenRadioss 2023.09.15CPU Power Consumption Monitor20406080100

JPEG XL libjxl

The JPEG XL Image Coding System is designed to provide next-generation JPEG image capabilities with JPEG XL offering better image quality and compression over legacy JPEG. This test profile is currently focused on the multi-threaded JPEG XL image encode performance using the reference libjxl library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterJPEG XL libjxl 0.7Input: PNG - Quality: 100AMD Preferred Core DisabledAMD Preferred Core Patched0.1980.3960.5940.7920.99SE +/- 0.01, N = 3SE +/- 0.01, N = 90.880.871. (CXX) g++ options: -fno-rtti -funwind-tables -O3 -O2 -fPIE -pie -lm -latomic
OpenBenchmarking.orgMP/s Per Watt, More Is BetterJPEG XL libjxl 0.7Input: PNG - Quality: 100AMD Preferred Core DisabledAMD Preferred Core Patched0.0070.0140.0210.0280.0350.0310.031
MinAvgMaxAMD Preferred Core Disabled9.628.446.8AMD Preferred Core Patched11.828.346.9OpenBenchmarking.orgWatts, Fewer Is BetterJPEG XL libjxl 0.7CPU Power Consumption Monitor1428425670

Apache Hadoop

This is a benchmark of the Apache Hadoop making use of its built-in name-node throughput benchmark (NNThroughputBenchmark). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps per sec, More Is BetterApache Hadoop 3.3.6Operation: Create - Threads: 50 - Files: 100000AMD Preferred Core DisabledAMD Preferred Core Patched15K30K45K60K75KSE +/- 540.98, N = 9SE +/- 793.68, N = 46769768474
OpenBenchmarking.orgOps per sec Per Watt, More Is BetterApache Hadoop 3.3.6Operation: Create - Threads: 50 - Files: 100000AMD Preferred Core DisabledAMD Preferred Core Patched50010001500200025002445.342476.40
MinAvgMaxAMD Preferred Core Disabled10.327.754.0AMD Preferred Core Patched11.327.754.1OpenBenchmarking.orgWatts, Fewer Is BetterApache Hadoop 3.3.6CPU Power Consumption Monitor1530456075

VVenC

VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterVVenC 1.9Video Input: Bosphorus 1080p - Video Preset: FasterAMD Preferred Core DisabledAMD Preferred Core Patched918273645SE +/- 0.12, N = 4SE +/- 0.10, N = 440.8041.241. (CXX) g++ options: -O3 -flto -fno-fat-lto-objects -flto=auto
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterVVenC 1.9Video Input: Bosphorus 1080p - Video Preset: FasterAMD Preferred Core DisabledAMD Preferred Core Patched0.15170.30340.45510.60680.75850.6720.674
MinAvgMaxAMD Preferred Core Disabled10.760.782.1AMD Preferred Core Patched10.761.282.1OpenBenchmarking.orgWatts, Fewer Is BetterVVenC 1.9CPU Power Consumption Monitor20406080100

AOM AV1

This is a test of the AOMedia AV1 encoder (libaom) developed by AOMedia and Google as the AV1 Codec Library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.7Encoder Mode: Speed 4 Two-Pass - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched3691215SE +/- 0.06, N = 3SE +/- 0.03, N = 311.1811.301. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterAOM AV1 3.7Encoder Mode: Speed 4 Two-Pass - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched0.03350.0670.10050.1340.16750.1470.149
MinAvgMaxAMD Preferred Core Disabled11.976.0102.3AMD Preferred Core Patched11.476.0101.1OpenBenchmarking.orgWatts, Fewer Is BetterAOM AV1 3.7CPU Power Consumption Monitor20406080100

PyPerformance

PyPerformance is the reference Python performance benchmark suite. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: 2to3AMD Preferred Core DisabledAMD Preferred Core Patched4080120160200SE +/- 0.00, N = 3SE +/- 0.33, N = 3191189
MinAvgMaxAMD Preferred Core Disabled12.226.528.7AMD Preferred Core Patched10.926.628.6OpenBenchmarking.orgWatts, Fewer Is BetterPyPerformance 1.0.0CPU Power Consumption Monitor918273645

RocksDB

This is a benchmark of Meta/Facebook's RocksDB as an embeddable persistent key-value store for fast storage based on Google's LevelDB. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 8.0Test: Random FillAMD Preferred Core DisabledAMD Preferred Core Patched300K600K900K1200K1500KSE +/- 1583.20, N = 3SE +/- 595.61, N = 3128357212706771. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread
OpenBenchmarking.orgOp/s Per Watt, More Is BetterRocksDB 8.0Test: Random FillAMD Preferred Core DisabledAMD Preferred Core Patched3K6K9K12K15K15846.1015626.31
MinAvgMaxAMD Preferred Core Disabled11.381.088.1AMD Preferred Core Patched13.081.388.1OpenBenchmarking.orgWatts, Fewer Is BetterRocksDB 8.0CPU Power Consumption Monitor20406080100

TiDB Community Server

This is a PingCAP TiDB Community Server benchmark facilitated using the sysbench OLTP database benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Second, More Is BetterTiDB Community Server 7.3Test: oltp_update_non_index - Threads: 1AMD Preferred Core DisabledAMD Preferred Core Patched400800120016002000SE +/- 2.40, N = 3SE +/- 2.38, N = 319921972
OpenBenchmarking.orgQueries Per Second Per Watt, More Is BetterTiDB Community Server 7.3Test: oltp_update_non_index - Threads: 1AMD Preferred Core DisabledAMD Preferred Core Patched112233445549.8749.37
OpenBenchmarking.orgWatts, Fewer Is BetterTiDB Community Server 7.3CPU Power Consumption MonitorAMD Preferred Core DisabledAMD Preferred Core Patched1020304050Min: 11.66 / Avg: 39.94 / Max: 48.8Min: 10.87 / Avg: 39.94 / Max: 49.31

Apache IoTDB

Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.2Device Count: 100 - Batch Size Per Write: 100 - Sensor Count: 200 - Client Number: 100AMD Preferred Core DisabledAMD Preferred Core Patched3M6M9M12M15MSE +/- 177739.87, N = 3SE +/- 39383.72, N = 31314273813272391

libavif avifenc

This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 1.0Encoder Speed: 0AMD Preferred Core DisabledAMD Preferred Core Patched20406080100SE +/- 0.18, N = 3SE +/- 0.11, N = 376.5875.831. (CXX) g++ options: -O3 -fPIC -lm
MinAvgMaxAMD Preferred Core Disabled10.979.6127.4AMD Preferred Core Patched11.880.1127.8OpenBenchmarking.orgWatts, Fewer Is Betterlibavif avifenc 1.0CPU Power Consumption Monitor4080120160200

OpenBenchmarking.orgSeconds, Fewer Is Betterlibavif avifenc 1.0Encoder Speed: 2AMD Preferred Core DisabledAMD Preferred Core Patched918273645SE +/- 0.10, N = 3SE +/- 0.25, N = 338.6838.311. (CXX) g++ options: -O3 -fPIC -lm
MinAvgMaxAMD Preferred Core Disabled12.366.8124.6AMD Preferred Core Patched11.667.8124.9OpenBenchmarking.orgWatts, Fewer Is Betterlibavif avifenc 1.0CPU Power Consumption Monitor4080120160200

SQLite Speedtest

This is a benchmark of SQLite's speedtest1 benchmark program with an increased problem size of 1,000. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterSQLite Speedtest 3.30Timed Time - Size 1,000AMD Preferred Core DisabledAMD Preferred Core Patched1122334455SE +/- 0.42, N = 3SE +/- 0.58, N = 448.1847.731. (CC) gcc options: -O2 -lz
MinAvgMaxAMD Preferred Core Disabled12.626.428.4AMD Preferred Core Patched11.726.428.3OpenBenchmarking.orgWatts, Fewer Is BetterSQLite Speedtest 3.30CPU Power Consumption Monitor918273645

SVT-AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 1.7Encoder Mode: Preset 4 - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched1.16572.33143.49714.66285.8285SE +/- 0.027, N = 3SE +/- 0.034, N = 35.1815.1331. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterSVT-AV1 1.7Encoder Mode: Preset 4 - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched0.01620.03240.04860.06480.0810.0720.072
MinAvgMaxAMD Preferred Core Disabled11.771.6115.4AMD Preferred Core Patched11.171.2115.1OpenBenchmarking.orgWatts, Fewer Is BetterSVT-AV1 1.7CPU Power Consumption Monitor4080120160200

Renaissance

Renaissance is a suite of benchmarks designed to test the Java JVM from Apache Spark to a Twitter-like service to Scala and other features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.14Test: Apache Spark PageRankAMD Preferred Core DisabledAMD Preferred Core Patched5001000150020002500SE +/- 17.65, N = 15SE +/- 23.82, N = 32179.92159.8MIN: 1901.17 / MAX: 2411.65MIN: 1892.22 / MAX: 2312.36
MinAvgMaxAMD Preferred Core Disabled9.039.254.9AMD Preferred Core Patched10.939.252.6OpenBenchmarking.orgWatts, Fewer Is BetterRenaissance 0.14CPU Power Consumption Monitor1530456075

Redis

Redis is an open-source in-memory data structure store, used as a database, cache, and message broker. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRequests Per Second, More Is BetterRedis 7.0.4Test: SET - Parallel Connections: 50AMD Preferred Core DisabledAMD Preferred Core Patched800K1600K2400K3200K4000KSE +/- 22710.85, N = 3SE +/- 32722.93, N = 73667648.003633880.251. (CXX) g++ options: -MM -MT -g3 -fvisibility=hidden -O3
OpenBenchmarking.orgRequests Per Second Per Watt, More Is BetterRedis 7.0.4Test: SET - Parallel Connections: 50AMD Preferred Core DisabledAMD Preferred Core Patched20K40K60K80K100K108853.82107496.48
MinAvgMaxAMD Preferred Core Disabled11.533.753.0AMD Preferred Core Patched12.233.853.9OpenBenchmarking.orgWatts, Fewer Is BetterRedis 7.0.4CPU Power Consumption Monitor1530456075

GIMP

GIMP is an open-source image manipulaton program. This test profile will use the system-provided GIMP program otherwise on Windows relys upon a pre-packaged Windows binary from upstream GIMP.org. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterGIMP 2.10.30Test: unsharp-maskAMD Preferred Core DisabledAMD Preferred Core Patched48121620SE +/- 0.05, N = 3SE +/- 0.05, N = 317.3217.16
MinAvgMaxAMD Preferred Core Disabled11.130.439.5AMD Preferred Core Patched12.530.537.9OpenBenchmarking.orgWatts, Fewer Is BetterGIMP 2.10.30CPU Power Consumption Monitor1122334455

PyPerformance

PyPerformance is the reference Python performance benchmark suite. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: floatAMD Preferred Core DisabledAMD Preferred Core Patched1530456075SE +/- 0.07, N = 3SE +/- 0.03, N = 365.765.1
MinAvgMaxAMD Preferred Core Disabled10.826.128.7AMD Preferred Core Patched12.926.128.7OpenBenchmarking.orgWatts, Fewer Is BetterPyPerformance 1.0.0CPU Power Consumption Monitor918273645

Neural Magic DeepSparse

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched246810SE +/- 0.0135, N = 3SE +/- 0.0242, N = 36.10306.0486
MinAvgMaxAMD Preferred Core Disabled10.165.077.4AMD Preferred Core Patched12.063.077.9OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor20406080100

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched4080120160200SE +/- 0.37, N = 3SE +/- 0.66, N = 3163.74165.21

JPEG XL libjxl

The JPEG XL Image Coding System is designed to provide next-generation JPEG image capabilities with JPEG XL offering better image quality and compression over legacy JPEG. This test profile is currently focused on the multi-threaded JPEG XL image encode performance using the reference libjxl library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMP/s, More Is BetterJPEG XL libjxl 0.7Input: PNG - Quality: 90AMD Preferred Core DisabledAMD Preferred Core Patched3691215SE +/- 0.03, N = 3SE +/- 0.06, N = 310.0210.111. (CXX) g++ options: -fno-rtti -funwind-tables -O3 -O2 -fPIE -pie -lm -latomic
OpenBenchmarking.orgMP/s Per Watt, More Is BetterJPEG XL libjxl 0.7Input: PNG - Quality: 90AMD Preferred Core DisabledAMD Preferred Core Patched0.06750.1350.20250.270.33750.2980.300
MinAvgMaxAMD Preferred Core Disabled12.733.745.5AMD Preferred Core Patched11.133.745.4OpenBenchmarking.orgWatts, Fewer Is BetterJPEG XL libjxl 0.7CPU Power Consumption Monitor1224364860

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: HMAC-SHA512AMD Preferred Core DisabledAMD Preferred Core Patched40M80M120M160M200MSE +/- 1038142.09, N = 3SE +/- 154855.85, N = 31844690001860940001. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt
OpenBenchmarking.orgReal C/S Per Watt, More Is BetterJohn The Ripper 2023.03.14Test: HMAC-SHA512AMD Preferred Core DisabledAMD Preferred Core Patched600K1200K1800K2400K3000K2627749.892642751.76
MinAvgMaxAMD Preferred Core Disabled11.170.292.1AMD Preferred Core Patched12.670.492.8OpenBenchmarking.orgWatts, Fewer Is BetterJohn The Ripper 2023.03.14CPU Power Consumption Monitor20406080100

OpenVINO

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.1Model: Vehicle Detection FP16 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched2004006008001000SE +/- 2.69, N = 3SE +/- 5.43, N = 31013.341022.221. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

TiDB Community Server

This is a PingCAP TiDB Community Server benchmark facilitated using the sysbench OLTP database benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Second, More Is BetterTiDB Community Server 7.3Test: oltp_point_select - Threads: 1AMD Preferred Core DisabledAMD Preferred Core Patched15003000450060007500SE +/- 24.08, N = 3SE +/- 25.88, N = 369976940
OpenBenchmarking.orgQueries Per Second Per Watt, More Is BetterTiDB Community Server 7.3Test: oltp_point_select - Threads: 1AMD Preferred Core DisabledAMD Preferred Core Patched4080120160200177.92175.95
OpenBenchmarking.orgWatts, Fewer Is BetterTiDB Community Server 7.3CPU Power Consumption MonitorAMD Preferred Core DisabledAMD Preferred Core Patched1020304050Min: 12.09 / Avg: 39.33 / Max: 47.89Min: 11.27 / Avg: 39.44 / Max: 48.69

VVenC

VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterVVenC 1.9Video Input: Bosphorus 4K - Video Preset: FasterAMD Preferred Core DisabledAMD Preferred Core Patched48121620SE +/- 0.02, N = 3SE +/- 0.06, N = 314.4714.591. (CXX) g++ options: -O3 -flto -fno-fat-lto-objects -flto=auto
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterVVenC 1.9Video Input: Bosphorus 4K - Video Preset: FasterAMD Preferred Core DisabledAMD Preferred Core Patched0.04390.08780.13170.17560.21950.1940.195
MinAvgMaxAMD Preferred Core Disabled11.174.6102.7AMD Preferred Core Patched11.574.6101.2OpenBenchmarking.orgWatts, Fewer Is BetterVVenC 1.9CPU Power Consumption Monitor20406080100

