Noctua NH-L12Sx77 Heatsink

Benchmarks 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 2404293-NE-NOCTUAHEA44
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

Limit displaying results to tests within:

AV1 3 Tests
C++ Boost Tests 2 Tests
Timed Code Compilation 6 Tests
C/C++ Compiler Tests 12 Tests
Compression Tests 3 Tests
CPU Massive 20 Tests
Creator Workloads 25 Tests
Cryptography 2 Tests
Encoding 8 Tests
Fortran Tests 2 Tests
Game Development 5 Tests
HPC - High Performance Computing 8 Tests
Imaging 5 Tests
Common Kernel Benchmarks 3 Tests
Machine Learning 5 Tests
Molecular Dynamics 2 Tests
MPI Benchmarks 2 Tests
Multi-Core 30 Tests
NVIDIA GPU Compute 3 Tests
Intel oneAPI 5 Tests
OpenMPI Tests 3 Tests
Programmer / Developer System Benchmarks 6 Tests
Python Tests 6 Tests
Raytracing 3 Tests
Renderers 5 Tests
Scientific Computing 2 Tests
Software Defined Radio 2 Tests
Server 3 Tests
Server CPU Tests 12 Tests
Texture Compression 2 Tests
Video Encoding 8 Tests
Common Workstation Benchmarks 3 Tests

Statistics

Show Overall Harmonic Mean(s)
Show Overall Geometric Mean
Show Geometric Means Per-Suite/Category
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
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
Hide
Result
Result
Identifier
Performance Per
Dollar
Date
Run
  Test
  Duration
Noctua NH-L9a-AM5
April 28
  19 Hours, 34 Minutes
Noctua NH-L12Sx77
April 29
  18 Hours, 55 Minutes
Invert Hiding All Results Option
  19 Hours, 14 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):


Noctua NH-L12Sx77 HeatsinkOpenBenchmarking.orgPhoronix Test SuiteAMD Ryzen 5 7600 6-Core @ 5.17GHz (6 Cores / 12 Threads)Gigabyte B650M DS3H (F24c BIOS)AMD Device 14d82 x 16GB DDR5-5600MT/s1000GB Sabrent Rocket 4.0 PlusGigabyte AMD Raphael 512MBAMD Rembrandt Radeon HD AudioASUS VP28URealtek RTL8125 2.5GbEUbuntu 24.046.8.0-31-generic (x86_64)GNOME Shell 46.0X Server + Wayland4.6 Mesa 24.0.5-1ubuntu1 (LLVM 17.0.6 DRM 3.57)GCC 13.2.0ext43840x2160ProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerOpenGLCompilerFile-SystemScreen ResolutionNoctua NH-L12Sx77 Heatsink BenchmarksSystem Logs- Transparent Huge Pages: madvise- --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-uJ7kn6/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-uJ7kn6/gcc-13-13.2.0/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-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 - Scaling Governor: amd-pstate-epp powersave (EPP: balance_performance) - CPU Microcode: 0xa601206 - Python 3.12.3- gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Mitigation of Safe RET + 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; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected

Noctua NH-L9a-AM5 vs. Noctua NH-L12Sx77 ComparisonPhoronix Test SuiteBaseline+2.3%+2.3%+4.6%+4.6%+6.9%+6.9%9.1%7.5%7.2%6.9%6.8%6.7%5.7%5.6%5.5%5.5%5.4%5.4%5.3%5.2%5.2%5.2%5.2%5.2%5.1%5.1%5.1%5.1%5.1%5%5%5%5%5%4.9%4.9%4.9%4.9%4.9%4.9%4.9%4.9%4.8%4.8%4.8%4.8%4.8%4.8%4.8%4.7%4.7%4.7%4.7%4.6%4.6%4.6%4.6%4.6%4.5%4.5%4.5%4.5%4.5%4.5%4.4%4.4%4.4%4.4%4.4%4.4%4.4%4.4%4.4%4.4%4.4%4.4%4.3%4.3%4.3%4.3%4.3%4.3%4.2%4.2%4.2%4.2%4.2%4.2%4.1%4.1%4.1%4%3.9%3.9%3.9%3.9%3.8%3.8%3.8%3.8%3.8%3.7%3.7%3.7%3.7%3.7%3.7%3.6%3.6%3.6%3.6%3.6%3.5%3.5%3.5%3.5%3.4%3.4%3.4%3.4%3.4%3.4%3.4%3.4%3.3%3.3%3.3%3.2%3.2%3.2%3.1%3.1%3.1%3.1%3.1%3.1%3.1%3%2.9%2.9%2.9%2.8%2.8%2.8%2.8%2.7%2.7%2.7%2.7%2.7%2.7%2.7%2.5%2.5%2.5%2.4%2.3%2.3%2.3%2.2%2.2%2.2%2.1%2.1%2.1%2%2%2%2%Update RandH.HD.RTomographic ModelOverwriteUpdate RandSpeed 6 Realtime - Bosphorus 4KEnhancedBosphorus 4KCTS2Bosphorus 4K - Medium2 - 1080p - 16 - Path Tracer - CPUR.N.N.T - CPU5.3%1e13R.R.W.RV.P.MallmodconfigBMW27 - CPU-Only3 - 1080p - 1 - Path Tracer - CPUBarbershop - CPU-Only2 - 4K - 32 - Path Tracer - CPU1 - 1080p - 32 - Path Tracer - CPU3 - 1080p - 32 - Path Tracer - CPU3 - 4K - 1 - Path Tracer - CPUTime To CompileCORAL2 P2Bosphorus 4K - Super Fast3 - 4K - 16 - Path Tracer - CPUBosphorus 4K - Very Fast1 - 4K - 1 - Path Tracer - CPUClassroom - CPU-Only2 - 1080p - 32 - Path Tracer - CPU3 - 1080p - 16 - Path Tracer - CPUSwirlRand Read8 - 256 - 512Pathtracer ISPC - CrownSpeed 0 Two-Pass - Bosphorus 1080p1 - 4K - 32 - Path Tracer - CPUFishy Cat - CPU-OnlyTime To CompilePabellon Barcelona - CPU-OnlyExhaustive1 - 1080p - 1 - Path Tracer - CPUgravity_spheres_volume/dim_512/pathtracer/real_time2 - 4K - 16 - Path Tracer - CPUJunkshop - CPU-OnlyBosphorus 4K - SlowResizing1 - 1080p - 16 - Path Tracer - CPU8 - 256 - 57W.l.HBosphorus 1080p - FasterSharpenPathtracer - CrownSpeed 6 Two-Pass - Bosphorus 1080p2 - 1080p - 1 - Path Tracer - CPUBosphorus 4K - Ultra Fastparticle_volume/scivis/real_time3 - 4K - 32 - Path Tracer - CPUPreset 13 - Bosphorus 4KVery ThoroughBosphorus 4K - Slow2 - 4K - 1 - Path Tracer - CPU1 - 4K - 16 - Path Tracer - CPUR.N.N.I - CPU4.4%SHA2564K - 16CORAL2 P1particle_volume/ao/real_time1e12Speed 4 Two-Pass - Bosphorus 4KRSA4096Bosphorus 4K - MediumSeq FillTime To CompileSHA5122defconfigBosphorus 1080p - SlowTime To CompileBosphorus 4K - Fastgravity_spheres_volume/dim_512/scivis/real_timeBosphorus 1080p - MediumSpeed 9 Realtime - Bosphorus 1080p4.1%6, LosslessBosphorus 4K - Very FastBosphorus 4K - Super FastR.R.W.R6ThoroughTime To CompileBosphorus 1080p - FastBosphorus 1080p - Slow12 - 256 - 512Speed 6 Two-Pass - Bosphorus 4KLayered HalfspaceChess BenchmarkPNG - 100CPUPreset 12 - Bosphorus 1080pSpeed 4 Two-Pass - Bosphorus 1080pPathtracer - Asian Dragon ObjPreset 4 - Bosphorus 4KJPEG - 100CPU - 64 - AlexNet3.6%gravity_spheres_volume/dim_512/ao/real_timeBosphorus 4K - FasterFastSpeed 8 Realtime - Bosphorus 4KBosphorus 1080p - MediumBosphorus 1080p - Very FastBosphorus 4K - Ultra FastAll1080p - 16Noise-GaussianBosphorus 1080p - Super FastSpeed 10 Realtime - Bosphorus 4KPathtracer ISPC - Asian Dragon12 - 256 - 32F.1.0.R.a.m.i.CRead While WritingMount St. Helenslibx265 - Video On DemandBosphorus 1080p - Ultra FastPreset 12 - Bosphorus 4Klibx265 - Platform1Rand FillPathtracer ISPC - Asian Dragon Obj8 - 256 - 32AES-128-GCM10, LosslessRSA4096Pathtracer - Asian DragonBosphorus 1080p - Very FastPNG - 80libx265 - UploadCompression RatingMediumAES-256-GCMHWB Color SpaceRead While Writing12 - 256 - 57S.w.1.0.6.AMulti-ThreadedPreset 8 - Bosphorus 4KBosphorus 1080pRand ReadJPEG - 90JPEG - 80Church FacadePreset 8 - Bosphorus 1080pCPU - 1 - ResNet-50Speed 8 Realtime - Bosphorus 1080pBosphorus 1080p - Super Fast1 - 256 - 512Speed 10 Realtime - Bosphorus 1080plibx264 - Livellava-v1.5-7b-q4 - CPUBosphorus 1080p - Ultra FastPNG - 90Preset 4 - Bosphorus 1080pChaCha20-Poly1305AVX-512 VNNICPU - 64 - ResNet-50RocksDBSPECFEM3D7-Zip CompressionSPECFEM3DRocksDBSpeedbAOM AV1GraphicsMagickx265Quicksilveruvg266OSPRay StudiooneDNNPrimesieveRocksDBBRL-CADTimed Linux Kernel CompilationBlenderOSPRay StudioBlenderOSPRay StudioOSPRay StudioOSPRay StudioOSPRay StudioTimed Node.js CompilationQuicksilveruvg266OSPRay Studiouvg266OSPRay StudioBlenderOSPRay StudioOSPRay StudioGraphicsMagickRocksDBLiquid-DSPEmbreeAOM AV1OSPRay StudioBlenderTimed Godot Game Engine CompilationBlenderASTC EncoderOSPRay StudioOSPRayOSPRay StudioBlenderuvg266GraphicsMagickOSPRay StudioLiquid-DSPSPECFEM3DVVenCGraphicsMagickEmbreelibavif avifencAOM AV1OSPRay Studiouvg266OSPRayOSPRay StudioSVT-AV1ASTC EncoderKvazaarOSPRay StudioOSPRay StudiooneDNNOpenSSLC-RayQuicksilverOSPRayPrimesieveAOM AV1OpenSSLKvazaarSpeedbTimed Mesa CompilationOpenSSLlibavif avifencTimed Linux Kernel Compilationuvg266Timed FFmpeg CompilationVVenCOSPRayuvg266AOM AV1libavif avifencKvazaarKvazaarSpeedblibavif avifencASTC EncoderTimed PHP CompilationVVenCKvazaarLiquid-DSPAOM AV1SPECFEM3DStockfishJPEG-XL libjxlChaos Group V-RAYSVT-AV1AOM AV1EmbreeSVT-AV1JPEG-XL libjxlTensorFlowOSPRayVVenCASTC EncoderAOM AV1Kvazaaruvg266KvazaarJPEG-XL Decoding libjxlC-RayGraphicsMagickuvg266AOM AV1EmbreeLiquid-DSPParallel BZIP2 CompressionSpeedbSPECFEM3DFFmpeguvg266SVT-AV1FFmpegJPEG-XL Decoding libjxlSpeedbEmbreeLiquid-DSPOpenSSLlibavif avifencOpenSSLEmbreeKvazaarJPEG-XL libjxlFFmpeg7-Zip CompressionASTC EncoderOpenSSLGraphicsMagickRocksDBLiquid-DSPNAMDQuantLibSVT-AV1x265SpeedbJPEG-XL libjxlJPEG-XL libjxlGoogle DracoSVT-AV1PyTorchAOM AV1KvazaarLiquid-DSPAOM AV1FFmpegLlamafileKvazaarJPEG-XL libjxlSVT-AV1OpenSSLStress-NGTensorFlowNoctua NH-L9a-AM5Noctua NH-L12Sx77