Apache HTTP Server

This is a test of the Apache HTTPD web server. This Apache HTTPD web server benchmark test profile makes use of the wrk program for facilitating the HTTP requests over a fixed period time with a configurable number of concurrent clients. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRequests Per Second, More Is BetterApache HTTP Server 2.4.56Concurrent Requests: 200AMD Preferred Core DisabledAMD Preferred Core Patched90K180K270K360K450KSE +/- 1196.94, N = 3SE +/- 528.46, N = 3401516.08398392.991. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2
OpenBenchmarking.orgRequests Per Second Per Watt, More Is BetterApache HTTP Server 2.4.56Concurrent Requests: 200AMD Preferred Core DisabledAMD Preferred Core Patched100020003000400050004751.124728.46
MinAvgMaxAMD Preferred Core Disabled11.584.590.7AMD Preferred Core Patched5.984.390.6OpenBenchmarking.orgWatts, Fewer Is BetterApache HTTP Server 2.4.56CPU Power Consumption Monitor20406080100

OpenVINO

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.1Model: Vehicle Detection FP16 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched246810SE +/- 0.02, N = 3SE +/- 0.04, N = 37.877.81MIN: 4.31 / MAX: 20.33MIN: 3.9 / MAX: 14.131. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
MinAvgMaxAMD Preferred Core Disabled11.0118.6127.7AMD Preferred Core Patched12.3119.7133.0OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.1CPU Power Consumption Monitor4080120160200

OpenRadioss

OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2023.09.15Model: Rubber O-Ring Seal InstallationAMD Preferred Core DisabledAMD Preferred Core Patched1530456075SE +/- 0.10, N = 3SE +/- 0.38, N = 369.1068.58
MinAvgMaxAMD Preferred Core Disabled12.6100.7110.0AMD Preferred Core Patched10.7100.3109.5OpenBenchmarking.orgWatts, Fewer Is BetterOpenRadioss 2023.09.15CPU Power Consumption Monitor20406080100

Neural Magic DeepSparse

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched246810SE +/- 0.0073, N = 3SE +/- 0.0173, N = 38.07738.0169
MinAvgMaxAMD Preferred Core Disabled11.385.8118.8AMD Preferred Core Patched11.884.4119.9OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor4080120160200

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: MediumAMD Preferred Core DisabledAMD Preferred Core Patched510152025SE +/- 0.02, N = 3SE +/- 0.03, N = 321.3621.521. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: MediumAMD Preferred Core DisabledAMD Preferred Core Patched0.04680.09360.14040.18720.2340.2070.208
MinAvgMaxAMD Preferred Core Disabled12.1103.1122.1AMD Preferred Core Patched12.4103.7122.5OpenBenchmarking.orgWatts, Fewer Is BetterKvazaar 2.2CPU Power Consumption Monitor4080120160200

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched2004006008001000SE +/- 0.91, N = 3SE +/- 2.15, N = 3988.97996.34

OpenVINO

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.1Model: Face Detection Retail FP16 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched7001400210028003500SE +/- 6.78, N = 3SE +/- 9.55, N = 33040.703018.311. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -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 2.12Device: CPU - Batch Size: 16 - Model: ResNet-50AMD Preferred Core DisabledAMD Preferred Core Patched918273645SE +/- 0.02, N = 3SE +/- 0.07, N = 339.7239.43
OpenBenchmarking.orgimages/sec Per Watt, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 16 - Model: ResNet-50AMD Preferred Core DisabledAMD Preferred Core Patched0.1190.2380.3570.4760.5950.5290.526
MinAvgMaxAMD Preferred Core Disabled7.575.086.7AMD Preferred Core Patched11.174.987.0OpenBenchmarking.orgWatts, Fewer Is BetterTensorFlow 2.12CPU Power Consumption Monitor20406080100

Apache HTTP Server

This is a test of the Apache HTTPD web server. This Apache HTTPD web server benchmark test profile makes use of the wrk program for facilitating the HTTP requests over a fixed period time with a configurable number of concurrent clients. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRequests Per Second, More Is BetterApache HTTP Server 2.4.56Concurrent Requests: 500AMD Preferred Core DisabledAMD Preferred Core Patched70K140K210K280K350KSE +/- 522.74, N = 3SE +/- 369.79, N = 3333352.80330962.461. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2
OpenBenchmarking.orgRequests Per Second Per Watt, More Is BetterApache HTTP Server 2.4.56Concurrent Requests: 500AMD Preferred Core DisabledAMD Preferred Core Patched80016002400320040003756.753766.56
MinAvgMaxAMD Preferred Core Disabled10.988.795.7AMD Preferred Core Patched11.587.994.6OpenBenchmarking.orgWatts, Fewer Is BetterApache HTTP Server 2.4.56CPU Power Consumption Monitor20406080100

RocksDB

This is a benchmark of Meta/Facebook's RocksDB as an embeddable persistent key-value store for fast storage based on Google's LevelDB. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 8.0Test: Read Random Write RandomAMD Preferred Core DisabledAMD Preferred Core Patched700K1400K2100K2800K3500KSE +/- 4601.20, N = 3SE +/- 8575.91, N = 3313507631127911. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread
OpenBenchmarking.orgOp/s Per Watt, More Is BetterRocksDB 8.0Test: Read Random Write RandomAMD Preferred Core DisabledAMD Preferred Core Patched7K14K21K28K35K33106.6633049.32
MinAvgMaxAMD Preferred Core Disabled11.894.7102.5AMD Preferred Core Patched11.194.2101.9OpenBenchmarking.orgWatts, Fewer Is BetterRocksDB 8.0CPU Power Consumption Monitor20406080100

Neural Magic DeepSparse

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched510152025SE +/- 0.01, N = 3SE +/- 0.14, N = 318.3918.52
MinAvgMaxAMD Preferred Core Disabled12.172.396.0AMD Preferred Core Patched10.373.099.0OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor20406080100

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched1224364860SE +/- 0.04, N = 3SE +/- 0.40, N = 354.3754.00

nginx

This is a benchmark of the lightweight Nginx HTTP(S) web-server. This Nginx web server benchmark test profile makes use of the wrk program for facilitating the HTTP requests over a fixed period time with a configurable number of concurrent clients/connections. HTTPS with a self-signed OpenSSL certificate is used by this test for local benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.23.2Connections: 200AMD Preferred Core DisabledAMD Preferred Core Patched30K60K90K120K150KSE +/- 51.48, N = 3SE +/- 163.99, N = 3143802.11144792.601. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2
OpenBenchmarking.orgRequests Per Second Per Watt, More Is Betternginx 1.23.2Connections: 200AMD Preferred Core DisabledAMD Preferred Core Patched300600900120015001342.211348.08
MinAvgMaxAMD Preferred Core Disabled11.9107.1115.7AMD Preferred Core Patched12.9107.4116.4OpenBenchmarking.orgWatts, Fewer Is Betternginx 1.23.2CPU Power Consumption Monitor4080120160200

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: SlowAMD Preferred Core DisabledAMD Preferred Core Patched510152025SE +/- 0.03, N = 3SE +/- 0.02, N = 320.8520.991. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: SlowAMD Preferred Core DisabledAMD Preferred Core Patched0.04570.09140.13710.18280.22850.2030.202
MinAvgMaxAMD Preferred Core Disabled12.0102.8122.5AMD Preferred Core Patched12.8103.9123.1OpenBenchmarking.orgWatts, Fewer Is BetterKvazaar 2.2CPU Power Consumption Monitor4080120160200

OpenVINO

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.1Model: Age Gender Recognition Retail 0013 FP16 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched7K14K21K28K35KSE +/- 20.93, N = 3SE +/- 36.74, N = 333780.3733562.191. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched60120180240300SE +/- 0.89, N = 3SE +/- 0.40, N = 3277.83276.05

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched0.81461.62922.44383.25844.073SE +/- 0.0115, N = 3SE +/- 0.0052, N = 33.59743.6205
MinAvgMaxAMD Preferred Core Disabled12.854.970.8AMD Preferred Core Patched12.456.470.6OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor20406080100

AOM AV1

This is a test of the AOMedia AV1 encoder (libaom) developed by AOMedia and Google as the AV1 Codec Library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.7Encoder Mode: Speed 4 Two-Pass - Input: Bosphorus 1080pAMD Preferred Core DisabledAMD Preferred Core Patched612182430SE +/- 0.14, N = 3SE +/- 0.02, N = 324.1824.331. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterAOM AV1 3.7Encoder Mode: Speed 4 Two-Pass - Input: Bosphorus 1080pAMD Preferred Core DisabledAMD Preferred Core Patched0.09320.18640.27960.37280.4660.4140.413
MinAvgMaxAMD Preferred Core Disabled11.258.471.1AMD Preferred Core Patched12.058.971.6OpenBenchmarking.orgWatts, Fewer Is BetterAOM AV1 3.7CPU Power Consumption Monitor20406080100

Apache CouchDB

This is a bulk insertion benchmark of Apache CouchDB. CouchDB is a document-oriented NoSQL database implemented in Erlang. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterApache CouchDB 3.3.2Bulk Size: 100 - Inserts: 1000 - Rounds: 30AMD Preferred Core DisabledAMD Preferred Core Patched20406080100SE +/- 0.32, N = 3SE +/- 0.23, N = 377.1477.601. (CXX) g++ options: -std=c++17 -lmozjs-91 -lm -lei -fPIC -MMD
MinAvgMaxAMD Preferred Core Disabled12.347.956.6AMD Preferred Core Patched12.549.957.7OpenBenchmarking.orgWatts, Fewer Is BetterApache CouchDB 3.3.2CPU Power Consumption Monitor1632486480

SQLite

This is a simple benchmark of SQLite. At present this test profile just measures the time to perform a pre-defined number of insertions on an indexed database with a variable number of concurrent repetitions -- up to the maximum number of CPU threads available. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterSQLite 3.41.2Threads / Copies: 4AMD Preferred Core DisabledAMD Preferred Core Patched0.57351.1471.72052.2942.8675SE +/- 0.017, N = 10SE +/- 0.009, N = 102.5492.5341. (CC) gcc options: -O2 -lz -lm
MinAvgMaxAMD Preferred Core Disabled12.121.034.1AMD Preferred Core Patched10.021.034.7OpenBenchmarking.orgWatts, Fewer Is BetterSQLite 3.41.2CPU Power Consumption Monitor1020304050

GIMP

GIMP is an open-source image manipulaton program. This test profile will use the system-provided GIMP program otherwise on Windows relys upon a pre-packaged Windows binary from upstream GIMP.org. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterGIMP 2.10.30Test: auto-levelsAMD Preferred Core DisabledAMD Preferred Core Patched48121620SE +/- 0.05, N = 4SE +/- 0.05, N = 414.1614.08
MinAvgMaxAMD Preferred Core Disabled12.528.135.0AMD Preferred Core Patched11.228.135.6OpenBenchmarking.orgWatts, Fewer Is BetterGIMP 2.10.30CPU Power Consumption Monitor1020304050

OpenVINO

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.1Model: Face Detection Retail FP16-INT8 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched0.78531.57062.35593.14123.9265SE +/- 0.00, N = 3SE +/- 0.01, N = 33.473.49MIN: 1.94 / MAX: 11MIN: 1.94 / MAX: 8.761. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
MinAvgMaxAMD Preferred Core Disabled12.8115.3125.1AMD Preferred Core Patched12.9113.4123.9OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.1CPU Power Consumption Monitor4080120160200

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.1Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched10K20K30K40K50KSE +/- 64.58, N = 3SE +/- 79.98, N = 348090.1647818.161. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Neural Magic DeepSparse

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: ResNet-50, Sparse INT8 - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched0.1860.3720.5580.7440.93SE +/- 0.0005, N = 3SE +/- 0.0019, N = 30.82230.8268
MinAvgMaxAMD Preferred Core Disabled11.365.175.2AMD Preferred Core Patched13.065.776.0OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor20406080100

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: ResNet-50, Sparse INT8 - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched30060090012001500SE +/- 0.74, N = 3SE +/- 2.81, N = 31212.801206.25

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 6.0Encoder: libx265 - Scenario: LiveAMD Preferred Core DisabledAMD Preferred Core Patched4080120160200SE +/- 1.26, N = 3SE +/- 0.60, N = 3176.16175.221. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
OpenBenchmarking.orgFPS Per Watt, More Is BetterFFmpeg 6.0Encoder: libx265 - Scenario: LiveAMD Preferred Core DisabledAMD Preferred Core Patched0.95831.91662.87493.83324.79154.2594.207
MinAvgMaxAMD Preferred Core Disabled11.641.467.6AMD Preferred Core Patched12.341.768.0OpenBenchmarking.orgWatts, Fewer Is BetterFFmpeg 6.0CPU Power Consumption Monitor20406080100

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: Very FastAMD Preferred Core DisabledAMD Preferred Core Patched4080120160200SE +/- 0.12, N = 8SE +/- 0.15, N = 8159.75160.60
OpenBenchmarking.orgFrames Per Second Per Watt, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: Very FastAMD Preferred Core DisabledAMD Preferred Core Patched0.67411.34822.02232.69643.37052.9962.972
MinAvgMaxAMD Preferred Core Disabled11.753.3100.7AMD Preferred Core Patched12.054.0101.2OpenBenchmarking.orgWatts, Fewer Is Betteruvg266 0.4.1CPU Power Consumption Monitor20406080100

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: Super FastAMD Preferred Core DisabledAMD Preferred Core Patched4080120160200SE +/- 0.13, N = 9SE +/- 0.09, N = 9173.51174.43
OpenBenchmarking.orgFrames Per Second Per Watt, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: Super FastAMD Preferred Core DisabledAMD Preferred Core Patched0.7671.5342.3013.0683.8353.4093.240
MinAvgMaxAMD Preferred Core Disabled11.050.9103.3AMD Preferred Core Patched11.453.8103.7OpenBenchmarking.orgWatts, Fewer Is Betteruvg266 0.4.1CPU Power Consumption Monitor20406080100

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.orgSeconds, Fewer Is BetterFFmpeg 6.0Encoder: libx265 - Scenario: LiveAMD Preferred Core DisabledAMD Preferred Core Patched714212835SE +/- 0.20, N = 3SE +/- 0.10, N = 328.6728.821. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

nginx

This is a benchmark of the lightweight Nginx HTTP(S) web-server. This Nginx web server benchmark test profile makes use of the wrk program for facilitating the HTTP requests over a fixed period time with a configurable number of concurrent clients/connections. HTTPS with a self-signed OpenSSL certificate is used by this test for local benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.23.2Connections: 100AMD Preferred Core DisabledAMD Preferred Core Patched30K60K90K120K150KSE +/- 148.74, N = 3SE +/- 493.50, N = 3144206.53143486.291. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2
OpenBenchmarking.orgRequests Per Second Per Watt, More Is Betternginx 1.23.2Connections: 100AMD Preferred Core DisabledAMD Preferred Core Patched300600900120015001337.501327.71
MinAvgMaxAMD Preferred Core Disabled11.9107.8116.7AMD Preferred Core Patched11.2108.1116.9OpenBenchmarking.orgWatts, Fewer Is Betternginx 1.23.2CPU Power Consumption Monitor4080120160200

Darktable

Darktable is an open-source photography / workflow application this will use any system-installed Darktable program or on Windows will automatically download the pre-built binary from the project. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 3.8.1Test: Masskrug - Acceleration: CPU-onlyAMD Preferred Core DisabledAMD Preferred Core Patched0.70131.40262.10392.80523.5065SE +/- 0.004, N = 8SE +/- 0.013, N = 83.1023.117
MinAvgMaxAMD Preferred Core Disabled10.231.962.3AMD Preferred Core Patched11.232.061.8OpenBenchmarking.orgWatts, Fewer Is BetterDarktable 3.8.1CPU Power Consumption Monitor20406080100

OpenVINO

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.1Model: Person Detection FP32 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched20406080100SE +/- 0.17, N = 3SE +/- 0.38, N = 395.2494.821. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

RocksDB

This is a benchmark of Meta/Facebook's RocksDB as an embeddable persistent key-value store for fast storage based on Google's LevelDB. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 8.0Test: Read While WritingAMD Preferred Core DisabledAMD Preferred Core Patched900K1800K2700K3600K4500KSE +/- 27775.49, N = 15SE +/- 58270.28, N = 3410594341235861. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread
OpenBenchmarking.orgOp/s Per Watt, More Is BetterRocksDB 8.0Test: Read While WritingAMD Preferred Core DisabledAMD Preferred Core Patched9K18K27K36K45K42448.4643188.38
MinAvgMaxAMD Preferred Core Disabled11.296.7121.9AMD Preferred Core Patched11.295.5119.7OpenBenchmarking.orgWatts, Fewer Is BetterRocksDB 8.0CPU Power Consumption Monitor4080120160200

OpenVINO

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.1Model: Person Detection FP32 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched20406080100SE +/- 0.16, N = 3SE +/- 0.34, N = 383.9384.28MIN: 42.48 / MAX: 113.62MIN: 43.67 / MAX: 110.891. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
MinAvgMaxAMD Preferred Core Disabled12.193.2106.2AMD Preferred Core Patched12.694.0106.1OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.1CPU Power Consumption Monitor20406080100

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 1080p - Video Preset: Super FastAMD Preferred Core DisabledAMD Preferred Core Patched50100150200250SE +/- 0.14, N = 10SE +/- 0.19, N = 10230.35231.301. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterKvazaar 2.2Video Input: Bosphorus 1080p - Video Preset: Super FastAMD Preferred Core DisabledAMD Preferred Core Patched1.14442.28883.43324.57765.7225.0745.086
MinAvgMaxAMD Preferred Core Disabled10.245.497.8AMD Preferred Core Patched11.045.597.8OpenBenchmarking.orgWatts, Fewer Is BetterKvazaar 2.2CPU Power Consumption Monitor20406080100

Apache HTTP Server

This is a test of the Apache HTTPD web server. This Apache HTTPD web server benchmark test profile makes use of the wrk program for facilitating the HTTP requests over a fixed period time with a configurable number of concurrent clients. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRequests Per Second, More Is BetterApache HTTP Server 2.4.56Concurrent Requests: 1000AMD Preferred Core DisabledAMD Preferred Core Patched70K140K210K280K350KSE +/- 217.20, N = 3SE +/- 307.25, N = 3329562.58328237.071. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2
OpenBenchmarking.orgRequests Per Second Per Watt, More Is BetterApache HTTP Server 2.4.56Concurrent Requests: 1000AMD Preferred Core DisabledAMD Preferred Core Patched80016002400320040003719.373741.95
MinAvgMaxAMD Preferred Core Disabled9.788.695.8AMD Preferred Core Patched12.787.794.7OpenBenchmarking.orgWatts, Fewer Is BetterApache HTTP Server 2.4.56CPU Power Consumption Monitor20406080100

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: Ultra FastAMD Preferred Core DisabledAMD Preferred Core Patched20406080100SE +/- 0.08, N = 6SE +/- 0.04, N = 677.2977.591. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: Ultra FastAMD Preferred Core DisabledAMD Preferred Core Patched0.24140.48280.72420.96561.2071.0731.065
MinAvgMaxAMD Preferred Core Disabled11.572.0106.8AMD Preferred Core Patched12.872.8107.9OpenBenchmarking.orgWatts, Fewer Is BetterKvazaar 2.2CPU Power Consumption Monitor20406080100

OpenVINO

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.1Model: Face Detection FP16-INT8 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched70140210280350SE +/- 0.71, N = 3SE +/- 0.04, N = 3312.26313.46MIN: 290.47 / MAX: 323.47MIN: 296.63 / MAX: 324.451. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
MinAvgMaxAMD Preferred Core Disabled12.8111.3122.7AMD Preferred Core Patched12.0111.4122.9OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.1CPU Power Consumption Monitor4080120160200

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: Very FastAMD Preferred Core DisabledAMD Preferred Core Patched1020304050SE +/- 0.03, N = 4SE +/- 0.04, N = 445.4745.641. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: Very FastAMD Preferred Core DisabledAMD Preferred Core Patched0.12690.25380.38070.50760.63450.5640.560
MinAvgMaxAMD Preferred Core Disabled11.780.6106.0AMD Preferred Core Patched11.681.5107.0OpenBenchmarking.orgWatts, Fewer Is BetterKvazaar 2.2CPU Power Consumption Monitor20406080100

Neural Magic DeepSparse

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: CV Detection, YOLOv5s COCO, Sparse INT8 - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched1530456075SE +/- 0.34, N = 3SE +/- 0.24, N = 365.5365.28
MinAvgMaxAMD Preferred Core Disabled12.690.1109.6AMD Preferred Core Patched11.893.3110.1OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor20406080100

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 2.12Device: CPU - Batch Size: 32 - Model: ResNet-50AMD Preferred Core DisabledAMD Preferred Core Patched918273645SE +/- 0.02, N = 3SE +/- 0.01, N = 340.5440.39
OpenBenchmarking.orgimages/sec Per Watt, More Is BetterTensorFlow 2.12Device: CPU - Batch Size: 32 - Model: ResNet-50AMD Preferred Core DisabledAMD Preferred Core Patched0.11520.23040.34560.46080.5760.5120.510
MinAvgMaxAMD Preferred Core Disabled8.879.288.1AMD Preferred Core Patched12.079.287.9OpenBenchmarking.orgWatts, Fewer Is BetterTensorFlow 2.12CPU Power Consumption Monitor20406080100

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: CV Detection, YOLOv5s COCO, Sparse INT8 - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched306090120150SE +/- 0.63, N = 3SE +/- 0.43, N = 3122.01122.45

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: SlowAMD Preferred Core DisabledAMD Preferred Core Patched48121620SE +/- 0.01, N = 3SE +/- 0.01, N = 313.8413.89
OpenBenchmarking.orgFrames Per Second Per Watt, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: SlowAMD Preferred Core DisabledAMD Preferred Core Patched0.02770.05540.08310.11080.13850.1220.123
MinAvgMaxAMD Preferred Core Disabled10.4113.4127.5AMD Preferred Core Patched10.8113.2127.9OpenBenchmarking.orgWatts, Fewer Is Betteruvg266 0.4.1CPU Power Consumption Monitor4080120160200

LAME MP3 Encoding

LAME is an MP3 encoder licensed under the LGPL. This test measures the time required to encode a WAV file to MP3 format. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterLAME MP3 Encoding 3.100WAV To MP3AMD Preferred Core DisabledAMD Preferred Core Patched1.20852.4173.62554.8346.0425SE +/- 0.015, N = 7SE +/- 0.010, N = 75.3715.3521. (CC) gcc options: -O3 -ffast-math -funroll-loops -fschedule-insns2 -fbranch-count-reg -fforce-addr -pipe -lncurses -lm
MinAvgMaxAMD Preferred Core Disabled9.221.328.6AMD Preferred Core Patched12.821.327.7OpenBenchmarking.orgWatts, Fewer Is BetterLAME MP3 Encoding 3.100CPU Power Consumption Monitor918273645

OpenVINO

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.1Model: Face Detection FP16-INT8 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched612182430SE +/- 0.06, N = 3SE +/- 0.01, N = 325.5725.481. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: Ultra FastAMD Preferred Core DisabledAMD Preferred Core Patched4080120160200SE +/- 0.24, N = 9SE +/- 0.15, N = 9193.53194.21
OpenBenchmarking.orgFrames Per Second Per Watt, More Is Betteruvg266 0.4.1Video Input: Bosphorus 1080p - Video Preset: Ultra FastAMD Preferred Core DisabledAMD Preferred Core Patched0.8981.7962.6943.5924.493.9903.991
MinAvgMaxAMD Preferred Core Disabled11.348.597.5AMD Preferred Core Patched11.848.797.6OpenBenchmarking.orgWatts, Fewer Is Betteruvg266 0.4.1CPU Power Consumption Monitor20406080100

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched90180270360450SE +/- 0.38, N = 3SE +/- 0.23, N = 3394.55395.90

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched510152025SE +/- 0.02, N = 3SE +/- 0.01, N = 320.2620.19
MinAvgMaxAMD Preferred Core Disabled10.687.9108.5AMD Preferred Core Patched12.888.0107.7OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor20406080100

OpenRadioss

OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2023.09.15Model: Bird Strike on WindshieldAMD Preferred Core DisabledAMD Preferred Core Patched306090120150SE +/- 1.73, N = 3SE +/- 2.07, N = 3150.00150.50
MinAvgMaxAMD Preferred Core Disabled10.395.1104.4AMD Preferred Core Patched12.395.4104.4OpenBenchmarking.orgWatts, Fewer Is BetterOpenRadioss 2023.09.15CPU Power Consumption Monitor20406080100

Renaissance

Renaissance is a suite of benchmarks designed to test the Java JVM from Apache Spark to a Twitter-like service to Scala and other features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.14Test: ALS Movie LensAMD Preferred Core DisabledAMD Preferred Core Patched2K4K6K8K10KSE +/- 10.91, N = 3SE +/- 55.33, N = 39125.29094.9MIN: 9105.29 / MAX: 9809.83MIN: 8997.96 / MAX: 9864.17
MinAvgMaxAMD Preferred Core Disabled12.942.364.2AMD Preferred Core Patched11.542.765.8OpenBenchmarking.orgWatts, Fewer Is BetterRenaissance 0.14CPU Power Consumption Monitor20406080100

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 9.12-MR1Java Test: TradebeansAMD Preferred Core DisabledAMD Preferred Core Patched400800120016002000SE +/- 5.72, N = 7SE +/- 5.35, N = 718441850
MinAvgMaxAMD Preferred Core Disabled9.630.859.7AMD Preferred Core Patched11.130.759.8OpenBenchmarking.orgWatts, Fewer Is BetterDaCapo Benchmark 9.12-MR1CPU Power Consumption Monitor1632486480

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: MediumAMD Preferred Core DisabledAMD Preferred Core Patched48121620SE +/- 0.01, N = 3SE +/- 0.01, N = 315.5515.60
OpenBenchmarking.orgFrames Per Second Per Watt, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: MediumAMD Preferred Core DisabledAMD Preferred Core Patched0.03130.06260.09390.12520.15650.1390.138
MinAvgMaxAMD Preferred Core Disabled9.9111.9126.4AMD Preferred Core Patched12.5112.8126.8OpenBenchmarking.orgWatts, Fewer Is Betteruvg266 0.4.1CPU Power Consumption Monitor4080120160200

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: Super FastAMD Preferred Core DisabledAMD Preferred Core Patched1326395265SE +/- 0.04, N = 5SE +/- 0.02, N = 559.7359.921. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterKvazaar 2.2Video Input: Bosphorus 4K - Video Preset: Super FastAMD Preferred Core DisabledAMD Preferred Core Patched0.16740.33480.50220.66960.8370.7440.743
MinAvgMaxAMD Preferred Core Disabled12.480.3113.7AMD Preferred Core Patched10.180.6113.6OpenBenchmarking.orgWatts, Fewer Is BetterKvazaar 2.2CPU Power Consumption Monitor4080120160200

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.1Algorithm: SHA256AMD Preferred Core DisabledAMD Preferred Core Patched7000M14000M21000M28000M35000MSE +/- 29855284.80, N = 3SE +/- 53681954.31, N = 333399261110332940333131. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
OpenBenchmarking.orgbyte/s Per Watt, More Is BetterOpenSSL 3.1Algorithm: SHA256AMD Preferred Core DisabledAMD Preferred Core Patched70M140M210M280M350M316837104.85313433150.96
MinAvgMaxAMD Preferred Core Disabled11.5105.4129.0AMD Preferred Core Patched10.3106.2128.8OpenBenchmarking.orgWatts, Fewer Is BetterOpenSSL 3.1CPU Power Consumption Monitor4080120160200

OpenVINO

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.1Model: Road Segmentation ADAS FP16 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched90180270360450SE +/- 1.87, N = 3SE +/- 0.69, N = 3435.18436.541. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenRadioss

OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2023.09.15Model: Bumper BeamAMD Preferred Core DisabledAMD Preferred Core Patched20406080100SE +/- 0.10, N = 3SE +/- 0.17, N = 390.7591.03
MinAvgMaxAMD Preferred Core Disabled10.3100.7107.4AMD Preferred Core Patched11.8100.9107.9OpenBenchmarking.orgWatts, Fewer Is BetterOpenRadioss 2023.09.15CPU Power Consumption Monitor20406080100

Memcached

Memcached is a high performance, distributed memory object caching system. This Memcached test profiles makes use of memtier_benchmark for excuting this CPU/memory-focused server benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps/sec, More Is BetterMemcached 1.6.19Set To Get Ratio: 1:10AMD Preferred Core DisabledAMD Preferred Core Patched800K1600K2400K3200K4000KSE +/- 5101.70, N = 3SE +/- 3231.14, N = 33776936.513765364.831. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre
OpenBenchmarking.orgOps/sec Per Watt, More Is BetterMemcached 1.6.19Set To Get Ratio: 1:10AMD Preferred Core DisabledAMD Preferred Core Patched9K18K27K36K45K43777.2743382.66
MinAvgMaxAMD Preferred Core Disabled12.686.3100.0AMD Preferred Core Patched11.986.8100.6OpenBenchmarking.orgWatts, Fewer Is BetterMemcached 1.6.19CPU Power Consumption Monitor20406080100

OpenVINO

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.1Model: Face Detection FP16 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched3691215SE +/- 0.00, N = 3SE +/- 0.01, N = 313.3513.311. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