Noctua NH-L12Sx77 Heatsinkbuild-linux-kernel: allmodconfigblender: Barbershop - CPU-Onlybuild-nodejs: Time To Compilexmrig: GhostRider - 1Mospray-studio: 3 - 4K - 32 - Path Tracer - CPUquicksilver: CTS2blender: Pabellon Barcelona - CPU-Onlybuild-godot: Time To Compileospray-studio: 2 - 4K - 32 - Path Tracer - CPUospray-studio: 1 - 4K - 32 - Path Tracer - CPUblender: Classroom - CPU-Onlyopenssl: RSA4096openssl: RSA4096c-ray: 4K - 16tensorflow: CPU - 64 - ResNet-50astcenc: Very Thoroughastcenc: Exhaustiveospray-studio: 3 - 4K - 16 - Path Tracer - CPUbrl-cad: VGR Performance Metricprimesieve: 1e13llamafile: wizardcoder-python-34b-v1.0.Q6_K - CPUspecfem3d: Layered Halfspaceospray-studio: 2 - 4K - 16 - Path Tracer - CPUffmpeg: libx264 - Uploadospray-studio: 1 - 4K - 16 - Path Tracer - CPUquicksilver: CORAL2 P2aom-av1: Speed 0 Two-Pass - Bosphorus 4Kblender: Junkshop - CPU-Onlyspeedb: Rand Fill Syncopenssl: AES-256-GCMopenssl: ChaCha20-Poly1305openssl: ChaCha20openssl: AES-128-GCMopenssl: SHA512openssl: SHA256ospray: particle_volume/scivis/real_timeblender: Fishy Cat - CPU-Onlyspecfem3d: Water-layered Halfspaceospray: particle_volume/pathtracer/real_timegromacs: MPI CPU - water_GMX50_bareffmpeg: libx265 - Platformffmpeg: libx265 - Video On Demandtensorflow: CPU - 32 - ResNet-50blender: BMW27 - CPU-Onlyffmpeg: libx265 - Uploadavifenc: 0ospray: particle_volume/ao/real_timevvenc: Bosphorus 4K - Fastffmpeg: libx264 - Platformffmpeg: libx264 - Video On Demandbuild-linux-kernel: defconfignamd: STMV with 1,066,628 Atomscachebench: Read / Modify / Writecachebench: Writecachebench: Readuvg266: Bosphorus 4K - Slowllamafile: mistral-7b-instruct-v0.2.Q8_0 - CPUospray-studio: 3 - 1080p - 32 - Path Tracer - CPUuvg266: Bosphorus 4K - Mediumhpcg: 104 104 104 - 60ospray-studio: 2 - 1080p - 32 - Path Tracer - CPUospray-studio: 1 - 1080p - 32 - Path Tracer - CPUaom-av1: Speed 4 Two-Pass - Bosphorus 4Ktensorflow: CPU - 64 - GoogLeNetospray-studio: 3 - 4K - 1 - Path Tracer - CPUquicksilver: CORAL2 P1stockfish: Chess Benchmarkquantlib: Single-Threadedoidn: RT.ldr_alb_nrm.3840x2160 - CPU-Onlyonednn: Recurrent Neural Network Training - CPUospray: gravity_spheres_volume/dim_512/pathtracer/real_timespeedb: Read While Writingspecfem3d: Homogeneous Halfspaceospray-studio: 3 - 1080p - 1 - Path Tracer - CPUonednn: Recurrent Neural Network Inference - CPUospray-studio: 2 - 1080p - 1 - Path Tracer - CPUliquid-dsp: 1 - 256 - 512liquid-dsp: 4 - 256 - 57ospray-studio: 1 - 1080p - 1 - Path Tracer - CPUc-ray: 1080p - 16v-ray: CPUembree: Pathtracer - Asian Dragon Objliquid-dsp: 1 - 256 - 57embree: Pathtracer ISPC - Asian Dragon Objastcenc: Thoroughospray-studio: 3 - 1080p - 16 - Path Tracer - CPUkvazaar: Bosphorus 4K - Slowavifenc: 2build-php: Time To Compilespecfem3d: Mount St. Helensspecfem3d: Tomographic Modelbuild-ffmpeg: Time To Compilekvazaar: Bosphorus 4K - Mediumembree: Pathtracer - Crowngraphics-magick: Sharpengraphics-magick: Enhancedjpegxl: PNG - 80embree: Pathtracer ISPC - Crowngraphics-magick: Noise-Gaussianjpegxl: JPEG - 80speedb: Rand Fillgraphics-magick: Swirlgraphics-magick: Rotatespeedb: Update Randgraphics-magick: Resizingrocksdb: Overwriterocksdb: Read While Writingspeedb: Read Rand Write Randgraphics-magick: HWB Color Spacespeedb: Rand Readrocksdb: Update Randrocksdb: Read Rand Write Randrocksdb: Rand Readospray-studio: 2 - 1080p - 16 - Path Tracer - CPUospray-studio: 2 - 4K - 1 - Path Tracer - CPUembree: Pathtracer - Asian Dragonospray-studio: 1 - 1080p - 16 - Path Tracer - CPUvvenc: Bosphorus 4K - Fasterospray-studio: 1 - 4K - 1 - Path Tracer - CPUembree: Pathtracer ISPC - Asian Dragonospray: gravity_spheres_volume/dim_512/ao/real_timeospray: gravity_spheres_volume/dim_512/scivis/real_timeliquid-dsp: 8 - 256 - 57tensorflow: CPU - 32 - GoogLeNettensorflow: CPU - 64 - AlexNetffmpeg: libx265 - Liveuvg266: Bosphorus 4K - Very Fastuvg266: Bosphorus 4K - Super Fastnamd: ATPase with 327,506 Atomsaom-av1: Speed 6 Realtime - Bosphorus 4Kaom-av1: Speed 6 Two-Pass - Bosphorus 4Kjpegxl: PNG - 90jpegxl: JPEG - 90vvenc: Bosphorus 1080p - Fastjpegxl: JPEG - 100jpegxl: PNG - 100uvg266: Bosphorus 4K - Ultra Fastquantlib: Multi-Threadedbuild-mesa: Time To Compilesvt-av1: Preset 4 - Bosphorus 4Ksvt-av1: Preset 13 - Bosphorus 4Kx265: Bosphorus 4Kaom-av1: Speed 4 Two-Pass - Bosphorus 1080py-cruncher: 1Bjpegxl-decode: 1aom-av1: Speed 0 Two-Pass - Bosphorus 1080pllamafile: llava-v1.5-7b-q4 - CPUffmpeg: libx264 - Liveaom-av1: Speed 8 Realtime - Bosphorus 4Kastcenc: Fastwebp: Quality 100, Lossless, Highest Compressionliquid-dsp: 12 - 256 - 512liquid-dsp: 2 - 256 - 512liquid-dsp: 2 - 256 - 57liquid-dsp: 8 - 256 - 512liquid-dsp: 4 - 256 - 512liquid-dsp: 1 - 256 - 32liquid-dsp: 2 - 256 - 32stress-ng: AVX-512 VNNIliquid-dsp: 12 - 256 - 57liquid-dsp: 12 - 256 - 32liquid-dsp: 4 - 256 - 32liquid-dsp: 8 - 256 - 32tensorflow: CPU - 32 - AlexNetuvg266: Bosphorus 1080p - Slowaom-av1: Speed 9 Realtime - Bosphorus 4Kaom-av1: Speed 10 Realtime - Bosphorus 4Kkvazaar: Bosphorus 4K - Very Fastcompress-lz4: 9 - Decompression Speedcompress-lz4: 9 - Compression Speeduvg266: Bosphorus 1080p - Mediumpytorch: CPU - 1 - ResNet-50compress-lz4: 3 - Decompression Speedcompress-lz4: 3 - Compression Speedcompress-lz4: 1 - Decompression Speedcompress-lz4: 1 - Compression Speedjpegxl-decode: Allonednn: Deconvolution Batch shapes_1d - CPUy-cruncher: 500Mastcenc: Mediumkvazaar: Bosphorus 4K - Super Fastsrsran: PUSCH Processor Benchmark, Throughput Threadonednn: IP Shapes 1D - CPUkvazaar: Bosphorus 4K - Ultra Fastprimesieve: 1e12compress-7zip: Decompression Ratingcompress-7zip: Compression Ratingvvenc: Bosphorus 1080p - Fasterkvazaar: Bosphorus 1080p - Slowsvt-av1: Preset 8 - Bosphorus 4Kkvazaar: Bosphorus 1080p - Mediumaom-av1: Speed 6 Two-Pass - Bosphorus 1080psvt-av1: Preset 4 - Bosphorus 1080pavifenc: 6, Losslessuvg266: Bosphorus 1080p - Very Fasttensorflow: CPU - 1 - ResNet-50speedb: Seq Fillaom-av1: Speed 11 Realtime - Bosphorus 4Kuvg266: Bosphorus 1080p - Super Fastwebp: Quality 100, Losslessonednn: IP Shapes 3D - CPUcompress-pbzip2: FreeBSD-13.0-RELEASE-amd64-memstick.img Compressionrnnoise: 26 Minute Long Talking Samplex265: Bosphorus 1080ptensorflow: CPU - 1 - AlexNetonednn: Convolution Batch Shapes Auto - CPUsrsran: PUSCH Processor Benchmark, Throughput Totaluvg266: Bosphorus 1080p - Ultra Fastkvazaar: Bosphorus 1080p - Very Fastsvt-av1: Preset 12 - Bosphorus 4Kavifenc: 6draco: Church Facadesrsran: PDSCH Processor Benchmark, Throughput Totalsvt-av1: Preset 8 - Bosphorus 1080paom-av1: Speed 8 Realtime - Bosphorus 1080paom-av1: Speed 11 Realtime - Bosphorus 1080paom-av1: Speed 9 Realtime - Bosphorus 1080pkvazaar: Bosphorus 1080p - Super Fastdraco: Lionavifenc: 10, Losslessaom-av1: Speed 10 Realtime - Bosphorus 1080pwebp: Quality 100, Highest Compressiontensorflow: CPU - 1 - GoogLeNetkvazaar: Bosphorus 1080p - Ultra Fastonednn: Deconvolution Batch shapes_3d - CPUaom-av1: Speed 6 Realtime - Bosphorus 1080psrsran: PDSCH Processor Benchmark, Throughput Threadwebp: Quality 100svt-av1: Preset 12 - Bosphorus 1080psvt-av1: Preset 13 - Bosphorus 1080pwebp: DefaultNoctua NH-L9a-AM5Noctua NH-L12Sx771828.8091294.46961.0591388.345435710976667439.91430.382388655384324365.47138462.45523.6306.09125.031.00140.6126229029163055213.4441.97174.48953158119645916.93193916232000000.19193.7211704819400510173487845171049073733997957882043974134485507127815384303.32581178.12163.880293981133.9371.23258.4558.4525.30142.8028.33141.2853.330194.43564.9265.02131.4830.37905143173.26748478257.90780312552.4424194.707.121173455.217.21199100707997327.1572.061410711533333161286924261.40.363183.773.36746231980180.12270704235521629.46303218314000284393333299375.9211697911.30307567372711.93597.3199603399.0965.79565.66963.40455557263.93163058864.9409.2911.6157223631.52512.06495931.5058417731021584654101959685361800027171344121852396335544815170134049697822522961198512.66505145910.2641185514.05672.867752.8090745679428673.21156.88160.5213.3013.451.3026475.5318.5029.21929.77615.03112.31912.44115.5031971.137.3584.47894.25816.5320.0431.02279.1980.6211.42275.7872.77144.91200.8015293666737316333152750000122230000736763335000233399749667994181.45578460000531756667199496667373806667130.8322.7290.2290.1323.455723.948.1424.8356.105543.0148.656114.1967.48369.6745.0151114.01257.295329.21183.42.6416039.6917.375637827914436.22343.9936.21345.5363.0516.33510.44358.1215.1597931791.6960.782.044.317487.8875977.76977.5814.368.252191205.769.1898.0696.7477.11151235443.4117.535253.33270.70274.23121.4939204.954270.014.9750.42160.115.75030239.80900.314.84414.595503.42023.221738.6451231.09914.9331372.543502511576667419.83410.573369695366594348.27144429.25694.5293.27525.521.04560.6418218192171522202.7142168.14753991318766117.15185746243666670.19185.0511827842288621333556804116749878545473987827384804310999687133404503973.47368169.91156.694149744134.7311.24060.3160.3525.68135.7729.14135.1593.475204.62265.2665.36126.1450.38928143158.25724878272.41983612552.5640264.927.201116485.497.2589395979948897.4672.981342412036667167278774253.50.363353.353.52703239680474.52416350733781701.05290118722333286455833285773.4241760811.71577688366712.31077.6054575129.4963.12263.22561.38113156459.82669284162.3099.6912.1431233832.42712.65076132.30386826410716049661320410343011849266178059322453785882594474178994852129110496351148013.04194920210.6311129714.53122.970362.9268547774333374.12151.46162.7513.9614.121.2778579.8619.2029.82030.53515.60412.76012.90216.1932829.635.8264.63998.41917.4420.7830.96281.6910.6511.66281.7975.34150.08250.80158733333373170001536766671281733337417200049996000997610001013790.15594253333549706667199310000385503333131.7623.6891.1393.1824.405728.547.9525.8557.415547.9148.716117.0966.29382.3565.0431013.96658.915930.37183.42.6074841.0816.650683488139137.88145.6637.18547.1365.9016.66610.03560.1615.38102142192.3262.842.054.236697.6315787.72879.6514.458.222461206.571.42100.9499.8756.84450005500.6120.332259.22275.24263.51124.2438604.804275.904.9850.21163.475.74921240.33902.114.86429.911501.53223.13OpenBenchmarking.org