RocksDB

This is a benchmark of Meta/Facebook's RocksDB as an embeddable persistent key-value store for fast storage based on Google's LevelDB. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 8.0Test: Random ReadAMD Preferred Core DisabledAMD Preferred Core Patched30M60M90M120M150MSE +/- 282168.48, N = 3SE +/- 238200.76, N = 31309004211305301741. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread
OpenBenchmarking.orgOp/s Per Watt, More Is BetterRocksDB 8.0Test: Random ReadAMD Preferred Core DisabledAMD Preferred Core Patched200K400K600K800K1000K1135514.711137280.10
MinAvgMaxAMD Preferred Core Disabled12.9115.3123.4AMD Preferred Core Patched12.1114.8122.9OpenBenchmarking.orgWatts, Fewer Is BetterRocksDB 8.0CPU Power Consumption Monitor4080120160200

Neural Magic DeepSparse

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched1530456075SE +/- 0.07, N = 3SE +/- 0.03, N = 368.0367.84
MinAvgMaxAMD Preferred Core Disabled11.893.1108.8AMD Preferred Core Patched13.092.7108.9OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor20406080100

TiDB Community Server

This is a PingCAP TiDB Community Server benchmark facilitated using the sysbench OLTP database benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Second, More Is BetterTiDB Community Server 7.3Test: oltp_read_write - Threads: 1AMD Preferred Core DisabledAMD Preferred Core Patched9001800270036004500SE +/- 7.98, N = 3SE +/- 6.08, N = 339934004
OpenBenchmarking.orgQueries Per Second Per Watt, More Is BetterTiDB Community Server 7.3Test: oltp_read_write - Threads: 1AMD Preferred Core DisabledAMD Preferred Core Patched2040608010097.5497.48
OpenBenchmarking.orgWatts, Fewer Is BetterTiDB Community Server 7.3CPU Power Consumption MonitorAMD Preferred Core DisabledAMD Preferred Core Patched1020304050Min: 11.93 / Avg: 40.94 / Max: 47.96Min: 10.07 / Avg: 41.08 / Max: 48.82

OpenVINO

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.1Model: Road Segmentation ADAS FP16 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched510152025SE +/- 0.08, N = 3SE +/- 0.03, N = 318.3418.29MIN: 10.04 / MAX: 27.73MIN: 11.41 / MAX: 26.841. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
MinAvgMaxAMD Preferred Core Disabled9.9112.2128.3AMD Preferred Core Patched11.9113.2127.1OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.1CPU Power Consumption Monitor4080120160200

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.1Model: Road Segmentation ADAS FP16-INT8 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched110220330440550SE +/- 1.05, N = 3SE +/- 1.48, N = 3523.99522.571. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched306090120150SE +/- 0.12, N = 3SE +/- 0.06, N = 3117.53117.85

OpenVINO

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.1Model: Road Segmentation ADAS FP16-INT8 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched48121620SE +/- 0.03, N = 3SE +/- 0.04, N = 315.2415.28MIN: 9.14 / MAX: 20.61MIN: 9.22 / MAX: 20.381. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
MinAvgMaxAMD Preferred Core Disabled12.0107.6117.4AMD Preferred Core Patched11.8107.5120.7OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.1CPU Power Consumption Monitor4080120160200

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.1Model: Face Detection FP16 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched130260390520650SE +/- 0.23, N = 3SE +/- 0.29, N = 3596.92598.44MIN: 573.91 / MAX: 624.05MIN: 573.8 / MAX: 625.981. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
MinAvgMaxAMD Preferred Core Disabled11.0116.3127.2AMD Preferred Core Patched11.3116.9127.4OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.1CPU Power Consumption Monitor4080120160200

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched306090120150SE +/- 0.06, N = 3SE +/- 0.39, N = 3111.92112.20

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 32 - Buffer Length: 256 - Filter Length: 512AMD Preferred Core DisabledAMD Preferred Core Patched80M160M240M320M400MSE +/- 332465.54, N = 3SE +/- 77960.10, N = 33723400003714133331. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
OpenBenchmarking.orgsamples/s Per Watt, More Is BetterLiquid-DSP 1.6Threads: 32 - Buffer Length: 256 - Filter Length: 512AMD Preferred Core DisabledAMD Preferred Core Patched800K1600K2400K3200K4000K3658760.713650205.13
MinAvgMaxAMD Preferred Core Disabled12.8101.8116.1AMD Preferred Core Patched12.8101.8115.8OpenBenchmarking.orgWatts, Fewer Is BetterLiquid-DSP 1.6CPU Power Consumption Monitor4080120160200

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.1Algorithm: ChaCha20AMD Preferred Core DisabledAMD Preferred Core Patched30000M60000M90000M120000M150000MSE +/- 33921017.78, N = 3SE +/- 67400734.64, N = 31257751432771254637724531. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
OpenBenchmarking.orgbyte/s Per Watt, More Is BetterOpenSSL 3.1Algorithm: ChaCha20AMD Preferred Core DisabledAMD Preferred Core Patched300M600M900M1200M1500M1233007369.281227653587.32
MinAvgMaxAMD Preferred Core Disabled10.0102.0116.0AMD Preferred Core Patched11.9102.2115.8OpenBenchmarking.orgWatts, Fewer Is BetterOpenSSL 3.1CPU Power Consumption Monitor4080120160200

Kvazaar

This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterKvazaar 2.2Video Input: Bosphorus 1080p - Video Preset: Ultra FastAMD Preferred Core DisabledAMD Preferred Core Patched60120180240300SE +/- 0.53, N = 11SE +/- 0.21, N = 11278.10278.791. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterKvazaar 2.2Video Input: Bosphorus 1080p - Video Preset: Ultra FastAMD Preferred Core DisabledAMD Preferred Core Patched2468107.1407.092
MinAvgMaxAMD Preferred Core Disabled6.539.088.6AMD Preferred Core Patched10.939.389.2OpenBenchmarking.orgWatts, Fewer Is BetterKvazaar 2.2CPU Power Consumption Monitor20406080100

Renaissance

Renaissance is a suite of benchmarks designed to test the Java JVM from Apache Spark to a Twitter-like service to Scala and other features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.14Test: Apache Spark BayesAMD Preferred Core DisabledAMD Preferred Core Patched2004006008001000SE +/- 1.49, N = 3SE +/- 0.50, N = 3905.4903.2MIN: 669.85 / MAX: 908.35MIN: 671.52 / MAX: 903.86
MinAvgMaxAMD Preferred Core Disabled10.663.381.5AMD Preferred Core Patched12.964.886.5OpenBenchmarking.orgWatts, Fewer Is BetterRenaissance 0.14CPU Power Consumption Monitor20406080100

Neural Magic DeepSparse

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched1632486480SE +/- 0.04, N = 3SE +/- 0.24, N = 371.4171.24
MinAvgMaxAMD Preferred Core Disabled11.895.9123.5AMD Preferred Core Patched11.099.6123.7OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor4080120160200

OpenVINO

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.1Model: Handwritten English Recognition FP16 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched510152025SE +/- 0.13, N = 3SE +/- 0.18, N = 321.3721.42MIN: 14.43 / MAX: 30.18MIN: 14.5 / MAX: 31.491. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
MinAvgMaxAMD Preferred Core Disabled12.9123.2136.1AMD Preferred Core Patched11.8124.6136.5OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.1CPU Power Consumption Monitor4080120160200

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.1Model: Vehicle Detection FP16-INT8 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched30060090012001500SE +/- 2.26, N = 3SE +/- 1.98, N = 31609.061605.401. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Blender

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.6Blend File: Pabellon Barcelona - Compute: CPU-OnlyAMD Preferred Core DisabledAMD Preferred Core Patched4080120160200SE +/- 0.36, N = 3SE +/- 0.25, N = 3170.05169.67
MinAvgMaxAMD Preferred Core Disabled13.0122.6127.4AMD Preferred Core Patched11.7123.0127.8OpenBenchmarking.orgWatts, Fewer Is BetterBlender 3.6CPU Power Consumption Monitor4080120160200

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 6.0Encoder: libx265 - Scenario: Video On DemandAMD Preferred Core DisabledAMD Preferred Core Patched1530456075SE +/- 0.10, N = 3SE +/- 0.10, N = 367.9368.081. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
OpenBenchmarking.orgFPS Per Watt, More Is BetterFFmpeg 6.0Encoder: libx265 - Scenario: Video On DemandAMD Preferred Core DisabledAMD Preferred Core Patched0.30080.60160.90241.20321.5041.3371.336
MinAvgMaxAMD Preferred Core Disabled12.350.872.9AMD Preferred Core Patched12.851.073.3OpenBenchmarking.orgWatts, Fewer Is BetterFFmpeg 6.0CPU Power Consumption Monitor20406080100

OpenBenchmarking.orgSeconds, Fewer Is BetterFFmpeg 6.0Encoder: libx265 - Scenario: Video On DemandAMD Preferred Core DisabledAMD Preferred Core Patched20406080100SE +/- 0.17, N = 3SE +/- 0.17, N = 3111.51111.271. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: Very FastAMD Preferred Core DisabledAMD Preferred Core Patched1020304050SE +/- 0.05, N = 4SE +/- 0.05, N = 442.0442.13
OpenBenchmarking.orgFrames Per Second Per Watt, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: Very FastAMD Preferred Core DisabledAMD Preferred Core Patched0.1040.2080.3120.4160.520.4610.462
MinAvgMaxAMD Preferred Core Disabled11.791.3119.0AMD Preferred Core Patched11.191.3120.2OpenBenchmarking.orgWatts, Fewer Is Betteruvg266 0.4.1CPU Power Consumption Monitor4080120160200

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.1Algorithm: ChaCha20-Poly1305AMD Preferred Core DisabledAMD Preferred Core Patched20000M40000M60000M80000M100000MSE +/- 19815620.43, N = 3SE +/- 22768062.85, N = 389411023847892271094071. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
OpenBenchmarking.orgbyte/s Per Watt, More Is BetterOpenSSL 3.1Algorithm: ChaCha20-Poly1305AMD Preferred Core DisabledAMD Preferred Core Patched200M400M600M800M1000M850750623.28848023797.92
MinAvgMaxAMD Preferred Core Disabled11.9105.1118.4AMD Preferred Core Patched10.4105.2118.4OpenBenchmarking.orgWatts, Fewer Is BetterOpenSSL 3.1CPU Power Consumption Monitor4080120160200

SVT-AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 1.7Encoder Mode: Preset 12 - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched4080120160200SE +/- 0.32, N = 8SE +/- 0.24, N = 8169.23169.581. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterSVT-AV1 1.7Encoder Mode: Preset 12 - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched0.89121.78242.67363.56484.4563.9613.896
MinAvgMaxAMD Preferred Core Disabled11.342.791.0AMD Preferred Core Patched10.443.590.8OpenBenchmarking.orgWatts, Fewer Is BetterSVT-AV1 1.7CPU Power Consumption Monitor20406080100

OpenVINO

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.1Model: Weld Porosity Detection FP16-INT8 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched6001200180024003000SE +/- 1.81, N = 3SE +/- 1.29, N = 32586.202581.071. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.1Algorithm: AES-256-GCMAMD Preferred Core DisabledAMD Preferred Core Patched50000M100000M150000M200000M250000MSE +/- 36600929.56, N = 3SE +/- 21724266.95, N = 32098575283532102669878171. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
OpenBenchmarking.orgbyte/s Per Watt, More Is BetterOpenSSL 3.1Algorithm: AES-256-GCMAMD Preferred Core DisabledAMD Preferred Core Patched400M800M1200M1600M2000M1821175025.401813155081.01
MinAvgMaxAMD Preferred Core Disabled11.2115.2122.6AMD Preferred Core Patched11.8116.0123.2OpenBenchmarking.orgWatts, Fewer Is BetterOpenSSL 3.1CPU Power Consumption Monitor4080120160200

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 Preferred Core DisabledAMD Preferred Core Patched30K60K90K120K150KSE +/- 68.33, N = 3SE +/- 259.94, N = 31600311597311. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt
OpenBenchmarking.orgReal C/S Per Watt, More Is BetterJohn The Ripper 2023.03.14Test: WPA PSKAMD Preferred Core DisabledAMD Preferred Core Patched4008001200160020001642.531639.26
MinAvgMaxAMD Preferred Core Disabled10.497.4110.8AMD Preferred Core Patched12.697.4110.7OpenBenchmarking.orgWatts, Fewer Is BetterJohn The Ripper 2023.03.14CPU Power Consumption Monitor20406080100

x265

This is a simple test of the x265 encoder run on the CPU with 1080p and 4K options for H.265 video encode performance with x265. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betterx265 3.4Video Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched714212835SE +/- 0.02, N = 3SE +/- 0.18, N = 332.2032.261. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
OpenBenchmarking.orgFrames Per Second Per Watt, More Is Betterx265 3.4Video Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched0.09920.19840.29760.39680.4960.4410.441
MinAvgMaxAMD Preferred Core Disabled12.673.195.2AMD Preferred Core Patched11.973.196.2OpenBenchmarking.orgWatts, Fewer Is Betterx265 3.4CPU Power Consumption Monitor20406080100

OpenVINO

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.1Model: Handwritten English Recognition FP16 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched160320480640800SE +/- 4.57, N = 3SE +/- 6.39, N = 3747.40746.051. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: BERT-Large, NLP Question Answering - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched510152025SE +/- 0.01, N = 3SE +/- 0.03, N = 318.8918.86

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: BERT-Large, NLP Question Answering - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched1224364860SE +/- 0.03, N = 3SE +/- 0.08, N = 352.9353.02
MinAvgMaxAMD Preferred Core Disabled13.057.798.4AMD Preferred Core Patched12.157.898.4OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor20406080100

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: BERT-Large, NLP Question Answering, Sparse INT8 - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched306090120150SE +/- 0.18, N = 3SE +/- 0.14, N = 3112.59112.39

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: BERT-Large, NLP Question Answering, Sparse INT8 - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched246810SE +/- 0.0142, N = 3SE +/- 0.0114, N = 38.87938.8950
MinAvgMaxAMD Preferred Core Disabled11.043.868.7AMD Preferred Core Patched12.444.269.0OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor20406080100

Blender

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.6Blend File: Fishy Cat - Compute: CPU-OnlyAMD Preferred Core DisabledAMD Preferred Core Patched1530456075SE +/- 0.01, N = 3SE +/- 0.05, N = 368.5868.46
MinAvgMaxAMD Preferred Core Disabled10.9118.1128.8AMD Preferred Core Patched12.7118.7129.2OpenBenchmarking.orgWatts, Fewer Is BetterBlender 3.6CPU Power Consumption Monitor4080120160200

Apache Hadoop

This is a benchmark of the Apache Hadoop making use of its built-in name-node throughput benchmark (NNThroughputBenchmark). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps per sec, More Is BetterApache Hadoop 3.3.6Operation: Create - Threads: 20 - Files: 1000000AMD Preferred Core DisabledAMD Preferred Core Patched20K40K60K80K100KSE +/- 210.68, N = 3SE +/- 366.31, N = 38561585764
OpenBenchmarking.orgOps per sec Per Watt, More Is BetterApache Hadoop 3.3.6Operation: Create - Threads: 20 - Files: 1000000AMD Preferred Core DisabledAMD Preferred Core Patched50010001500200025002505.352505.67
MinAvgMaxAMD Preferred Core Disabled11.134.251.2AMD Preferred Core Patched10.834.251.5OpenBenchmarking.orgWatts, Fewer Is BetterApache Hadoop 3.3.6CPU Power Consumption Monitor1530456075

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgverify/s, More Is BetterOpenSSL 3.1Algorithm: RSA4096AMD Preferred Core DisabledAMD Preferred Core Patched80K160K240K320K400KSE +/- 101.07, N = 3SE +/- 51.43, N = 3358591.0357968.11. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
OpenBenchmarking.orgverify/s Per Watt, More Is BetterOpenSSL 3.1Algorithm: RSA4096AMD Preferred Core DisabledAMD Preferred Core Patched70014002100280035003317.313325.39
MinAvgMaxAMD Preferred Core Disabled13.0108.1137.1AMD Preferred Core Patched11.0107.6137.4OpenBenchmarking.orgWatts, Fewer Is BetterOpenSSL 3.1CPU Power Consumption Monitor4080120160200

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 4 - Buffer Length: 256 - Filter Length: 512AMD Preferred Core DisabledAMD Preferred Core Patched14M28M42M56M70MSE +/- 409704.09, N = 3SE +/- 873432.50, N = 365200667653130001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
OpenBenchmarking.orgsamples/s Per Watt, More Is BetterLiquid-DSP 1.6Threads: 4 - Buffer Length: 256 - Filter Length: 512AMD Preferred Core DisabledAMD Preferred Core Patched400K800K1200K1600K2000K1775351.851881565.43
MinAvgMaxAMD Preferred Core Disabled12.836.740.8AMD Preferred Core Patched12.934.740.7OpenBenchmarking.orgWatts, Fewer Is BetterLiquid-DSP 1.6CPU Power Consumption Monitor1122334455

Renaissance

Renaissance is a suite of benchmarks designed to test the Java JVM from Apache Spark to a Twitter-like service to Scala and other features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.14Test: In-Memory Database ShootoutAMD Preferred Core DisabledAMD Preferred Core Patched5001000150020002500SE +/- 27.17, N = 3SE +/- 24.73, N = 52526.92531.2MIN: 2294.89 / MAX: 2775.39MIN: 2257.54 / MAX: 2836.12
MinAvgMaxAMD Preferred Core Disabled10.847.262.4AMD Preferred Core Patched12.047.863.3OpenBenchmarking.orgWatts, Fewer Is BetterRenaissance 0.14CPU Power Consumption Monitor20406080100

Timed Linux Kernel Compilation

This test times how long it takes to build the Linux kernel in a default configuration (defconfig) for the architecture being tested or alternatively an allmodconfig for building all possible kernel modules for the build. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 6.1Build: defconfigAMD Preferred Core DisabledAMD Preferred Core Patched1020304050SE +/- 0.46, N = 3SE +/- 0.50, N = 343.0943.17
MinAvgMaxAMD Preferred Core Disabled11.183.2115.9AMD Preferred Core Patched11.982.9116.0OpenBenchmarking.orgWatts, Fewer Is BetterTimed Linux Kernel Compilation 6.1CPU Power Consumption Monitor4080120160200

Neural Magic DeepSparse

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: BERT-Large, NLP Question Answering, Sparse INT8 - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched48121620SE +/- 0.03, N = 3SE +/- 0.02, N = 317.2817.31
MinAvgMaxAMD Preferred Core Disabled10.680.5115.0AMD Preferred Core Patched13.082.1115.0OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor4080120160200

OpenVINO

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.1Model: Weld Porosity Detection FP16-INT8 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched246810SE +/- 0.00, N = 3SE +/- 0.00, N = 36.136.12MIN: 3.16 / MAX: 12.91MIN: 3.17 / MAX: 11.721. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
MinAvgMaxAMD Preferred Core Disabled11.5114.7124.2AMD Preferred Core Patched11.3114.9124.4OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.1CPU Power Consumption Monitor4080120160200

Apache Hadoop

This is a benchmark of the Apache Hadoop making use of its built-in name-node throughput benchmark (NNThroughputBenchmark). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps per sec, More Is BetterApache Hadoop 3.3.6Operation: Open - Threads: 20 - Files: 100000AMD Preferred Core DisabledAMD Preferred Core Patched200K400K600K800K1000KSE +/- 12247.66, N = 15SE +/- 9447.38, N = 15822149820810
OpenBenchmarking.orgOps per sec Per Watt, More Is BetterApache Hadoop 3.3.6Operation: Open - Threads: 20 - Files: 100000AMD Preferred Core DisabledAMD Preferred Core Patched6K12K18K24K30K28880.4728705.04
MinAvgMaxAMD Preferred Core Disabled12.328.554.6AMD Preferred Core Patched10.728.653.8OpenBenchmarking.orgWatts, Fewer Is BetterApache Hadoop 3.3.6CPU Power Consumption Monitor1530456075

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: BERT-Large, NLP Question Answering, Sparse INT8 - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched100200300400500SE +/- 0.86, N = 3SE +/- 0.50, N = 3462.47461.72

PyPerformance

PyPerformance is the reference Python performance benchmark suite. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: python_startupAMD Preferred Core DisabledAMD Preferred Core Patched246810SE +/- 0.01, N = 3SE +/- 0.01, N = 36.216.20
MinAvgMaxAMD Preferred Core Disabled11.927.028.1AMD Preferred Core Patched11.127.028.0OpenBenchmarking.orgWatts, Fewer Is BetterPyPerformance 1.0.0CPU Power Consumption Monitor918273645

Neural Magic DeepSparse

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched1326395265SE +/- 0.04, N = 3SE +/- 0.03, N = 357.4657.54
MinAvgMaxAMD Preferred Core Disabled10.180.4105.6AMD Preferred Core Patched12.982.7105.7OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor20406080100

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched48121620SE +/- 0.01, N = 3SE +/- 0.01, N = 317.4017.38

OpenRadioss

OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2023.09.15Model: INIVOL and Fluid Structure Interaction Drop ContainerAMD Preferred Core DisabledAMD Preferred Core Patched60120180240300SE +/- 1.01, N = 3SE +/- 1.57, N = 3274.18273.77
MinAvgMaxAMD Preferred Core Disabled11.694.1102.5AMD Preferred Core Patched12.494.6104.9OpenBenchmarking.orgWatts, Fewer Is BetterOpenRadioss 2023.09.15CPU Power Consumption Monitor20406080100

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched60120180240300SE +/- 0.35, N = 3SE +/- 0.29, N = 3267.26266.87

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched4080120160200SE +/- 0.06, N = 3SE +/- 0.04, N = 3180.59180.34

Apache Hadoop

This is a benchmark of the Apache Hadoop making use of its built-in name-node throughput benchmark (NNThroughputBenchmark). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps per sec, More Is BetterApache Hadoop 3.3.6Operation: Delete - Threads: 20 - Files: 1000000AMD Preferred Core DisabledAMD Preferred Core Patched30K60K90K120K150KSE +/- 244.85, N = 3SE +/- 406.98, N = 3136850137039
OpenBenchmarking.orgOps per sec Per Watt, More Is BetterApache Hadoop 3.3.6Operation: Delete - Threads: 20 - Files: 1000000AMD Preferred Core DisabledAMD Preferred Core Patched80016002400320040003597.743595.51
MinAvgMaxAMD Preferred Core Disabled9.838.052.5AMD Preferred Core Patched10.638.151.2OpenBenchmarking.orgWatts, Fewer Is BetterApache Hadoop 3.3.6CPU Power Consumption Monitor1530456075

OpenVINO

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.1Model: Handwritten English Recognition FP16-INT8 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched130260390520650SE +/- 1.19, N = 3SE +/- 1.78, N = 3589.30588.541. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Neural Magic DeepSparse

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched714212835SE +/- 0.04, N = 3SE +/- 0.03, N = 329.9229.96
MinAvgMaxAMD Preferred Core Disabled11.490.8105.8AMD Preferred Core Patched12.984.7105.4OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor20406080100

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched1020304050SE +/- 0.01, N = 3SE +/- 0.01, N = 344.2844.33
MinAvgMaxAMD Preferred Core Disabled9.490.0112.3AMD Preferred Core Patched12.793.0112.1OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor4080120160200

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: ResNet-50, Baseline - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched1.16612.33223.49834.66445.8305SE +/- 0.0048, N = 3SE +/- 0.0058, N = 35.18285.1765

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: ResNet-50, Baseline - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched4080120160200SE +/- 0.18, N = 3SE +/- 0.22, N = 3192.88193.11

SVT-AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 1.7Encoder Mode: Preset 8 - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched20406080100SE +/- 0.21, N = 5SE +/- 0.12, N = 579.5679.651. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterSVT-AV1 1.7Encoder Mode: Preset 8 - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched0.27340.54680.82021.09361.3671.1761.215
MinAvgMaxAMD Preferred Core Disabled11.167.6111.7AMD Preferred Core Patched11.465.6112.0OpenBenchmarking.orgWatts, Fewer Is BetterSVT-AV1 1.7CPU Power Consumption Monitor4080120160200

Unpacking Firefox

This simple test profile measures how long it takes to extract the .tar.xz source package of the Mozilla Firefox Web Browser. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterUnpacking Firefox 84.0Extracting: firefox-84.0.source.tar.xzAMD Preferred Core DisabledAMD Preferred Core Patched48121620SE +/- 0.05, N = 4SE +/- 0.04, N = 414.8414.85
MinAvgMaxAMD Preferred Core Disabled10.625.228.9AMD Preferred Core Patched12.825.331.3OpenBenchmarking.orgWatts, Fewer Is BetterUnpacking Firefox 84.0CPU Power Consumption Monitor918273645

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: bcryptAMD Preferred Core DisabledAMD Preferred Core Patched9K18K27K36K45KSE +/- 33.35, N = 3SE +/- 17.90, N = 342032419841. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt
OpenBenchmarking.orgReal C/S Per Watt, More Is BetterJohn The Ripper 2023.03.14Test: bcryptAMD Preferred Core DisabledAMD Preferred Core Patched110220330440550490.40489.38

Neural Magic DeepSparse

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Text Classification, DistilBERT mnli - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched3691215SE +/- 0.0100, N = 3SE +/- 0.0256, N = 39.04599.0562
MinAvgMaxAMD Preferred Core Disabled12.174.188.8AMD Preferred Core Patched12.174.089.8OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor20406080100

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: Ultra FastAMD Preferred Core DisabledAMD Preferred Core Patched1224364860SE +/- 0.04, N = 5SE +/- 0.08, N = 553.5453.60
OpenBenchmarking.orgFrames Per Second Per Watt, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: Ultra FastAMD Preferred Core DisabledAMD Preferred Core Patched0.14510.29020.43530.58040.72550.6160.645
MinAvgMaxAMD Preferred Core Disabled12.686.9118.6AMD Preferred Core Patched12.983.2117.6OpenBenchmarking.orgWatts, Fewer Is Betteruvg266 0.4.1CPU Power Consumption Monitor4080120160200

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Text Classification, DistilBERT mnli - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched20406080100SE +/- 0.12, N = 3SE +/- 0.31, N = 3110.50110.37

OpenVINO

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.1Model: Handwritten English Recognition FP16-INT8 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched612182430SE +/- 0.06, N = 3SE +/- 0.08, N = 327.1127.14MIN: 18.08 / MAX: 35.83MIN: 18.44 / MAX: 35.41. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
MinAvgMaxAMD Preferred Core Disabled13.1118.3128.5AMD Preferred Core Patched13.2118.9129.7OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.1CPU Power Consumption Monitor4080120160200

nginx

This is a benchmark of the lightweight Nginx HTTP(S) web-server. This Nginx web server benchmark test profile makes use of the wrk program for facilitating the HTTP requests over a fixed period time with a configurable number of concurrent clients/connections. HTTPS with a self-signed OpenSSL certificate is used by this test for local benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.23.2Connections: 500AMD Preferred Core DisabledAMD Preferred Core Patched30K60K90K120K150KSE +/- 102.64, N = 3SE +/- 78.91, N = 3146476.98146317.311. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2
OpenBenchmarking.orgRequests Per Second Per Watt, More Is Betternginx 1.23.2Connections: 500AMD Preferred Core DisabledAMD Preferred Core Patched300600900120015001366.381358.24
MinAvgMaxAMD Preferred Core Disabled8.7107.2116.4AMD Preferred Core Patched11.6107.7116.8OpenBenchmarking.orgWatts, Fewer Is Betternginx 1.23.2CPU Power Consumption Monitor4080120160200

uvg266

uvg266 is an open-source VVC/H.266 (Versatile Video Coding) encoder based on Kvazaar as part of the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: Super FastAMD Preferred Core DisabledAMD Preferred Core Patched1020304050SE +/- 0.04, N = 4SE +/- 0.03, N = 445.9145.86
OpenBenchmarking.orgFrames Per Second Per Watt, More Is Betteruvg266 0.4.1Video Input: Bosphorus 4K - Video Preset: Super FastAMD Preferred Core DisabledAMD Preferred Core Patched0.11140.22280.33420.44560.5570.4950.492
MinAvgMaxAMD Preferred Core Disabled12.692.8122.0AMD Preferred Core Patched11.293.1122.0OpenBenchmarking.orgWatts, Fewer Is Betteruvg266 0.4.1CPU Power Consumption Monitor4080120160200

Neural Magic DeepSparse

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched48121620SE +/- 0.04, N = 3SE +/- 0.05, N = 316.5716.59
MinAvgMaxAMD Preferred Core Disabled11.971.285.5AMD Preferred Core Patched11.071.286.4OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor20406080100

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched1428425670SE +/- 0.13, N = 3SE +/- 0.20, N = 360.3260.26

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched80160240320400SE +/- 0.16, N = 3SE +/- 0.31, N = 3372.05372.42
MinAvgMaxAMD Preferred Core Disabled11.688.5119.0AMD Preferred Core Patched11.488.9119.7OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor4080120160200

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: CV Detection, YOLOv5s COCO - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched20406080100SE +/- 0.02, N = 3SE +/- 0.08, N = 389.3289.41

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: CV Detection, YOLOv5s COCO - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched3691215SE +/- 0.00, N = 3SE +/- 0.01, N = 311.1911.18
MinAvgMaxAMD Preferred Core Disabled10.885.2100.2AMD Preferred Core Patched10.086.0100.6OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor20406080100

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: ResNet-50, Sparse INT8 - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched0.59631.19261.78892.38522.9815SE +/- 0.0067, N = 3SE +/- 0.0093, N = 32.64762.6501
MinAvgMaxAMD Preferred Core Disabled11.093.2108.0AMD Preferred Core Patched13.193.0107.9OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor20406080100

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: BERT-Large, NLP Question Answering - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched612182430SE +/- 0.03, N = 3SE +/- 0.02, N = 326.9526.92