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 6.8Build: allmodconfigNoctua NH-L9a-AM5Noctua NH-L12Sx77400800120016002000SE +/- 2.16, N = 3SE +/- 2.01, N = 31828.811738.65

Blender

Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: Barbershop - Compute: CPU-OnlyNoctua NH-L9a-AM5Noctua NH-L12Sx7730060090012001500SE +/- 2.39, N = 3SE +/- 0.65, N = 31294.461231.09

Timed Node.js Compilation

This test profile times how long it takes to build/compile Node.js itself from source. Node.js is a JavaScript run-time built from the Chrome V8 JavaScript engine while itself is written in C/C++. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Node.js Compilation 21.7.2Time To CompileNoctua NH-L9a-AM5Noctua NH-L12Sx772004006008001000SE +/- 0.79, N = 3SE +/- 0.41, N = 3961.06914.93

Xmrig

Xmrig is an open-source cross-platform CPU/GPU miner for RandomX, KawPow, CryptoNight and AstroBWT. This test profile is setup to measure the Xmrig CPU mining performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgH/s, More Is BetterXmrig 6.21Variant: GhostRider - Hash Count: 1MNoctua NH-L12Sx77Noctua NH-L9a-AM530060090012001500SE +/- 6.97, N = 3SE +/- 4.45, N = 31372.51388.31. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 3 - Resolution: 4K - Samples Per Pixel: 32 - Renderer: Path Tracer - Acceleration: CPUNoctua NH-L9a-AM5Noctua NH-L12Sx77100K200K300K400K500KSE +/- 539.49, N = 3SE +/- 349.30, N = 3454357435025

Quicksilver

Quicksilver is a proxy application that represents some elements of the Mercury workload by solving a simplified dynamic Monte Carlo particle transport problem. Quicksilver is developed by Lawrence Livermore National Laboratory (LLNL) and this test profile currently makes use of the OpenMP CPU threaded code path. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFigure Of Merit, More Is BetterQuicksilver 20230818Input: CTS2Noctua NH-L9a-AM5Noctua NH-L12Sx772M4M6M8M10MSE +/- 8819.17, N = 3SE +/- 8819.17, N = 310976667115766671. (CXX) g++ options: -fopenmp -O3 -march=native

Blender

Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: Pabellon Barcelona - Compute: CPU-OnlyNoctua NH-L9a-AM5Noctua NH-L12Sx77100200300400500SE +/- 0.42, N = 3SE +/- 1.45, N = 3439.91419.83

Timed Godot Game Engine Compilation

This test times how long it takes to compile the Godot Game Engine. Godot is a popular, open-source, cross-platform 2D/3D game engine and is built using the SCons build system and targeting the X11 platform. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Godot Game Engine Compilation 4.0Time To CompileNoctua NH-L9a-AM5Noctua NH-L12Sx7790180270360450SE +/- 0.19, N = 3SE +/- 0.18, N = 3430.38410.57

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 2 - Resolution: 4K - Samples Per Pixel: 32 - Renderer: Path Tracer - Acceleration: CPUNoctua NH-L9a-AM5Noctua NH-L12Sx7780K160K240K320K400KSE +/- 363.15, N = 3SE +/- 319.84, N = 3388655369695

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 1 - Resolution: 4K - Samples Per Pixel: 32 - Renderer: Path Tracer - Acceleration: CPUNoctua NH-L9a-AM5Noctua NH-L12Sx7780K160K240K320K400KSE +/- 615.82, N = 3SE +/- 623.06, N = 3384324366594

Blender

Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: Classroom - Compute: CPU-OnlyNoctua NH-L9a-AM5Noctua NH-L12Sx7780160240320400SE +/- 0.18, N = 3SE +/- 0.54, N = 3365.47348.27

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 3.3Algorithm: RSA4096Noctua NH-L9a-AM5Noctua NH-L12Sx7730K60K90K120K150KSE +/- 395.07, N = 3SE +/- 169.14, N = 3138462.4144429.21. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

OpenBenchmarking.orgsign/s, More Is BetterOpenSSL 3.3Algorithm: RSA4096Noctua NH-L9a-AM5Noctua NH-L12Sx7712002400360048006000SE +/- 37.20, N = 3SE +/- 10.54, N = 35523.65694.51. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

C-Ray

OpenBenchmarking.orgSeconds, Fewer Is BetterC-Ray 2.0Resolution: 4K - Rays Per Pixel: 16Noctua NH-L9a-AM5Noctua NH-L12Sx7770140210280350SE +/- 0.41, N = 3SE +/- 0.12, N = 3306.09293.281. (CC) gcc options: -lpthread -lm

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 2.16.1Device: CPU - Batch Size: 64 - Model: ResNet-50Noctua NH-L9a-AM5Noctua NH-L12Sx77612182430SE +/- 0.02, N = 3SE +/- 0.01, N = 325.0325.52

ASTC Encoder

ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.7Preset: Very ThoroughNoctua NH-L9a-AM5Noctua NH-L12Sx770.23530.47060.70590.94121.1765SE +/- 0.0035, N = 3SE +/- 0.0005, N = 31.00141.04561. (CXX) g++ options: -O3 -flto -pthread