Apache Hadoop

This is a benchmark of the Apache Hadoop making use of its built-in name-node throughput benchmark (NNThroughputBenchmark). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps per sec, More Is BetterApache Hadoop 3.3.6Operation: Delete - Threads: 50 - Files: 1000000AMD Preferred Core DisabledAMD Preferred Core Patched30K60K90K120K150KSE +/- 1248.44, N = 3SE +/- 1423.03, N = 3130289130409
OpenBenchmarking.orgOps per sec Per Watt, More Is BetterApache Hadoop 3.3.6Operation: Delete - Threads: 50 - Files: 1000000AMD Preferred Core DisabledAMD Preferred Core Patched70014002100280035003351.183344.91
MinAvgMaxAMD Preferred Core Disabled11.338.951.8AMD Preferred Core Patched12.939.051.0OpenBenchmarking.orgWatts, Fewer Is BetterApache Hadoop 3.3.6CPU Power Consumption Monitor1530456075

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: ResNet-50, Sparse INT8 - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched6001200180024003000SE +/- 7.50, N = 3SE +/- 10.38, N = 33010.963008.22

Blender

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.6Blend File: BMW27 - Compute: CPU-OnlyAMD Preferred Core DisabledAMD Preferred Core Patched1224364860SE +/- 0.04, N = 3SE +/- 0.04, N = 355.0555.10
MinAvgMaxAMD Preferred Core Disabled13.0113.4124.1AMD Preferred Core Patched11.2113.7125.0OpenBenchmarking.orgWatts, Fewer Is BetterBlender 3.6CPU Power Consumption Monitor4080120160200

Neural Magic DeepSparse

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: BERT-Large, NLP Question Answering - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched60120180240300SE +/- 0.17, N = 3SE +/- 0.15, N = 3296.22296.48
MinAvgMaxAMD Preferred Core Disabled12.057.6115.9AMD Preferred Core Patched12.258.1115.9OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor4080120160200

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: ResNet-50, Baseline - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched60120180240300SE +/- 0.08, N = 3SE +/- 0.15, N = 3267.08266.86

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: ResNet-50, Baseline - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched714212835SE +/- 0.01, N = 3SE +/- 0.02, N = 329.9329.96

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched80160240320400SE +/- 0.34, N = 3SE +/- 0.12, N = 3372.63372.36
MinAvgMaxAMD Preferred Core Disabled12.988.7119.2AMD Preferred Core Patched11.789.3119.7OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor4080120160200

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched510152025SE +/- 0.01, N = 3SE +/- 0.01, N = 321.4521.46

OpenRadioss

OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenRadioss 2023.09.15Model: Chrysler Neon 1MAMD Preferred Core DisabledAMD Preferred Core Patched2004006008001000SE +/- 0.89, N = 3SE +/- 0.89, N = 3809.03808.48
OpenBenchmarking.orgWatts, Fewer Is BetterOpenRadioss 2023.09.15CPU Power Consumption MonitorAMD Preferred Core DisabledAMD Preferred Core Patched1530456075Min: 11.63 / Avg: 70.13 / Max: 79.49Min: 11.07 / Avg: 70.23 / Max: 77.57

Neural Magic DeepSparse

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: CV Detection, YOLOv5s COCO, Sparse INT8 - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched3691215SE +/- 0.01, N = 3SE +/- 0.01, N = 311.0511.05
MinAvgMaxAMD Preferred Core Disabled13.385.299.7AMD Preferred Core Patched12.685.799.8OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor20406080100

OpenVINO

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.1Model: Weld Porosity Detection FP16 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched30060090012001500SE +/- 0.88, N = 3SE +/- 1.24, N = 31345.161344.261. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: CV Detection, YOLOv5s COCO, Sparse INT8 - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched20406080100SE +/- 0.04, N = 3SE +/- 0.05, N = 390.4590.51

OpenVINO

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.1Model: Machine Translation EN To DE FP16 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched306090120150SE +/- 0.31, N = 3SE +/- 0.17, N = 3129.89129.971. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Neural Magic DeepSparse

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: CV Classification, ResNet-50 ImageNet - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched1.16632.33263.49894.66525.8315SE +/- 0.0033, N = 3SE +/- 0.0026, N = 35.18045.1835
MinAvgMaxAMD Preferred Core Disabled12.381.295.5AMD Preferred Core Patched11.178.894.9OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor20406080100

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.1Algorithm: AES-128-GCMAMD Preferred Core DisabledAMD Preferred Core Patched50000M100000M150000M200000M250000MSE +/- 21357128.87, N = 3SE +/- 64573682.12, N = 32463002916202461537631331. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
OpenBenchmarking.orgbyte/s Per Watt, More Is BetterOpenSSL 3.1Algorithm: AES-128-GCMAMD Preferred Core DisabledAMD Preferred Core Patched400M800M1200M1600M2000M2073186003.662074825863.28
MinAvgMaxAMD Preferred Core Disabled13.0118.8126.6AMD Preferred Core Patched11.9118.6126.7OpenBenchmarking.orgWatts, Fewer Is BetterOpenSSL 3.1CPU Power Consumption Monitor4080120160200

OpenBenchmarking.orgsign/s, More Is BetterOpenSSL 3.1Algorithm: RSA4096AMD Preferred Core DisabledAMD Preferred Core Patched3K6K9K12K15KSE +/- 29.06, N = 3SE +/- 12.82, N = 314049.114041.11. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -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: BlowfishAMD Preferred Core DisabledAMD Preferred Core Patched9K18K27K36K45KSE +/- 5.77, N = 3SE +/- 4.67, N = 342182421581. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -lm -lrt -lz -ldl -lcrypt
OpenBenchmarking.orgReal C/S Per Watt, More Is BetterJohn The Ripper 2023.03.14Test: BlowfishAMD Preferred Core DisabledAMD Preferred Core Patched110220330440550494.38492.33
MinAvgMaxAMD Preferred Core Disabled12.385.396.7AMD Preferred Core Patched12.685.897.0OpenBenchmarking.orgWatts, Fewer Is BetterJohn The Ripper 2023.03.14CPU Power Consumption Monitor20406080100

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: CV Classification, ResNet-50 ImageNet - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched4080120160200SE +/- 0.12, N = 3SE +/- 0.10, N = 3192.96192.86

Memcached

Memcached is a high performance, distributed memory object caching system. This Memcached test profiles makes use of memtier_benchmark for excuting this CPU/memory-focused server benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps/sec, More Is BetterMemcached 1.6.19Set To Get Ratio: 1:100AMD Preferred Core DisabledAMD Preferred Core Patched800K1600K2400K3200K4000KSE +/- 8544.53, N = 3SE +/- 9906.74, N = 33748498.213746607.051. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre
OpenBenchmarking.orgOps/sec Per Watt, More Is BetterMemcached 1.6.19Set To Get Ratio: 1:100AMD Preferred Core DisabledAMD Preferred Core Patched9K18K27K36K45K43701.1743502.75
MinAvgMaxAMD Preferred Core Disabled9.385.8100.3AMD Preferred Core Patched12.186.1100.7OpenBenchmarking.orgWatts, Fewer Is BetterMemcached 1.6.19CPU Power Consumption Monitor20406080100

OpenVINO

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.1Model: Machine Translation EN To DE FP16 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched1428425670SE +/- 0.14, N = 3SE +/- 0.08, N = 361.4761.44MIN: 27.88 / MAX: 72.47MIN: 31.36 / MAX: 72.291. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
MinAvgMaxAMD Preferred Core Disabled12.7102.1112.6AMD Preferred Core Patched9.9102.6113.3OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.1CPU Power Consumption Monitor4080120160200

Neural Magic DeepSparse

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched20406080100SE +/- 0.09, N = 3SE +/- 0.08, N = 387.0787.10
MinAvgMaxAMD Preferred Core Disabled10.786.4121.8AMD Preferred Core Patched12.988.6121.9OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor4080120160200

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched510152025SE +/- 0.01, N = 3SE +/- 0.01, N = 321.4721.46

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test profile makes use of the built-in "openssl speed" benchmarking capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.1Algorithm: SHA512AMD Preferred Core DisabledAMD Preferred Core Patched2000M4000M6000M8000M10000MSE +/- 4002929.96, N = 3SE +/- 649251.32, N = 310577375097105812458071. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl
OpenBenchmarking.orgbyte/s Per Watt, More Is BetterOpenSSL 3.1Algorithm: SHA512AMD Preferred Core DisabledAMD Preferred Core Patched20M40M60M80M100M100559841.80100534903.79
MinAvgMaxAMD Preferred Core Disabled11.4105.2118.9AMD Preferred Core Patched12.9105.2119.7OpenBenchmarking.orgWatts, Fewer Is BetterOpenSSL 3.1CPU Power Consumption Monitor4080120160200

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.orgSeconds, Fewer Is BetterFFmpeg 6.0Encoder: libx265 - Scenario: PlatformAMD Preferred Core DisabledAMD Preferred Core Patched306090120150SE +/- 0.11, N = 3SE +/- 0.19, N = 3111.72111.761. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

Blender

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.6Blend File: Classroom - Compute: CPU-OnlyAMD Preferred Core DisabledAMD Preferred Core Patched306090120150SE +/- 0.07, N = 3SE +/- 0.13, N = 3140.74140.79
MinAvgMaxAMD Preferred Core Disabled12.4117.3122.1AMD Preferred Core Patched11.8117.6122.7OpenBenchmarking.orgWatts, Fewer Is BetterBlender 3.6CPU Power Consumption Monitor4080120160200

Neural Magic DeepSparse

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched50100150200250SE +/- 0.33, N = 3SE +/- 0.06, N = 3210.00209.93
MinAvgMaxAMD Preferred Core Disabled12.384.5112.0AMD Preferred Core Patched11.886.6111.7OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor4080120160200

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 6.0Encoder: libx265 - Scenario: PlatformAMD Preferred Core DisabledAMD Preferred Core Patched1530456075SE +/- 0.07, N = 3SE +/- 0.11, N = 367.8067.781. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
OpenBenchmarking.orgFPS Per Watt, More Is BetterFFmpeg 6.0Encoder: libx265 - Scenario: PlatformAMD Preferred Core DisabledAMD Preferred Core Patched0.29990.59980.89971.19961.49951.3331.332
MinAvgMaxAMD Preferred Core Disabled12.250.972.6AMD Preferred Core Patched10.850.974.5OpenBenchmarking.orgWatts, Fewer Is BetterFFmpeg 6.0CPU Power Consumption Monitor20406080100

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 16 - Buffer Length: 256 - Filter Length: 512AMD Preferred Core DisabledAMD Preferred Core Patched50M100M150M200M250MSE +/- 770216.42, N = 3SE +/- 695006.00, N = 32522300002523000001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
OpenBenchmarking.orgsamples/s Per Watt, More Is BetterLiquid-DSP 1.6Threads: 16 - Buffer Length: 256 - Filter Length: 512AMD Preferred Core DisabledAMD Preferred Core Patched700K1400K2100K2800K3500K3191296.953315337.52
MinAvgMaxAMD Preferred Core Disabled10.179.089.8AMD Preferred Core Patched11.976.190.1OpenBenchmarking.orgWatts, Fewer Is BetterLiquid-DSP 1.6CPU Power Consumption Monitor20406080100

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched714212835SE +/- 0.02, N = 3SE +/- 0.02, N = 328.9828.97

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched816243240SE +/- 0.02, N = 3SE +/- 0.03, N = 334.5034.51
MinAvgMaxAMD Preferred Core Disabled13.179.397.0AMD Preferred Core Patched11.979.297.3OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor20406080100

OpenVINO

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.1Model: Person Detection FP16 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched20406080100SE +/- 0.38, N = 3SE +/- 0.40, N = 384.1184.13MIN: 28.34 / MAX: 114.33MIN: 49.14 / MAX: 110.451. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
MinAvgMaxAMD Preferred Core Disabled11.793.4107.4AMD Preferred Core Patched12.593.4107.5OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.1CPU Power Consumption Monitor20406080100

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched918273645SE +/- 0.06, N = 3SE +/- 0.01, N = 338.0838.09

OpenVINO

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2023.1Model: Person Detection FP16 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched20406080100SE +/- 0.43, N = 3SE +/- 0.45, N = 395.0395.011. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched48121620SE +/- 0.02, N = 3SE +/- 0.00, N = 317.3817.37

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Synchronous Single-StreamAMD Preferred Core DisabledAMD Preferred Core Patched1326395265SE +/- 0.06, N = 3SE +/- 0.01, N = 357.5557.56
MinAvgMaxAMD Preferred Core Disabled12.178.0105.3AMD Preferred Core Patched12.482.6105.4OpenBenchmarking.orgWatts, Fewer Is BetterNeural Magic DeepSparse 1.5CPU Power Consumption Monitor20406080100

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-StreamAMD Preferred Core DisabledAMD Preferred Core Patched20406080100SE +/- 0.08, N = 3SE +/- 0.08, N = 391.8391.82

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.orgSeconds, Fewer Is BetterFFmpeg 6.0Encoder: libx265 - Scenario: UploadAMD Preferred Core DisabledAMD Preferred Core Patched20406080100SE +/- 0.25, N = 3SE +/- 0.31, N = 376.0576.061. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

OpenBenchmarking.orgFPS, More Is BetterFFmpeg 6.0Encoder: libx265 - Scenario: UploadAMD Preferred Core DisabledAMD Preferred Core Patched816243240SE +/- 0.11, N = 3SE +/- 0.14, N = 333.2033.201. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma
OpenBenchmarking.orgFPS Per Watt, More Is BetterFFmpeg 6.0Encoder: libx265 - Scenario: UploadAMD Preferred Core DisabledAMD Preferred Core Patched0.16880.33760.50640.67520.8440.7500.749
MinAvgMaxAMD Preferred Core Disabled12.744.377.9AMD Preferred Core Patched11.144.391.4OpenBenchmarking.orgWatts, Fewer Is BetterFFmpeg 6.0CPU Power Consumption Monitor20406080100

AOM AV1

This is a test of the AOMedia AV1 encoder (libaom) developed by AOMedia and Google as the AV1 Codec Library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.7Encoder Mode: Speed 0 Two-Pass - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched0.09230.18460.27690.36920.4615SE +/- 0.00, N = 3SE +/- 0.00, N = 30.410.411. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterAOM AV1 3.7Encoder Mode: Speed 0 Two-Pass - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched0.00110.00220.00330.00440.00550.0050.005
MinAvgMaxAMD Preferred Core Disabled12.388.5138.7AMD Preferred Core Patched11.088.9138.6OpenBenchmarking.orgWatts, Fewer Is BetterAOM AV1 3.7CPU Power Consumption Monitor4080120160200

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.7Encoder Mode: Speed 6 Two-Pass - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched612182430SE +/- 0.22, N = 3SE +/- 0.03, N = 324.8524.851. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterAOM AV1 3.7Encoder Mode: Speed 6 Two-Pass - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched0.08010.16020.24030.32040.40050.3560.351
MinAvgMaxAMD Preferred Core Disabled11.569.990.5AMD Preferred Core Patched10.570.889.3OpenBenchmarking.orgWatts, Fewer Is BetterAOM AV1 3.7CPU Power Consumption Monitor20406080100