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.7Preset: ExhaustiveNoctua NH-L9a-AM5Noctua NH-L12Sx770.14440.28880.43320.57760.722SE +/- 0.0013, N = 3SE +/- 0.0005, N = 30.61260.64181. (CXX) g++ options: -O3 -flto -pthread

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 3 - Resolution: 4K - Samples Per Pixel: 16 - Renderer: Path Tracer - Acceleration: CPUNoctua NH-L9a-AM5Noctua NH-L12Sx7750K100K150K200K250KSE +/- 323.35, N = 3SE +/- 372.60, N = 3229029218192

BRL-CAD

BRL-CAD is a cross-platform, open-source solid modeling system with built-in benchmark mode. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgVGR Performance Metric, More Is BetterBRL-CAD 7.38.2VGR Performance MetricNoctua NH-L9a-AM5Noctua NH-L12Sx7740K80K120K160K200K1630551715221. (CXX) g++ options: -std=c++17 -pipe -fvisibility=hidden -fno-strict-aliasing -fno-common -fexceptions -ftemplate-depth-128 -m64 -ggdb3 -O3 -fipa-pta -fstrength-reduce -finline-functions -flto -ltcl8.6 -lnetpbm -lregex_brl -lz_brl -lassimp -ldl -lm -ltk8.6

Primesieve

Primesieve generates prime numbers using a highly optimized sieve of Eratosthenes implementation. Primesieve primarily benchmarks the CPU's L1/L2 cache performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterPrimesieve 12.1Length: 1e13Noctua NH-L9a-AM5Noctua NH-L12Sx7750100150200250SE +/- 0.34, N = 3SE +/- 0.14, N = 3213.44202.711. (CXX) g++ options: -O3

Llamafile

Mozilla's Llamafile allows distributing and running large language models (LLMs) as a single file. Llamafile aims to make open-source LLMs more accessible to developers and users. Llamafile supports a variety of models, CPUs and GPUs, and other options. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.7Test: wizardcoder-python-34b-v1.0.Q6_K - Acceleration: CPUNoctua NH-L9a-AM5Noctua NH-L12Sx770.450.91.351.82.25SE +/- 0.00, N = 3SE +/- 0.00, N = 31.972.00

SPECFEM3D

simulates acoustic (fluid), elastic (solid), coupled acoustic/elastic, poroelastic or seismic wave propagation in any type of conforming mesh of hexahedra. This test profile currently relies on CPU-based execution for SPECFEM3D and using a variety of their built-in examples/models for benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.1.1Model: Layered HalfspaceNoctua NH-L9a-AM5Noctua NH-L12Sx774080120160200SE +/- 1.92, N = 3SE +/- 2.06, N = 4174.49168.151. (F9X) gfortran options: -O2 -fopenmp -std=f2008 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 2 - Resolution: 4K - Samples Per Pixel: 16 - Renderer: Path Tracer - Acceleration: CPUNoctua NH-L9a-AM5Noctua NH-L12Sx7740K80K120K160K200KSE +/- 168.49, N = 3SE +/- 108.54, N = 3196459187661

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 7.0Encoder: libx264 - Scenario: UploadNoctua NH-L9a-AM5Noctua NH-L12Sx7748121620SE +/- 0.05, N = 3SE +/- 0.02, N = 316.9317.151. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 1 - Resolution: 4K - Samples Per Pixel: 16 - Renderer: Path Tracer - Acceleration: CPUNoctua NH-L9a-AM5Noctua NH-L12Sx7740K80K120K160K200KSE +/- 203.99, N = 3SE +/- 383.05, N = 3193916185746

Quicksilver

Quicksilver is a proxy application that represents some elements of the Mercury workload by solving a simplified dynamic Monte Carlo particle transport problem. Quicksilver is developed by Lawrence Livermore National Laboratory (LLNL) and this test profile currently makes use of the OpenMP CPU threaded code path. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFigure Of Merit, More Is BetterQuicksilver 20230818Input: CORAL2 P2Noctua NH-L9a-AM5Noctua NH-L12Sx775M10M15M20M25MSE +/- 28867.51, N = 3SE +/- 72188.03, N = 323200000243666671. (CXX) g++ options: -fopenmp -O3 -march=native

AOM AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.9Encoder Mode: Speed 0 Two-Pass - Input: Bosphorus 4KNoctua NH-L9a-AM5Noctua NH-L12Sx770.04280.08560.12840.17120.214SE +/- 0.00, N = 3SE +/- 0.00, N = 80.190.191. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm

Blender

Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: Junkshop - Compute: CPU-OnlyNoctua NH-L9a-AM5Noctua NH-L12Sx774080120160200SE +/- 0.65, N = 3SE +/- 0.29, N = 3193.72185.05

Speedb

Speedb is a next-generation key value storage engine that is RocksDB compatible and aiming for stability, efficiency, and performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Random Fill SyncNoctua NH-L9a-AM5Noctua NH-L12Sx773K6K9K12K15KSE +/- 81.96, N = 15SE +/- 33.01, N = 311704118271. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti

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 3.3Algorithm: AES-256-GCMNoctua NH-L9a-AM5Noctua NH-L12Sx7720000M40000M60000M80000M100000MSE +/- 53831533.36, N = 3SE +/- 61023915.20, N = 381940051017842288621331. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.3Algorithm: ChaCha20-Poly1305Noctua NH-L9a-AM5Noctua NH-L12Sx778000M16000M24000M32000M40000MSE +/- 9263127.14, N = 3SE +/- 4691298.45, N = 334878451710355680411671. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.3Algorithm: ChaCha20Noctua NH-L9a-AM5Noctua NH-L12Sx7711000M22000M33000M44000M55000MSE +/- 4327976.13, N = 3SE +/- 14724634.33, N = 349073733997498785454731. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.3Algorithm: AES-128-GCMNoctua NH-L9a-AM5Noctua NH-L12Sx7720000M40000M60000M80000M100000MSE +/- 73901143.94, N = 3SE +/- 11567880.82, N = 395788204397987827384801. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.3Algorithm: SHA512Noctua NH-L9a-AM5Noctua NH-L12Sx77900M1800M2700M3600M4500MSE +/- 3014246.12, N = 3SE +/- 1471079.75, N = 3413448550743109996871. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

OpenBenchmarking.orgbyte/s, More Is BetterOpenSSL 3.3Algorithm: SHA256Noctua NH-L9a-AM5Noctua NH-L12Sx773000M6000M9000M12000M15000MSE +/- 28997219.45, N = 3SE +/- 5836324.01, N = 312781538430133404503971. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

OSPRay

Intel OSPRay is a portable ray-tracing engine for high-performance, high-fidelity scientific visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 3.1Benchmark: particle_volume/scivis/real_timeNoctua NH-L9a-AM5Noctua NH-L12Sx770.78161.56322.34483.12643.908SE +/- 0.00125, N = 3SE +/- 0.00509, N = 33.325813.47368

Blender

Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: Fishy Cat - Compute: CPU-OnlyNoctua NH-L9a-AM5Noctua NH-L12Sx774080120160200SE +/- 0.17, N = 3SE +/- 0.09, N = 3178.12169.91

SPECFEM3D

simulates acoustic (fluid), elastic (solid), coupled acoustic/elastic, poroelastic or seismic wave propagation in any type of conforming mesh of hexahedra. This test profile currently relies on CPU-based execution for SPECFEM3D and using a variety of their built-in examples/models for benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.1.1Model: Water-layered HalfspaceNoctua NH-L9a-AM5Noctua NH-L12Sx774080120160200SE +/- 0.39, N = 3SE +/- 0.35, N = 3163.88156.691. (F9X) gfortran options: -O2 -fopenmp -std=f2008 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

OSPRay

Intel OSPRay is a portable ray-tracing engine for high-performance, high-fidelity scientific visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 3.1Benchmark: particle_volume/pathtracer/real_timeNoctua NH-L9a-AM5Noctua NH-L12Sx77306090120150SE +/- 0.94, N = 3SE +/- 0.06, N = 3133.94134.73

GROMACS

The GROMACS (GROningen MAchine for Chemical Simulations) molecular dynamics package testing with the water_GMX50 data. This test profile allows selecting between CPU and GPU-based GROMACS builds. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNs Per Day, More Is BetterGROMACS 2024Implementation: MPI CPU - Input: water_GMX50_bareNoctua NH-L9a-AM5Noctua NH-L12Sx770.2790.5580.8371.1161.395SE +/- 0.001, N = 3SE +/- 0.002, N = 31.2321.2401. (CXX) g++ options: -O3 -lm

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 7.0Encoder: libx265 - Scenario: PlatformNoctua NH-L9a-AM5Noctua NH-L12Sx771326395265SE +/- 0.08, N = 3SE +/- 0.03, N = 358.4560.311. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

OpenBenchmarking.orgFPS, More Is BetterFFmpeg 7.0Encoder: libx265 - Scenario: Video On DemandNoctua NH-L9a-AM5Noctua NH-L12Sx771428425670SE +/- 0.01, N = 3SE +/- 0.01, N = 358.4560.351. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

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 2.16.1Device: CPU - Batch Size: 32 - Model: ResNet-50Noctua NH-L9a-AM5Noctua NH-L12Sx77612182430SE +/- 0.03, N = 3SE +/- 0.01, N = 325.3025.68

Blender

Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.1Blend File: BMW27 - Compute: CPU-OnlyNoctua NH-L9a-AM5Noctua NH-L12Sx77306090120150SE +/- 0.32, N = 3SE +/- 0.06, N = 3142.80135.77

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 7.0Encoder: libx265 - Scenario: UploadNoctua NH-L9a-AM5Noctua NH-L12Sx77714212835SE +/- 0.05, N = 3SE +/- 0.02, N = 328.3329.141. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

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: 0Noctua NH-L9a-AM5Noctua NH-L12Sx77306090120150SE +/- 0.39, N = 3SE +/- 0.16, N = 3141.29135.161. (CXX) g++ options: -O3 -fPIC -lm

OSPRay

Intel OSPRay is a portable ray-tracing engine for high-performance, high-fidelity scientific visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 3.1Benchmark: particle_volume/ao/real_timeNoctua NH-L9a-AM5Noctua NH-L12Sx770.78191.56382.34573.12763.9095SE +/- 0.00073, N = 3SE +/- 0.00320, N = 33.330193.47520

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 1.11Video Input: Bosphorus 4K - Video Preset: FastNoctua NH-L9a-AM5Noctua NH-L12Sx771.042.083.124.165.2SE +/- 0.012, N = 3SE +/- 0.017, N = 34.4354.6221. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects

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 7.0Encoder: libx264 - Scenario: PlatformNoctua NH-L9a-AM5Noctua NH-L12Sx771530456075SE +/- 0.11, N = 3SE +/- 0.06, N = 364.9265.261. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

OpenBenchmarking.orgFPS, More Is BetterFFmpeg 7.0Encoder: libx264 - Scenario: Video On DemandNoctua NH-L9a-AM5Noctua NH-L12Sx771530456075SE +/- 0.02, N = 3SE +/- 0.11, N = 365.0265.361. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

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 6.8Build: defconfigNoctua NH-L9a-AM5Noctua NH-L12Sx77306090120150SE +/- 0.23, N = 3SE +/- 0.20, N = 3131.48126.15

NAMD

NAMD is a parallel molecular dynamics code designed for high-performance simulation of large biomolecular systems. NAMD was developed by the Theoretical and Computational Biophysics Group in the Beckman Institute for Advanced Science and Technology at the University of Illinois at Urbana-Champaign. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgns/day, More Is BetterNAMD 3.0b6Input: STMV with 1,066,628 AtomsNoctua NH-L9a-AM5Noctua NH-L12Sx770.08760.17520.26280.35040.438SE +/- 0.00005, N = 3SE +/- 0.00032, N = 30.379050.38928

CacheBench

This is a performance test of CacheBench, which is part of LLCbench. CacheBench is designed to test the memory and cache bandwidth performance Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterCacheBenchTest: Read / Modify / WriteNoctua NH-L12Sx77Noctua NH-L9a-AM530K60K90K120K150KSE +/- 41.80, N = 3SE +/- 58.09, N = 3143158.26143173.27MIN: 112699.84 / MAX: 155475.21MIN: 113186.19 / MAX: 155462.761. (CC) gcc options: -O3 -lrt

OpenBenchmarking.orgMB/s, More Is BetterCacheBenchTest: WriteNoctua NH-L9a-AM5Noctua NH-L12Sx7720K40K60K80K100KSE +/- 21.71, N = 3SE +/- 9.29, N = 378257.9178272.42MIN: 77933.85 / MAX: 78362.75MIN: 77614.46 / MAX: 78367.251. (CC) gcc options: -O3 -lrt

OpenBenchmarking.orgMB/s, More Is BetterCacheBenchTest: ReadNoctua NH-L9a-AM5Noctua NH-L12Sx773K6K9K12K15KSE +/- 0.99, N = 3SE +/- 0.03, N = 312552.4412552.56MIN: 12549.21 / MAX: 12599.84MIN: 12550.6 / MAX: 12553.061. (CC) gcc options: -O3 -lrt

uvg266

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.8.0Video Input: Bosphorus 4K - Video Preset: SlowNoctua NH-L9a-AM5Noctua NH-L12Sx771.1072.2143.3214.4285.535SE +/- 0.01, N = 3SE +/- 0.01, N = 34.704.92

Llamafile

Mozilla's Llamafile allows distributing and running large language models (LLMs) as a single file. Llamafile aims to make open-source LLMs more accessible to developers and users. Llamafile supports a variety of models, CPUs and GPUs, and other options. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.7Test: mistral-7b-instruct-v0.2.Q8_0 - Acceleration: CPUNoctua NH-L9a-AM5Noctua NH-L12Sx77246810SE +/- 0.04, N = 3SE +/- 0.01, N = 37.127.20

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 3 - Resolution: 1080p - Samples Per Pixel: 32 - Renderer: Path Tracer - Acceleration: CPUNoctua NH-L9a-AM5Noctua NH-L12Sx7730K60K90K120K150KSE +/- 207.03, N = 3SE +/- 9.02, N = 3117345111648

uvg266

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.8.0Video Input: Bosphorus 4K - Video Preset: MediumNoctua NH-L9a-AM5Noctua NH-L12Sx771.23532.47063.70594.94126.1765SE +/- 0.01, N = 3SE +/- 0.01, N = 35.215.49

High Performance Conjugate Gradient

HPCG is the High Performance Conjugate Gradient and is a new scientific benchmark from Sandia National Lans focused for super-computer testing with modern real-world workloads compared to HPCC. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHigh Performance Conjugate Gradient 3.1X Y Z: 104 104 104 - RT: 60Noctua NH-L9a-AM5Noctua NH-L12Sx77246810SE +/- 0.00254, N = 3SE +/- 0.00514, N = 37.211997.258931. (CXX) g++ options: -O3 -ffast-math -ftree-vectorize -lmpi_cxx -lmpi

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 2 - Resolution: 1080p - Samples Per Pixel: 32 - Renderer: Path Tracer - Acceleration: CPUNoctua NH-L9a-AM5Noctua NH-L12Sx7720K40K60K80K100KSE +/- 315.38, N = 3SE +/- 356.50, N = 310070795979

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 1 - Resolution: 1080p - Samples Per Pixel: 32 - Renderer: Path Tracer - Acceleration: CPUNoctua NH-L9a-AM5Noctua NH-L12Sx7720K40K60K80K100KSE +/- 81.43, N = 3SE +/- 275.03, N = 39973294889

AOM AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.9Encoder Mode: Speed 4 Two-Pass - Input: Bosphorus 4KNoctua NH-L9a-AM5Noctua NH-L12Sx77246810SE +/- 0.01, N = 3SE +/- 0.02, N = 37.157.461. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm

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 2.16.1Device: CPU - Batch Size: 64 - Model: GoogLeNetNoctua NH-L9a-AM5Noctua NH-L12Sx771632486480SE +/- 0.05, N = 3SE +/- 0.03, N = 372.0672.98

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 3 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path Tracer - Acceleration: CPUNoctua NH-L9a-AM5Noctua NH-L12Sx773K6K9K12K15KSE +/- 9.81, N = 3SE +/- 32.65, N = 31410713424

Quicksilver

Quicksilver is a proxy application that represents some elements of the Mercury workload by solving a simplified dynamic Monte Carlo particle transport problem. Quicksilver is developed by Lawrence Livermore National Laboratory (LLNL) and this test profile currently makes use of the OpenMP CPU threaded code path. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFigure Of Merit, More Is BetterQuicksilver 20230818Input: CORAL2 P1Noctua NH-L9a-AM5Noctua NH-L12Sx773M6M9M12M15MSE +/- 48074.02, N = 3SE +/- 16666.67, N = 311533333120366671. (CXX) g++ options: -fopenmp -O3 -march=native

Stockfish

This is a test of Stockfish, an advanced open-source C++11 chess benchmark that can scale up to 1024 CPU threads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNodes Per Second, More Is BetterStockfish 16.1Chess BenchmarkNoctua NH-L9a-AM5Noctua NH-L12Sx774M8M12M16M20MSE +/- 30490.11, N = 3SE +/- 97334.34, N = 316128692167278771. (CXX) g++ options: -lgcov -m64 -lpthread -fno-exceptions -std=c++17 -fno-peel-loops -fno-tracer -pedantic -O3 -funroll-loops -msse -msse3 -mpopcnt -mavx2 -mbmi -mavx512f -mavx512bw -mavx512vnni -mavx512dq -mavx512vl -msse4.1 -mssse3 -msse2 -mbmi2 -flto -flto-partition=one -flto=jobserver

QuantLib

QuantLib is an open-source library/framework around quantitative finance for modeling, trading and risk management scenarios. QuantLib is written in C++ with Boost and its built-in benchmark used reports the QuantLib Benchmark Index benchmark score. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMFLOPS, More Is BetterQuantLib 1.32Configuration: Single-ThreadedNoctua NH-L12Sx77Noctua NH-L9a-AM59001800270036004500SE +/- 50.56, N = 13SE +/- 43.33, N = 124253.54261.41. (CXX) g++ options: -O3 -march=native -fPIE -pie

Intel Open Image Denoise

Open Image Denoise is a denoising library for ray-tracing and part of the Intel oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 2.2Run: RT.ldr_alb_nrm.3840x2160 - Device: CPU-OnlyNoctua NH-L9a-AM5Noctua NH-L12Sx770.0810.1620.2430.3240.405SE +/- 0.00, N = 3SE +/- 0.00, N = 30.360.36

oneDNN

This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.4Harness: Recurrent Neural Network Training - Engine: CPUNoctua NH-L12Sx77Noctua NH-L9a-AM57001400210028003500SE +/- 5.14, N = 3SE +/- 3.10, N = 33353.353183.77MIN: 3333.48MIN: 3155.451. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

OSPRay

Intel OSPRay is a portable ray-tracing engine for high-performance, high-fidelity scientific visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 3.1Benchmark: gravity_spheres_volume/dim_512/pathtracer/real_timeNoctua NH-L9a-AM5Noctua NH-L12Sx770.79361.58722.38083.17443.968SE +/- 0.00212, N = 3SE +/- 0.00127, N = 33.367463.52703

Speedb

Speedb is a next-generation key value storage engine that is RocksDB compatible and aiming for stability, efficiency, and performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Read While WritingNoctua NH-L9a-AM5Noctua NH-L12Sx77500K1000K1500K2000K2500KSE +/- 25833.92, N = 4SE +/- 29574.10, N = 4231980123968041. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti

SPECFEM3D

simulates acoustic (fluid), elastic (solid), coupled acoustic/elastic, poroelastic or seismic wave propagation in any type of conforming mesh of hexahedra. This test profile currently relies on CPU-based execution for SPECFEM3D and using a variety of their built-in examples/models for benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.1.1Model: Homogeneous HalfspaceNoctua NH-L9a-AM5Noctua NH-L12Sx7720406080100SE +/- 0.10, N = 3SE +/- 0.56, N = 380.1274.521. (F9X) gfortran options: -O2 -fopenmp -std=f2008 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 3 - Resolution: 1080p - Samples Per Pixel: 1 - Renderer: Path Tracer - Acceleration: CPUNoctua NH-L9a-AM5Noctua NH-L12Sx778001600240032004000SE +/- 2.52, N = 3SE +/- 6.36, N = 335523378

oneDNN

This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.4Harness: Recurrent Neural Network Inference - Engine: CPUNoctua NH-L12Sx77Noctua NH-L9a-AM5400800120016002000SE +/- 6.05, N = 3SE +/- 0.45, N = 31701.051629.46MIN: 1677.63MIN: 1617.521. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 2 - Resolution: 1080p - Samples Per Pixel: 1 - Renderer: Path Tracer - Acceleration: CPUNoctua NH-L9a-AM5Noctua NH-L12Sx777001400210028003500SE +/- 4.16, N = 3SE +/- 3.00, N = 330322901