OpenVINO

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.1Model: Face Detection Retail FP16 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched0.5671.1341.7012.2682.835SE +/- 0.01, N = 3SE +/- 0.01, N = 32.522.52MIN: 1.34 / MAX: 9.5MIN: 1.35 / MAX: 7.781. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
MinAvgMaxAMD Preferred Core Disabled12.5113.0123.1AMD Preferred Core Patched12.7112.9124.4OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.1CPU Power Consumption Monitor4080120160200

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.1Model: Vehicle Detection FP16-INT8 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched1.10932.21863.32794.43725.5465SE +/- 0.01, N = 3SE +/- 0.00, N = 34.934.93MIN: 2.76 / MAX: 9.29MIN: 2.76 / MAX: 9.921. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
MinAvgMaxAMD Preferred Core Disabled10.4115.9126.6AMD Preferred Core Patched13.2117.1128.4OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.1CPU Power Consumption Monitor4080120160200

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.1Model: Weld Porosity Detection FP16 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched3691215SE +/- 0.01, N = 3SE +/- 0.01, N = 311.8611.86MIN: 6.03 / MAX: 19.5MIN: 6.58 / MAX: 18.441. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
MinAvgMaxAMD Preferred Core Disabled12.2119.6129.5AMD Preferred Core Patched11.0120.7130.7OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.1CPU Power Consumption Monitor4080120160200

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2023.1Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPUAMD Preferred Core DisabledAMD Preferred Core Patched0.06530.13060.19590.26120.3265SE +/- 0.00, N = 3SE +/- 0.00, N = 30.290.29MIN: 0.16 / MAX: 7.41MIN: 0.16 / MAX: 7.491. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
MinAvgMaxAMD Preferred Core Disabled12.6109.3119.0AMD Preferred Core Patched11.5109.3118.7OpenBenchmarking.orgWatts, Fewer Is BetterOpenVINO 2023.1CPU Power Consumption Monitor4080120160200

Darktable

MinAvgMaxAMD Preferred Core Disabled11.316.529.7AMD Preferred Core Patched12.716.732.2OpenBenchmarking.orgWatts, Fewer Is BetterDarktable 3.8.1CPU Power Consumption Monitor918273645

CPU Power Consumption Monitor

OpenBenchmarking.orgWattsCPU Power Consumption MonitorPhoronix Test Suite System MonitoringAMD Preferred Core DisabledAMD Preferred Core Patched306090120150Min: 6.45 / Avg: 48.65 / Max: 104.02Min: 5.88 / Avg: 62.27 / Max: 138.62

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance. Learn more via the OpenBenchmarking.org test page.

Java Test: Eclipse

AMD Preferred Core Patched: The test quit with a non-zero exit status.

AMD Preferred Core Disabled: The test quit with a non-zero exit status.

Renaissance

MinAvgMaxAMD Preferred Core Disabled10.733.174.7AMD Preferred Core Patched11.032.973.9OpenBenchmarking.orgWatts, Fewer Is BetterRenaissance 0.14CPU Power Consumption Monitor20406080100

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.14Test: Scala DottyAMD Preferred Core DisabledAMD Preferred Core Patched120240360480600SE +/- 11.82, N = 15SE +/- 8.15, N = 12538.0565.8MIN: 401.58 / MAX: 825.06MIN: 382.98 / MAX: 836.92
MinAvgMaxAMD Preferred Core Disabled10.733.174.7AMD Preferred Core Patched11.032.973.9OpenBenchmarking.orgWatts, Fewer Is BetterRenaissance 0.14CPU Power Consumption Monitor20406080100

AOM AV1

MinAvgMaxAMD Preferred Core Disabled12.140.961.4AMD Preferred Core Patched11.540.960.9OpenBenchmarking.orgWatts, Fewer Is BetterAOM AV1 3.7CPU Power Consumption Monitor20406080100

OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterAOM AV1 3.7Encoder Mode: Speed 6 Realtime - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched0.53961.07921.61882.15842.6982.3982.322

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.7Encoder Mode: Speed 6 Realtime - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched20406080100SE +/- 2.01, N = 15SE +/- 0.79, N = 1598.0995.021. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterAOM AV1 3.7Encoder Mode: Speed 6 Realtime - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched0.53961.07921.61882.15842.6982.3982.322
MinAvgMaxAMD Preferred Core Disabled12.140.961.4AMD Preferred Core Patched11.540.960.9OpenBenchmarking.orgWatts, Fewer Is BetterAOM AV1 3.7CPU Power Consumption Monitor20406080100

MinAvgMaxAMD Preferred Core Disabled10.730.759.7AMD Preferred Core Patched12.830.860.4OpenBenchmarking.orgWatts, Fewer Is BetterAOM AV1 3.7CPU Power Consumption Monitor20406080100

OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterAOM AV1 3.7Encoder Mode: Speed 6 Realtime - Input: Bosphorus 1080pAMD Preferred Core DisabledAMD Preferred Core Patched2468106.8567.136

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.7Encoder Mode: Speed 6 Realtime - Input: Bosphorus 1080pAMD Preferred Core DisabledAMD Preferred Core Patched50100150200250SE +/- 4.80, N = 15SE +/- 5.35, N = 15210.83219.521. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterAOM AV1 3.7Encoder Mode: Speed 6 Realtime - Input: Bosphorus 1080pAMD Preferred Core DisabledAMD Preferred Core Patched2468106.8567.136
MinAvgMaxAMD Preferred Core Disabled10.730.759.7AMD Preferred Core Patched12.830.860.4OpenBenchmarking.orgWatts, Fewer Is BetterAOM AV1 3.7CPU Power Consumption Monitor20406080100

MinAvgMaxAMD Preferred Core Disabled8.842.762.8AMD Preferred Core Patched10.542.762.6OpenBenchmarking.orgWatts, Fewer Is BetterAOM AV1 3.7CPU Power Consumption Monitor20406080100

OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterAOM AV1 3.7Encoder Mode: Speed 8 Realtime - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched0.48560.97121.45681.94242.4282.1582.145

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.7Encoder Mode: Speed 8 Realtime - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched20406080100SE +/- 1.58, N = 15SE +/- 0.84, N = 1592.1591.531. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterAOM AV1 3.7Encoder Mode: Speed 8 Realtime - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched0.48560.97121.45681.94242.4282.1582.145
MinAvgMaxAMD Preferred Core Disabled8.842.762.8AMD Preferred Core Patched10.542.762.6OpenBenchmarking.orgWatts, Fewer Is BetterAOM AV1 3.7CPU Power Consumption Monitor20406080100

MinAvgMaxAMD Preferred Core Disabled8.729.355.3AMD Preferred Core Patched11.429.353.7OpenBenchmarking.orgWatts, Fewer Is BetterAOM AV1 3.7CPU Power Consumption Monitor1530456075

OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterAOM AV1 3.7Encoder Mode: Speed 8 Realtime - Input: Bosphorus 1080pAMD Preferred Core DisabledAMD Preferred Core Patched2468107.7417.407

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.7Encoder Mode: Speed 8 Realtime - Input: Bosphorus 1080pAMD Preferred Core DisabledAMD Preferred Core Patched50100150200250SE +/- 5.51, N = 15SE +/- 2.39, N = 15226.48216.901. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterAOM AV1 3.7Encoder Mode: Speed 8 Realtime - Input: Bosphorus 1080pAMD Preferred Core DisabledAMD Preferred Core Patched2468107.7417.407
MinAvgMaxAMD Preferred Core Disabled8.729.355.3AMD Preferred Core Patched11.429.353.7OpenBenchmarking.orgWatts, Fewer Is BetterAOM AV1 3.7CPU Power Consumption Monitor1530456075

MinAvgMaxAMD Preferred Core Disabled11.038.258.0AMD Preferred Core Patched12.938.257.7OpenBenchmarking.orgWatts, Fewer Is BetterAOM AV1 3.7CPU Power Consumption Monitor1632486480

OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterAOM AV1 3.7Encoder Mode: Speed 9 Realtime - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched0.60321.20641.80962.41283.0162.6812.638

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.7Encoder Mode: Speed 9 Realtime - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched20406080100SE +/- 1.97, N = 15SE +/- 0.99, N = 6102.43100.661. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterAOM AV1 3.7Encoder Mode: Speed 9 Realtime - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched0.60321.20641.80962.41283.0162.6812.638
MinAvgMaxAMD Preferred Core Disabled11.038.258.0AMD Preferred Core Patched12.938.257.7OpenBenchmarking.orgWatts, Fewer Is BetterAOM AV1 3.7CPU Power Consumption Monitor1632486480

MinAvgMaxAMD Preferred Core Disabled10.127.852.5AMD Preferred Core Patched12.128.051.6OpenBenchmarking.orgWatts, Fewer Is BetterAOM AV1 3.7CPU Power Consumption Monitor1530456075

OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterAOM AV1 3.7Encoder Mode: Speed 9 Realtime - Input: Bosphorus 1080pAMD Preferred Core DisabledAMD Preferred Core Patched2468108.7318.104

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.7Encoder Mode: Speed 9 Realtime - Input: Bosphorus 1080pAMD Preferred Core DisabledAMD Preferred Core Patched50100150200250SE +/- 5.36, N = 15SE +/- 5.04, N = 15242.35226.641. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterAOM AV1 3.7Encoder Mode: Speed 9 Realtime - Input: Bosphorus 1080pAMD Preferred Core DisabledAMD Preferred Core Patched2468108.7318.104
MinAvgMaxAMD Preferred Core Disabled10.127.852.5AMD Preferred Core Patched12.128.051.6OpenBenchmarking.orgWatts, Fewer Is BetterAOM AV1 3.7CPU Power Consumption Monitor1530456075

MinAvgMaxAMD Preferred Core Disabled11.237.656.5AMD Preferred Core Patched10.937.957.1OpenBenchmarking.orgWatts, Fewer Is BetterAOM AV1 3.7CPU Power Consumption Monitor1632486480

OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterAOM AV1 3.7Encoder Mode: Speed 10 Realtime - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched0.6751.352.0252.73.3753.0002.705

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.7Encoder Mode: Speed 10 Realtime - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched306090120150SE +/- 0.95, N = 7SE +/- 1.80, N = 15112.90102.481. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterAOM AV1 3.7Encoder Mode: Speed 10 Realtime - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched0.6751.352.0252.73.3753.0002.705
MinAvgMaxAMD Preferred Core Disabled11.237.656.5AMD Preferred Core Patched10.937.957.1OpenBenchmarking.orgWatts, Fewer Is BetterAOM AV1 3.7CPU Power Consumption Monitor1632486480

MinAvgMaxAMD Preferred Core Disabled11.627.853.5AMD Preferred Core Patched10.627.852.8OpenBenchmarking.orgWatts, Fewer Is BetterAOM AV1 3.7CPU Power Consumption Monitor1530456075

OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterAOM AV1 3.7Encoder Mode: Speed 10 Realtime - Input: Bosphorus 1080pAMD Preferred Core DisabledAMD Preferred Core Patched2468108.5088.590

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.7Encoder Mode: Speed 10 Realtime - Input: Bosphorus 1080pAMD Preferred Core DisabledAMD Preferred Core Patched50100150200250SE +/- 7.11, N = 15SE +/- 3.89, N = 15236.30238.891. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterAOM AV1 3.7Encoder Mode: Speed 10 Realtime - Input: Bosphorus 1080pAMD Preferred Core DisabledAMD Preferred Core Patched2468108.5088.590
MinAvgMaxAMD Preferred Core Disabled11.627.853.5AMD Preferred Core Patched10.627.852.8OpenBenchmarking.orgWatts, Fewer Is BetterAOM AV1 3.7CPU Power Consumption Monitor1530456075

MinAvgMaxAMD Preferred Core Disabled12.827.452.6AMD Preferred Core Patched12.227.652.1OpenBenchmarking.orgWatts, Fewer Is BetterAOM AV1 3.7CPU Power Consumption Monitor1530456075

OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterAOM AV1 3.7Encoder Mode: Speed 11 Realtime - Input: Bosphorus 1080pAMD Preferred Core DisabledAMD Preferred Core Patched2468108.8208.016

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.7Encoder Mode: Speed 11 Realtime - Input: Bosphorus 1080pAMD Preferred Core DisabledAMD Preferred Core Patched50100150200250SE +/- 7.56, N = 15SE +/- 5.65, N = 15241.52221.271. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterAOM AV1 3.7Encoder Mode: Speed 11 Realtime - Input: Bosphorus 1080pAMD Preferred Core DisabledAMD Preferred Core Patched2468108.8208.016
MinAvgMaxAMD Preferred Core Disabled12.827.452.6AMD Preferred Core Patched12.227.652.1OpenBenchmarking.orgWatts, Fewer Is BetterAOM AV1 3.7CPU Power Consumption Monitor1530456075

Apache Hadoop

MinAvgMaxAMD Preferred Core Disabled10.628.854.8AMD Preferred Core Patched11.328.754.1OpenBenchmarking.orgWatts, Fewer Is BetterApache Hadoop 3.3.6CPU Power Consumption Monitor1530456075

OpenBenchmarking.orgOps per sec Per Watt, More Is BetterApache Hadoop 3.3.6Operation: Open - Threads: 50 - Files: 100000AMD Preferred Core DisabledAMD Preferred Core Patched6K12K18K24K30K25451.8326365.62

OpenBenchmarking.orgOps per sec, More Is BetterApache Hadoop 3.3.6Operation: Open - Threads: 50 - Files: 100000AMD Preferred Core DisabledAMD Preferred Core Patched160K320K480K640K800KSE +/- 11390.27, N = 15SE +/- 11501.08, N = 12732532757585
OpenBenchmarking.orgOps per sec Per Watt, More Is BetterApache Hadoop 3.3.6Operation: Open - Threads: 50 - Files: 100000AMD Preferred Core DisabledAMD Preferred Core Patched6K12K18K24K30K25451.8326365.62
MinAvgMaxAMD Preferred Core Disabled10.628.854.8AMD Preferred Core Patched11.328.754.1OpenBenchmarking.orgWatts, Fewer Is BetterApache Hadoop 3.3.6CPU Power Consumption Monitor1530456075

Apache HTTP Server

This is a test of the Apache HTTPD web server. This Apache HTTPD web server benchmark test profile makes use of the wrk program for facilitating the HTTP requests over a fixed period time with a configurable number of concurrent clients. Learn more via the OpenBenchmarking.org test page.

Concurrent Requests: 20

AMD Preferred Core Patched: The test quit with a non-zero exit status.

AMD Preferred Core Disabled: The test quit with a non-zero exit status.

nginx

This is a benchmark of the lightweight Nginx HTTP(S) web-server. This Nginx web server benchmark test profile makes use of the wrk program for facilitating the HTTP requests over a fixed period time with a configurable number of concurrent clients/connections. HTTPS with a self-signed OpenSSL certificate is used by this test for local benchmarking. Learn more via the OpenBenchmarking.org test page.

Connections: 20

AMD Preferred Core Patched: The test quit with a non-zero exit status.

AMD Preferred Core Disabled: The test quit with a non-zero exit status.

SVT-AV1

MinAvgMaxAMD Preferred Core Disabled8.143.491.9AMD Preferred Core Patched11.043.492.7OpenBenchmarking.orgWatts, Fewer Is BetterSVT-AV1 1.7CPU Power Consumption Monitor20406080100

OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterSVT-AV1 1.7Encoder Mode: Preset 13 - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched0.85611.71222.56833.42444.28053.8053.788

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 1.7Encoder Mode: Preset 13 - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched4080120160200SE +/- 4.45, N = 15SE +/- 4.41, N = 15165.00164.441. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenBenchmarking.orgFrames Per Second Per Watt, More Is BetterSVT-AV1 1.7Encoder Mode: Preset 13 - Input: Bosphorus 4KAMD Preferred Core DisabledAMD Preferred Core Patched0.85611.71222.56833.42444.28053.8053.788
MinAvgMaxAMD Preferred Core Disabled8.143.491.9AMD Preferred Core Patched11.043.492.7OpenBenchmarking.orgWatts, Fewer Is BetterSVT-AV1 1.7CPU Power Consumption Monitor20406080100

SQLite

This is a simple benchmark of SQLite. At present this test profile just measures the time to perform a pre-defined number of insertions on an indexed database with a variable number of concurrent repetitions -- up to the maximum number of CPU threads available. Learn more via the OpenBenchmarking.org test page.

Threads / Copies: 1

AMD Preferred Core Patched: The test run did not produce a result.

AMD Preferred Core Disabled: The test run did not produce a result.

Darktable

Darktable is an open-source photography / workflow application this will use any system-installed Darktable program or on Windows will automatically download the pre-built binary from the project. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 3.8.1Test: Server Rack - Acceleration: CPU-onlyAMD Preferred Core DisabledAMD Preferred Core Patched0.02610.05220.07830.10440.1305SE +/- 0.000, N = 14SE +/- 0.002, N = 150.1130.116
MinAvgMaxAMD Preferred Core Disabled11.316.529.7AMD Preferred Core Patched12.716.732.2OpenBenchmarking.orgWatts, Fewer Is BetterDarktable 3.8.1CPU Power Consumption Monitor918273645

OpenFOAM

OpenFOAM is the leading free, open-source software for computational fluid dynamics (CFD). This test profile currently uses the drivaerFastback test case for analyzing automotive aerodynamics or alternatively the older motorBike input. Learn more via the OpenBenchmarking.org test page.

Input: drivaerFastback, Medium Mesh Size

AMD Preferred Core Patched: The test quit with a non-zero exit status. E: [0] --> FOAM FATAL ERROR:

AMD Preferred Core Disabled: The test quit with a non-zero exit status. E: [0] --> FOAM FATAL ERROR:

Input: drivaerFastback, Small Mesh Size

AMD Preferred Core Patched: The test quit with a non-zero exit status. E: [0] --> FOAM FATAL ERROR:

AMD Preferred Core Disabled: The test quit with a non-zero exit status. E: [0] --> FOAM FATAL ERROR:

310 Results Shown

Meta Performance Per Watts
Stargate Digital Audio Workstation:
  192000 - 1024
  192000 - 512
AOM AV1
TiDB Community Server:
  oltp_update_non_index - 32
  oltp_update_index - 16
  oltp_update_non_index - 16
  oltp_read_write - 16
RocksDB
PyPerformance
JPEG XL Decoding libjxl
Apache Hadoop
Renaissance
AOM AV1
PyPerformance
DaCapo Benchmark
SQLite
x265
Renaissance
SQLite
PyBench
Apache Hadoop
GIMP
PyPerformance
DaCapo Benchmark
VVenC
Renaissance
OpenVINO
PHPBench
libavif avifenc
DaCapo Benchmark
Kvazaar
SQLite
uvg266
JPEG XL libjxl
Darktable
uvg266
PyPerformance:
  nbody
  pathlib
OpenVINO
RocksDB
PyPerformance
Apache HTTP Server
Liquid-DSP
Apache IoTDB
Kvazaar
Liquid-DSP
Darktable
PyPerformance
Apache CouchDB
nginx
libavif avifenc
AOM AV1
PyPerformance
Liquid-DSP
PyPerformance
Renaissance:
  Genetic Algorithm Using Jenetics + Futures
  Savina Reactors.IO
Apache Hadoop
OpenVINO
Apache CouchDB
PyPerformance
Apache CouchDB
VVenC
Apache Hadoop
Timed FFmpeg Compilation
OpenVINO
Kvazaar
JPEG XL Decoding libjxl
libavif avifenc
John The Ripper
Apache CouchDB
JPEG XL libjxl
GIMP
SQLite
Gcrypt Library
Renaissance
OpenRadioss
JPEG XL libjxl
Apache Hadoop
VVenC
AOM AV1
PyPerformance
RocksDB
TiDB Community Server
Apache IoTDB
libavif avifenc:
  0
  2
SQLite Speedtest
SVT-AV1
Renaissance
Redis
GIMP
PyPerformance
Neural Magic DeepSparse:
  NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Synchronous Single-Stream:
    ms/batch
    items/sec
JPEG XL libjxl
John The Ripper
OpenVINO
TiDB Community Server
VVenC
Apache HTTP Server
OpenVINO
OpenRadioss
Neural Magic DeepSparse
Kvazaar
Neural Magic DeepSparse
OpenVINO
TensorFlow
Apache HTTP Server
RocksDB
Neural Magic DeepSparse:
  NLP Text Classification, BERT base uncased SST2 - Synchronous Single-Stream:
    ms/batch
    items/sec
nginx
Kvazaar
OpenVINO
Neural Magic DeepSparse:
  NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Synchronous Single-Stream:
    items/sec
    ms/batch
AOM AV1
Apache CouchDB
SQLite
GIMP
OpenVINO:
  Face Detection Retail FP16-INT8 - CPU
  Age Gender Recognition Retail 0013 FP16-INT8 - CPU
Neural Magic DeepSparse:
  ResNet-50, Sparse INT8 - Synchronous Single-Stream:
    ms/batch
    items/sec
FFmpeg
uvg266:
  Bosphorus 1080p - Very Fast
  Bosphorus 1080p - Super Fast
FFmpeg
nginx
Darktable
OpenVINO
RocksDB
OpenVINO
Kvazaar
Apache HTTP Server
Kvazaar
OpenVINO
Kvazaar
Neural Magic DeepSparse
TensorFlow
Neural Magic DeepSparse
uvg266
LAME MP3 Encoding
OpenVINO
uvg266
Neural Magic DeepSparse:
  NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Asynchronous Multi-Stream:
    items/sec
    ms/batch
OpenRadioss
Renaissance
DaCapo Benchmark
uvg266
Kvazaar
OpenSSL
OpenVINO
OpenRadioss
Memcached
OpenVINO
RocksDB
Neural Magic DeepSparse
TiDB Community Server
OpenVINO:
  Road Segmentation ADAS FP16 - CPU
  Road Segmentation ADAS FP16-INT8 - CPU
Neural Magic DeepSparse
OpenVINO:
  Road Segmentation ADAS FP16-INT8 - CPU
  Face Detection FP16 - CPU
Neural Magic DeepSparse
Liquid-DSP
OpenSSL
Kvazaar
Renaissance
Neural Magic DeepSparse
OpenVINO:
  Handwritten English Recognition FP16 - CPU
  Vehicle Detection FP16-INT8 - CPU
Blender
FFmpeg:
  libx265 - Video On Demand:
    FPS
    Seconds
uvg266
OpenSSL
SVT-AV1
OpenVINO
OpenSSL
John The Ripper
x265
OpenVINO
Neural Magic DeepSparse:
  BERT-Large, NLP Question Answering - Synchronous Single-Stream:
    items/sec
    ms/batch
  BERT-Large, NLP Question Answering, Sparse INT8 - Synchronous Single-Stream:
    items/sec
    ms/batch
Blender
Apache Hadoop
OpenSSL
Liquid-DSP
Renaissance
Timed Linux Kernel Compilation
Neural Magic DeepSparse
OpenVINO
Apache Hadoop
Neural Magic DeepSparse
PyPerformance
Neural Magic DeepSparse:
  NLP Token Classification, BERT base uncased conll2003 - Synchronous Single-Stream:
    ms/batch
    items/sec
OpenRadioss
Neural Magic DeepSparse:
  CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Stream
  NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Stream
Apache Hadoop
OpenVINO
Neural Magic DeepSparse:
  CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Stream
  NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Stream
  ResNet-50, Baseline - Synchronous Single-Stream
  ResNet-50, Baseline - Synchronous Single-Stream
SVT-AV1
Unpacking Firefox
John The Ripper
Neural Magic DeepSparse
uvg266
Neural Magic DeepSparse
OpenVINO
nginx
uvg266
Neural Magic DeepSparse:
  NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Synchronous Single-Stream:
    ms/batch
    items/sec
  NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Stream:
    ms/batch
  CV Detection, YOLOv5s COCO - Synchronous Single-Stream:
    items/sec
    ms/batch
  ResNet-50, Sparse INT8 - Asynchronous Multi-Stream:
    ms/batch
  BERT-Large, NLP Question Answering - Asynchronous Multi-Stream:
    items/sec
Apache Hadoop
Neural Magic DeepSparse
Blender
Neural Magic DeepSparse:
  BERT-Large, NLP Question Answering - Asynchronous Multi-Stream
  ResNet-50, Baseline - Asynchronous Multi-Stream
  ResNet-50, Baseline - Asynchronous Multi-Stream
  NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Stream
  NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Stream
OpenRadioss
Neural Magic DeepSparse
OpenVINO
Neural Magic DeepSparse
OpenVINO
Neural Magic DeepSparse
OpenSSL:
  AES-128-GCM
  RSA4096
John The Ripper
Neural Magic DeepSparse
Memcached
OpenVINO
Neural Magic DeepSparse:
  NLP Text Classification, BERT base uncased SST2 - Asynchronous Multi-Stream
  NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Stream
OpenSSL
FFmpeg
Blender
Neural Magic DeepSparse
FFmpeg
Liquid-DSP
Neural Magic DeepSparse:
  CV Segmentation, 90% Pruned YOLACT Pruned - Synchronous Single-Stream:
    items/sec
    ms/batch
OpenVINO
Neural Magic DeepSparse
OpenVINO
Neural Magic DeepSparse:
  NLP Document Classification, oBERT base uncased on IMDB - Synchronous Single-Stream:
    items/sec
    ms/batch
  NLP Text Classification, BERT base uncased SST2 - Asynchronous Multi-Stream:
    items/sec
FFmpeg:
  libx265 - Upload:
    Seconds
    FPS
AOM AV1:
  Speed 0 Two-Pass - Bosphorus 4K
  Speed 6 Two-Pass - Bosphorus 4K
OpenVINO:
  Face Detection Retail FP16 - CPU
  Vehicle Detection FP16-INT8 - CPU
  Weld Porosity Detection FP16 - CPU
  Age Gender Recognition Retail 0013 FP16-INT8 - CPU
Darktable:
  CPU Power Consumption Monitor
  Phoronix Test Suite System Monitoring
  CPU Power Consumption Monitor
Renaissance
AOM AV1:
  CPU Power Consumption Monitor
  Speed 6 Realtime - Bosphorus 4K
AOM AV1
AOM AV1:
  CPU Power Consumption Monitor
  Speed 6 Realtime - Bosphorus 1080p
AOM AV1
AOM AV1:
  CPU Power Consumption Monitor
  Speed 8 Realtime - Bosphorus 4K
AOM AV1
AOM AV1:
  CPU Power Consumption Monitor
  Speed 8 Realtime - Bosphorus 1080p
AOM AV1
AOM AV1:
  CPU Power Consumption Monitor
  Speed 9 Realtime - Bosphorus 4K
AOM AV1
AOM AV1:
  CPU Power Consumption Monitor
  Speed 9 Realtime - Bosphorus 1080p
AOM AV1
AOM AV1:
  CPU Power Consumption Monitor
  Speed 10 Realtime - Bosphorus 4K
AOM AV1
AOM AV1:
  CPU Power Consumption Monitor
  Speed 10 Realtime - Bosphorus 1080p
AOM AV1
AOM AV1:
  CPU Power Consumption Monitor
  Speed 11 Realtime - Bosphorus 1080p
AOM AV1
Apache Hadoop:
  CPU Power Consumption Monitor
  Open - 50 - 100000
Apache Hadoop
SVT-AV1:
  CPU Power Consumption Monitor
  Preset 13 - Bosphorus 4K
SVT-AV1
Darktable

AMD Preferred Core Patched

Kernel Notes: Transparent Huge Pages: madvise
Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-XeT9lY/gcc-11-11.4.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-XeT9lY/gcc-11-11.4.0/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-build-config=bootstrap-lto-lean --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
Disk Notes: NONE / errors=remount-ro,relatime,rw,stripe=64 / Block Size: 4096
Processor Notes: Scaling Governor: amd-pstate-epp powersave (EPP: balance_performance) - CPU Microcode: 0xa601203
Java Notes: OpenJDK Runtime Environment (build 11.0.20+8-post-Ubuntu-1ubuntu122.04)
Python Notes: Python 3.10.12
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 + retbleed: Not affected + spec_rstack_overflow: Mitigation of safe RET no microcode + 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 + srbds: Not affected + tsx_async_abort: Not affected

Testing initiated at 7 October 2023 17:39 by user root.

AMD Preferred Core Disabled

Processor: AMD Ryzen 9 7950X3D 16-Core @ 5.76GHz (16 Cores / 32 Threads), Motherboard: ASRockRack B650D4U-2L2T/BCM (2.09 BIOS), Chipset: AMD Device 14d8, Memory: 2 x 32 GB DDR5-4800MT/s MTC20C2085S1EC48BA1, Disk: 3201GB Micron_7450_MTFDKCC3T2TFS + 0GB Virtual HDisk0 + 0GB Virtual HDisk1 + 0GB Virtual HDisk2 + 0GB Virtual HDisk3, Graphics: ASPEED 512MB, Audio: AMD Device 1640, Monitor: VA2431, Network: 2 x Intel I210 + 2 x Broadcom BCM57416 NetXtreme-E Dual-Media 10G RDMA

OS: Ubuntu 22.04, Kernel: 6.6.0-rc4-phx-amd-pref-core (x86_64), Desktop: GNOME Shell 42.9, Display Server: X Server, Vulkan: 1.3.238, Compiler: GCC 11.4.0, File-System: ext4, Screen Resolution: 1920x1200

Kernel Notes: Transparent Huge Pages: madvise
Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-XeT9lY/gcc-11-11.4.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-XeT9lY/gcc-11-11.4.0/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-build-config=bootstrap-lto-lean --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
Disk Notes: NONE / errors=remount-ro,relatime,rw,stripe=64 / Block Size: 4096
Processor Notes: Scaling Governor: amd-pstate-epp powersave (EPP: balance_performance) - CPU Microcode: 0xa601203
Java Notes: OpenJDK Runtime Environment (build 11.0.20+8-post-Ubuntu-1ubuntu122.04)
Python Notes: Python 3.10.12
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 + retbleed: Not affected + spec_rstack_overflow: Mitigation of safe RET no microcode + 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 + srbds: Not affected + tsx_async_abort: Not affected

Testing initiated at 8 October 2023 11:28 by user root.