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: 512Noctua NH-L9a-AM5Noctua NH-L12Sx774M8M12M16M20MSE +/- 305766.10, N = 12SE +/- 28415.57, N = 318314000187223331. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 4 - Buffer Length: 256 - Filter Length: 57Noctua NH-L9a-AM5Noctua NH-L12Sx7760M120M180M240M300MSE +/- 1540739.36, N = 3SE +/- 3412871.24, N = 122843933332864558331. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 1 - Resolution: 1080p - Samples Per Pixel: 1 - Renderer: Path Tracer - Acceleration: CPUNoctua NH-L9a-AM5Noctua NH-L12Sx776001200180024003000SE +/- 1.53, N = 3SE +/- 1.00, N = 329932857

C-Ray

OpenBenchmarking.orgSeconds, Fewer Is BetterC-Ray 2.0Resolution: 1080p - Rays Per Pixel: 16Noctua NH-L9a-AM5Noctua NH-L12Sx7720406080100SE +/- 0.21, N = 3SE +/- 0.11, N = 375.9273.421. (CC) gcc options: -lpthread -lm

Chaos Group V-RAY

This is a test of Chaos Group's V-RAY benchmark. V-RAY is a commercial renderer that can integrate with various creator software products like SketchUp and 3ds Max. The V-RAY benchmark is standalone and supports CPU and NVIDIA CUDA/RTX based rendering. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgvsamples, More Is BetterChaos Group V-RAY 6.0Mode: CPUNoctua NH-L9a-AM5Noctua NH-L12Sx774K8K12K16K20KSE +/- 65.38, N = 3SE +/- 17.50, N = 31697917608

Embree

Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs (and GPUs via SYCL) and supporting instruction sets such as SSE, AVX, AVX2, and AVX-512. Embree also supports making use of the Intel SPMD Program Compiler (ISPC). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 4.3Binary: Pathtracer - Model: Asian Dragon ObjNoctua NH-L9a-AM5Noctua NH-L12Sx773691215SE +/- 0.01, N = 3SE +/- 0.01, N = 311.3011.72MIN: 11.23 / MAX: 11.52MIN: 11.65 / MAX: 11.88

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: 57Noctua NH-L9a-AM5Noctua NH-L12Sx7716M32M48M64M80MSE +/- 568285.01, N = 11SE +/- 18764.62, N = 375673727768836671. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

Embree

Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs (and GPUs via SYCL) and supporting instruction sets such as SSE, AVX, AVX2, and AVX-512. Embree also supports making use of the Intel SPMD Program Compiler (ISPC). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 4.3Binary: Pathtracer ISPC - Model: Asian Dragon ObjNoctua NH-L9a-AM5Noctua NH-L12Sx773691215SE +/- 0.01, N = 3SE +/- 0.01, N = 311.9412.31MIN: 11.85 / MAX: 12.18MIN: 12.23 / MAX: 12.54

ASTC Encoder

ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.7Preset: ThoroughNoctua NH-L9a-AM5Noctua NH-L12Sx77246810SE +/- 0.0278, N = 3SE +/- 0.0107, N = 37.31997.60541. (CXX) g++ options: -O3 -flto -pthread

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 3 - Resolution: 1080p - Samples Per Pixel: 16 - Renderer: Path Tracer - Acceleration: CPUNoctua NH-L9a-AM5Noctua NH-L12Sx7713K26K39K52K65KSE +/- 64.95, N = 3SE +/- 50.80, N = 36033957512

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: SlowNoctua NH-L9a-AM5Noctua NH-L12Sx773691215SE +/- 0.01, N = 3SE +/- 0.02, N = 39.099.491. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

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: 2Noctua NH-L9a-AM5Noctua NH-L12Sx771530456075SE +/- 0.14, N = 3SE +/- 0.11, N = 365.8063.121. (CXX) g++ options: -O3 -fPIC -lm

Timed PHP Compilation

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

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed PHP Compilation 8.3.4Time To CompileNoctua NH-L9a-AM5Noctua NH-L12Sx771530456075SE +/- 0.07, N = 3SE +/- 0.07, N = 365.6763.23

SPECFEM3D

simulates acoustic (fluid), elastic (solid), coupled acoustic/elastic, poroelastic or seismic wave propagation in any type of conforming mesh of hexahedra. This test profile currently relies on CPU-based execution for SPECFEM3D and using a variety of their built-in examples/models for benchmarking. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.1.1Model: Mount St. HelensNoctua NH-L9a-AM5Noctua NH-L12Sx771428425670SE +/- 0.45, N = 3SE +/- 0.87, N = 363.4061.381. (F9X) gfortran options: -O2 -fopenmp -std=f2008 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

OpenBenchmarking.orgSeconds, Fewer Is BetterSPECFEM3D 4.1.1Model: Tomographic ModelNoctua NH-L9a-AM5Noctua NH-L12Sx771428425670SE +/- 0.68, N = 3SE +/- 0.45, N = 363.9359.831. (F9X) gfortran options: -O2 -fopenmp -std=f2008 -fimplicit-none -fmax-errors=10 -pedantic -pedantic-errors -O3 -finline-functions -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz

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 7.0Time To CompileNoctua NH-L9a-AM5Noctua NH-L12Sx771428425670SE +/- 0.23, N = 3SE +/- 0.06, N = 364.9462.31

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: MediumNoctua NH-L9a-AM5Noctua NH-L12Sx773691215SE +/- 0.01, N = 3SE +/- 0.00, N = 39.299.691. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

Embree

Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs (and GPUs via SYCL) and supporting instruction sets such as SSE, AVX, AVX2, and AVX-512. Embree also supports making use of the Intel SPMD Program Compiler (ISPC). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 4.3Binary: Pathtracer - Model: CrownNoctua NH-L9a-AM5Noctua NH-L12Sx773691215SE +/- 0.02, N = 3SE +/- 0.02, N = 311.6212.14MIN: 11.48 / MAX: 11.95MIN: 12.03 / MAX: 12.38

GraphicsMagick

This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: SharpenNoctua NH-L9a-AM5Noctua NH-L12Sx77612182430SE +/- 0.00, N = 3SE +/- 0.00, N = 322231. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: EnhancedNoctua NH-L9a-AM5Noctua NH-L12Sx77918273645SE +/- 0.00, N = 3SE +/- 0.00, N = 336381. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp

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 0.10.1Input: PNG - Quality: 80Noctua NH-L9a-AM5Noctua NH-L12Sx77816243240SE +/- 0.34, N = 3SE +/- 0.23, N = 331.5332.431. (CXX) g++ options: -fno-rtti -O3 -fPIE -pie -lm

Embree

Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs (and GPUs via SYCL) and supporting instruction sets such as SSE, AVX, AVX2, and AVX-512. Embree also supports making use of the Intel SPMD Program Compiler (ISPC). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 4.3Binary: Pathtracer ISPC - Model: CrownNoctua NH-L9a-AM5Noctua NH-L12Sx773691215SE +/- 0.01, N = 3SE +/- 0.01, N = 312.0612.65MIN: 11.93 / MAX: 12.4MIN: 12.52 / MAX: 12.98

GraphicsMagick

This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: Noise-GaussianNoctua NH-L9a-AM5Noctua NH-L12Sx771428425670SE +/- 0.33, N = 3SE +/- 0.00, N = 359611. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp

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 0.10.1Input: JPEG - Quality: 80Noctua NH-L9a-AM5Noctua NH-L12Sx77816243240SE +/- 0.12, N = 3SE +/- 0.03, N = 331.5132.301. (CXX) g++ options: -fno-rtti -O3 -fPIE -pie -lm

Speedb

Speedb is a next-generation key value storage engine that is RocksDB compatible and aiming for stability, efficiency, and performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Random FillNoctua NH-L9a-AM5Noctua NH-L12Sx77200K400K600K800K1000KSE +/- 702.57, N = 3SE +/- 1112.81, N = 38417738682641. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti

GraphicsMagick

This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: SwirlNoctua NH-L9a-AM5Noctua NH-L12Sx7720406080100SE +/- 0.00, N = 3SE +/- 0.33, N = 31021071. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: RotateNoctua NH-L9a-AM5Noctua NH-L12Sx774080120160200SE +/- 0.33, N = 3SE +/- 0.33, N = 31581601. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp

Speedb

Speedb is a next-generation key value storage engine that is RocksDB compatible and aiming for stability, efficiency, and performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Update RandomNoctua NH-L9a-AM5Noctua NH-L12Sx77110K220K330K440K550KSE +/- 677.51, N = 3SE +/- 1118.64, N = 34654104966131. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti

GraphicsMagick

This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: ResizingNoctua NH-L9a-AM5Noctua NH-L12Sx774080120160200SE +/- 0.33, N = 3SE +/- 0.00, N = 31952041. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp

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 9.0Test: OverwriteNoctua NH-L9a-AM5Noctua NH-L12Sx77200K400K600K800K1000KSE +/- 3283.45, N = 3SE +/- 2049.50, N = 396853610343011. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: Read While WritingNoctua NH-L9a-AM5Noctua NH-L12Sx77400K800K1200K1600K2000KSE +/- 3999.65, N = 3SE +/- 2791.28, N = 3180002718492661. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti

Speedb

Speedb is a next-generation key value storage engine that is RocksDB compatible and aiming for stability, efficiency, and performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Read Random Write RandomNoctua NH-L9a-AM5Noctua NH-L12Sx77400K800K1200K1600K2000KSE +/- 2105.22, N = 3SE +/- 1366.09, N = 3171344117805931. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti

GraphicsMagick

This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIterations Per Minute, More Is BetterGraphicsMagick 1.3.43Operation: HWB Color SpaceNoctua NH-L9a-AM5Noctua NH-L12Sx7750100150200250SE +/- 0.33, N = 3SE +/- 0.33, N = 32182241. (CC) gcc options: -fopenmp -O2 -ltiff -ljbig -lsharpyuv -lwebp -lwebpmux -lfreetype -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lzstd -llzma -lbz2 -lz -lm -lpthread -lgomp

Speedb

Speedb is a next-generation key value storage engine that is RocksDB compatible and aiming for stability, efficiency, and performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Random ReadNoctua NH-L9a-AM5Noctua NH-L12Sx7712M24M36M48M60MSE +/- 219893.87, N = 3SE +/- 163539.62, N = 352396335537858821. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti

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 9.0Test: Update RandomNoctua NH-L9a-AM5Noctua NH-L12Sx77130K260K390K520K650KSE +/- 283.54, N = 3SE +/- 2417.86, N = 35448155944741. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: Read Random Write RandomNoctua NH-L9a-AM5Noctua NH-L12Sx77400K800K1200K1600K2000KSE +/- 12289.53, N = 3SE +/- 4621.81, N = 3170134017899481. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: Random ReadNoctua NH-L9a-AM5Noctua NH-L12Sx7711M22M33M44M55MSE +/- 66730.35, N = 3SE +/- 2870.48, N = 349697822521291101. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 2 - Resolution: 1080p - Samples Per Pixel: 16 - Renderer: Path Tracer - Acceleration: CPUNoctua NH-L9a-AM5Noctua NH-L12Sx7711K22K33K44K55KSE +/- 59.90, N = 3SE +/- 155.92, N = 35229649635

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 2 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path Tracer - Acceleration: CPUNoctua NH-L9a-AM5Noctua NH-L12Sx773K6K9K12K15KSE +/- 12.88, N = 3SE +/- 19.86, N = 31198511480

Embree

Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs (and GPUs via SYCL) and supporting instruction sets such as SSE, AVX, AVX2, and AVX-512. Embree also supports making use of the Intel SPMD Program Compiler (ISPC). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 4.3Binary: Pathtracer - Model: Asian DragonNoctua NH-L9a-AM5Noctua NH-L12Sx773691215SE +/- 0.04, N = 3SE +/- 0.00, N = 312.6713.04MIN: 12.56 / MAX: 12.96MIN: 12.99 / MAX: 13.18

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 1 - Resolution: 1080p - Samples Per Pixel: 16 - Renderer: Path Tracer - Acceleration: CPUNoctua NH-L9a-AM5Noctua NH-L12Sx7711K22K33K44K55KSE +/- 198.34, N = 3SE +/- 133.21, N = 35145949202

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 1.11Video Input: Bosphorus 4K - Video Preset: FasterNoctua NH-L9a-AM5Noctua NH-L12Sx773691215SE +/- 0.03, N = 3SE +/- 0.01, N = 310.2610.631. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects

OSPRay Studio

Intel OSPRay Studio is an open-source, interactive visualization and ray-tracing software package. OSPRay Studio makes use of Intel OSPRay, a portable ray-tracing engine for high-performance, high-fidelity visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterOSPRay Studio 1.0Camera: 1 - Resolution: 4K - Samples Per Pixel: 1 - Renderer: Path Tracer - Acceleration: CPUNoctua NH-L9a-AM5Noctua NH-L12Sx773K6K9K12K15KSE +/- 19.91, N = 3SE +/- 32.19, N = 31185511297

Embree

Intel Embree is a collection of high-performance ray-tracing kernels for execution on CPUs (and GPUs via SYCL) and supporting instruction sets such as SSE, AVX, AVX2, and AVX-512. Embree also supports making use of the Intel SPMD Program Compiler (ISPC). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 4.3Binary: Pathtracer ISPC - Model: Asian DragonNoctua NH-L9a-AM5Noctua NH-L12Sx7748121620SE +/- 0.01, N = 3SE +/- 0.03, N = 314.0614.53MIN: 13.97 / MAX: 14.29MIN: 14.39 / MAX: 14.87

OSPRay

Intel OSPRay is a portable ray-tracing engine for high-performance, high-fidelity scientific visualizations. OSPRay builds off Intel's Embree and Intel SPMD Program Compiler (ISPC) components as part of the oneAPI rendering toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 3.1Benchmark: gravity_spheres_volume/dim_512/ao/real_timeNoctua NH-L9a-AM5Noctua NH-L12Sx770.66831.33662.00492.67323.3415SE +/- 0.00310, N = 3SE +/- 0.00569, N = 32.867752.97036

OpenBenchmarking.orgItems Per Second, More Is BetterOSPRay 3.1Benchmark: gravity_spheres_volume/dim_512/scivis/real_timeNoctua NH-L9a-AM5Noctua NH-L12Sx770.65851.3171.97552.6343.2925SE +/- 0.00436, N = 3SE +/- 0.00253, N = 32.809072.92685

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: 57Noctua NH-L9a-AM5Noctua NH-L12Sx77100M200M300M400M500MSE +/- 4290906.70, N = 7SE +/- 1977518.09, N = 34567942864777433331. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

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 2.16.1Device: CPU - Batch Size: 32 - Model: GoogLeNetNoctua NH-L9a-AM5Noctua NH-L12Sx771632486480SE +/- 0.06, N = 3SE +/- 0.04, N = 373.2174.12

OpenBenchmarking.orgimages/sec, More Is BetterTensorFlow 2.16.1Device: CPU - Batch Size: 64 - Model: AlexNetNoctua NH-L12Sx77Noctua NH-L9a-AM5306090120150SE +/- 0.05, N = 3SE +/- 0.23, N = 3151.46156.88

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 7.0Encoder: libx265 - Scenario: LiveNoctua NH-L9a-AM5Noctua NH-L12Sx774080120160200SE +/- 0.45, N = 3SE +/- 0.43, N = 3160.52162.751. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

uvg266

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.8.0Video Input: Bosphorus 4K - Video Preset: Very FastNoctua NH-L9a-AM5Noctua NH-L12Sx7748121620SE +/- 0.01, N = 3SE +/- 0.02, N = 313.3013.96

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.8.0Video Input: Bosphorus 4K - Video Preset: Super FastNoctua NH-L9a-AM5Noctua NH-L12Sx7748121620SE +/- 0.01, N = 3SE +/- 0.02, N = 313.4514.12

NAMD

NAMD is a parallel molecular dynamics code designed for high-performance simulation of large biomolecular systems. NAMD was developed by the Theoretical and Computational Biophysics Group in the Beckman Institute for Advanced Science and Technology at the University of Illinois at Urbana-Champaign. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgns/day, More Is BetterNAMD 3.0b6Input: ATPase with 327,506 AtomsNoctua NH-L12Sx77Noctua NH-L9a-AM50.29310.58620.87931.17241.4655SE +/- 0.00661, N = 3SE +/- 0.00105, N = 31.277851.30264

AOM AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.9Encoder Mode: Speed 6 Realtime - Input: Bosphorus 4KNoctua NH-L9a-AM5Noctua NH-L12Sx7720406080100SE +/- 0.97, N = 15SE +/- 1.18, N = 1575.5379.861. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.9Encoder Mode: Speed 6 Two-Pass - Input: Bosphorus 4KNoctua NH-L9a-AM5Noctua NH-L12Sx77510152025SE +/- 0.06, N = 3SE +/- 0.02, N = 318.5019.201. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm

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 0.10.1Input: PNG - Quality: 90Noctua NH-L9a-AM5Noctua NH-L12Sx77714212835SE +/- 0.08, N = 3SE +/- 0.12, N = 329.2229.821. (CXX) g++ options: -fno-rtti -O3 -fPIE -pie -lm

OpenBenchmarking.orgMP/s, More Is BetterJPEG-XL libjxl 0.10.1Input: JPEG - Quality: 90Noctua NH-L9a-AM5Noctua NH-L12Sx77714212835SE +/- 0.10, N = 3SE +/- 0.06, N = 329.7830.541. (CXX) g++ options: -fno-rtti -O3 -fPIE -pie -lm

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 1.11Video Input: Bosphorus 1080p - Video Preset: FastNoctua NH-L9a-AM5Noctua NH-L12Sx7748121620SE +/- 0.02, N = 3SE +/- 0.03, N = 315.0315.601. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects

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 0.10.1Input: JPEG - Quality: 100Noctua NH-L9a-AM5Noctua NH-L12Sx773691215SE +/- 0.01, N = 3SE +/- 0.00, N = 312.3212.761. (CXX) g++ options: -fno-rtti -O3 -fPIE -pie -lm

OpenBenchmarking.orgMP/s, More Is BetterJPEG-XL libjxl 0.10.1Input: PNG - Quality: 100Noctua NH-L9a-AM5Noctua NH-L12Sx773691215SE +/- 0.01, N = 3SE +/- 0.01, N = 312.4412.901. (CXX) g++ options: -fno-rtti -O3 -fPIE -pie -lm

uvg266

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.8.0Video Input: Bosphorus 4K - Video Preset: Ultra FastNoctua NH-L9a-AM5Noctua NH-L12Sx7748121620SE +/- 0.03, N = 3SE +/- 0.00, N = 315.5016.19

QuantLib

QuantLib is an open-source library/framework around quantitative finance for modeling, trading and risk management scenarios. QuantLib is written in C++ with Boost and its built-in benchmark used reports the QuantLib Benchmark Index benchmark score. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMFLOPS, More Is BetterQuantLib 1.32Configuration: Multi-ThreadedNoctua NH-L9a-AM5Noctua NH-L12Sx777K14K21K28K35KSE +/- 189.38, N = 3SE +/- 43.52, N = 331971.132829.61. (CXX) g++ options: -O3 -march=native -fPIE -pie

Timed Mesa Compilation

This test profile times how long it takes to compile Mesa with Meson/Ninja. For minimizing build dependencies and avoid versioning conflicts, test this is just the core Mesa build without LLVM or the extra Gallium3D/Mesa drivers enabled. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Mesa Compilation 24.0Time To CompileNoctua NH-L9a-AM5Noctua NH-L12Sx77918273645SE +/- 0.03, N = 3SE +/- 0.02, N = 337.3635.83

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 4 - Input: Bosphorus 4KNoctua NH-L9a-AM5Noctua NH-L12Sx771.04382.08763.13144.17525.219SE +/- 0.013, N = 3SE +/- 0.004, N = 34.4784.6391. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 13 - Input: Bosphorus 4KNoctua NH-L9a-AM5Noctua NH-L12Sx7720406080100SE +/- 1.55, N = 15SE +/- 1.38, N = 1594.2698.421. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

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 3.6Video Input: Bosphorus 4KNoctua NH-L9a-AM5Noctua NH-L12Sx7748121620SE +/- 0.10, N = 3SE +/- 0.09, N = 316.5317.441. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

AOM AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.9Encoder Mode: Speed 4 Two-Pass - Input: Bosphorus 1080pNoctua NH-L9a-AM5Noctua NH-L12Sx77510152025SE +/- 0.04, N = 3SE +/- 0.02, N = 320.0420.781. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm

Y-Cruncher

Y-Cruncher is a multi-threaded Pi benchmark capable of computing Pi to trillions of digits. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterY-Cruncher 0.8.3Pi Digits To Calculate: 1BNoctua NH-L9a-AM5Noctua NH-L12Sx77714212835SE +/- 0.07, N = 3SE +/- 0.23, N = 331.0230.96

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 0.10.1CPU Threads: 1Noctua NH-L9a-AM5Noctua NH-L12Sx7720406080100SE +/- 0.82, N = 3SE +/- 0.27, N = 379.2081.69

AOM AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.9Encoder Mode: Speed 0 Two-Pass - Input: Bosphorus 1080pNoctua NH-L9a-AM5Noctua NH-L12Sx770.14630.29260.43890.58520.7315SE +/- 0.00, N = 3SE +/- 0.00, N = 30.620.651. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm

Llamafile

Mozilla's Llamafile allows distributing and running large language models (LLMs) as a single file. Llamafile aims to make open-source LLMs more accessible to developers and users. Llamafile supports a variety of models, CPUs and GPUs, and other options. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.7Test: llava-v1.5-7b-q4 - Acceleration: CPUNoctua NH-L9a-AM5Noctua NH-L12Sx773691215SE +/- 0.01, N = 3SE +/- 0.01, N = 311.4211.66

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 7.0Encoder: libx264 - Scenario: LiveNoctua NH-L9a-AM5Noctua NH-L12Sx7760120180240300SE +/- 1.24, N = 3SE +/- 1.31, N = 3275.78281.791. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

AOM AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.9Encoder Mode: Speed 8 Realtime - Input: Bosphorus 4KNoctua NH-L9a-AM5Noctua NH-L12Sx7720406080100SE +/- 0.77, N = 15SE +/- 0.76, N = 672.7775.341. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm

ASTC Encoder

ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.7Preset: FastNoctua NH-L9a-AM5Noctua NH-L12Sx77306090120150SE +/- 0.25, N = 3SE +/- 0.16, N = 3144.91150.081. (CXX) g++ options: -O3 -flto -pthread

WebP Image Encode

OpenBenchmarking.orgMP/s, More Is BetterWebP Image Encode 1.4Encode Settings: Quality 100, Lossless, Highest CompressionNoctua NH-L9a-AM5Noctua NH-L12Sx770.180.360.540.720.9SE +/- 0.00, N = 3SE +/- 0.00, N = 30.800.801. (CC) gcc options: -fvisibility=hidden -O2 -lm

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: 12 - Buffer Length: 256 - Filter Length: 512Noctua NH-L9a-AM5Noctua NH-L12Sx7730M60M90M120M150MSE +/- 274792.85, N = 3SE +/- 143100.59, N = 31529366671587333331. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 2 - Buffer Length: 256 - Filter Length: 512Noctua NH-L9a-AM5Noctua NH-L12Sx778M16M24M32M40MSE +/- 214222.88, N = 3SE +/- 226124.60, N = 337316333373170001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 2 - Buffer Length: 256 - Filter Length: 57Noctua NH-L9a-AM5Noctua NH-L12Sx7730M60M90M120M150MSE +/- 512477.64, N = 3SE +/- 37118.43, N = 31527500001536766671. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 8 - Buffer Length: 256 - Filter Length: 512Noctua NH-L9a-AM5Noctua NH-L12Sx7730M60M90M120M150MSE +/- 450148.12, N = 3SE +/- 150591.43, N = 31222300001281733331. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 4 - Buffer Length: 256 - Filter Length: 512Noctua NH-L9a-AM5Noctua NH-L12Sx7716M32M48M64M80MSE +/- 303444.85, N = 3SE +/- 419261.65, N = 373676333741720001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 1 - Buffer Length: 256 - Filter Length: 32Noctua NH-L12Sx77Noctua NH-L9a-AM511M22M33M44M55MSE +/- 3511.88, N = 3SE +/- 18550.23, N = 349996000500023331. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 2 - Buffer Length: 256 - Filter Length: 32Noctua NH-L9a-AM5Noctua NH-L12Sx7720M40M60M80M100MSE +/- 36241.47, N = 3SE +/- 7505.55, N = 399749667997610001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

Stress-NG

Stress-NG is a Linux stress tool developed by Colin Ian King. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: AVX-512 VNNINoctua NH-L9a-AM5Noctua NH-L12Sx77200K400K600K800K1000KSE +/- 3387.88, N = 3SE +/- 2132.79, N = 3994181.451013790.151. (CXX) g++ options: -O2 -std=gnu99 -lc

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: 12 - Buffer Length: 256 - Filter Length: 57Noctua NH-L9a-AM5Noctua NH-L12Sx77130M260M390M520M650MSE +/- 3220750.43, N = 3SE +/- 5739527.66, N = 35784600005942533331. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 12 - Buffer Length: 256 - Filter Length: 32Noctua NH-L9a-AM5Noctua NH-L12Sx77120M240M360M480M600MSE +/- 751228.62, N = 3SE +/- 2218560.59, N = 35317566675497066671. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 4 - Buffer Length: 256 - Filter Length: 32Noctua NH-L12Sx77Noctua NH-L9a-AM540M80M120M160M200MSE +/- 36055.51, N = 3SE +/- 27284.51, N = 31993100001994966671. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 8 - Buffer Length: 256 - Filter Length: 32Noctua NH-L9a-AM5Noctua NH-L12Sx7780M160M240M320M400MSE +/- 399597.02, N = 3SE +/- 391932.53, N = 33738066673855033331. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

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 2.16.1Device: CPU - Batch Size: 32 - Model: AlexNetNoctua NH-L9a-AM5Noctua NH-L12Sx77306090120150SE +/- 0.16, N = 3SE +/- 0.22, N = 3130.83131.76

uvg266

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.8.0Video Input: Bosphorus 1080p - Video Preset: SlowNoctua NH-L9a-AM5Noctua NH-L12Sx77612182430SE +/- 0.04, N = 3SE +/- 0.05, N = 322.7223.68

AOM AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.9Encoder Mode: Speed 9 Realtime - Input: Bosphorus 4KNoctua NH-L9a-AM5Noctua NH-L12Sx7720406080100SE +/- 0.44, N = 6SE +/- 1.00, N = 1590.2291.131. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.9Encoder Mode: Speed 10 Realtime - Input: Bosphorus 4KNoctua NH-L9a-AM5Noctua NH-L12Sx7720406080100SE +/- 1.01, N = 15SE +/- 0.28, N = 690.1393.181. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm

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 FastNoctua NH-L9a-AM5Noctua NH-L12Sx77612182430SE +/- 0.01, N = 3SE +/- 0.02, N = 323.4524.401. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

LZ4 Compression

This test measures the time needed to compress/decompress a sample file (silesia archive) using LZ4 compression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterLZ4 Compression 1.9.4Compression Level: 9 - Decompression SpeedNoctua NH-L9a-AM5Noctua NH-L12Sx7712002400360048006000SE +/- 3.57, N = 3SE +/- 3.27, N = 35723.95728.51. (CC) gcc options: -O3

OpenBenchmarking.orgMB/s, More Is BetterLZ4 Compression 1.9.4Compression Level: 9 - Compression SpeedNoctua NH-L12Sx77Noctua NH-L9a-AM51122334455SE +/- 0.14, N = 3SE +/- 0.20, N = 347.9548.141. (CC) gcc options: -O3

uvg266

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.8.0Video Input: Bosphorus 1080p - Video Preset: MediumNoctua NH-L9a-AM5Noctua NH-L12Sx77612182430SE +/- 0.06, N = 3SE +/- 0.01, N = 324.8325.85

PyTorch

This is a benchmark of PyTorch making use of pytorch-benchmark [https://github.com/LukasHedegaard/pytorch-benchmark]. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgbatches/sec, More Is BetterPyTorch 2.2.1Device: CPU - Batch Size: 1 - Model: ResNet-50Noctua NH-L9a-AM5Noctua NH-L12Sx771326395265SE +/- 0.71, N = 3SE +/- 0.69, N = 356.1057.41MIN: 50.95 / MAX: 59.36MIN: 53.92 / MAX: 59.49

LZ4 Compression

This test measures the time needed to compress/decompress a sample file (silesia archive) using LZ4 compression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterLZ4 Compression 1.9.4Compression Level: 3 - Decompression SpeedNoctua NH-L9a-AM5Noctua NH-L12Sx7712002400360048006000SE +/- 21.54, N = 3SE +/- 21.18, N = 35543.05547.91. (CC) gcc options: -O3

OpenBenchmarking.orgMB/s, More Is BetterLZ4 Compression 1.9.4Compression Level: 3 - Compression SpeedNoctua NH-L9a-AM5Noctua NH-L12Sx77306090120150SE +/- 0.64, N = 3SE +/- 0.54, N = 3148.65148.711. (CC) gcc options: -O3

OpenBenchmarking.orgMB/s, More Is BetterLZ4 Compression 1.9.4Compression Level: 1 - Decompression SpeedNoctua NH-L9a-AM5Noctua NH-L12Sx7713002600390052006500SE +/- 1.29, N = 3SE +/- 4.91, N = 36114.16117.01. (CC) gcc options: -O3

OpenBenchmarking.orgMB/s, More Is BetterLZ4 Compression 1.9.4Compression Level: 1 - Compression SpeedNoctua NH-L12Sx77Noctua NH-L9a-AM52004006008001000SE +/- 0.00, N = 3SE +/- 0.31, N = 3966.29967.481. (CC) gcc options: -O3

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 0.10.1CPU Threads: AllNoctua NH-L9a-AM5Noctua NH-L12Sx7780160240320400SE +/- 0.96, N = 3SE +/- 0.18, N = 3369.67382.36

oneDNN

This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.4Harness: Deconvolution Batch shapes_1d - Engine: CPUNoctua NH-L12Sx77Noctua NH-L9a-AM51.13472.26943.40414.53885.6735SE +/- 0.00346, N = 3SE +/- 0.00815, N = 35.043105.01511MIN: 4.61MIN: 4.781. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

Y-Cruncher

Y-Cruncher is a multi-threaded Pi benchmark capable of computing Pi to trillions of digits. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterY-Cruncher 0.8.3Pi Digits To Calculate: 500MNoctua NH-L9a-AM5Noctua NH-L12Sx7748121620SE +/- 0.06, N = 4SE +/- 0.09, N = 414.0113.97

ASTC Encoder

ASTC Encoder (astcenc) is for the Adaptive Scalable Texture Compression (ASTC) format commonly used with OpenGL, OpenGL ES, and Vulkan graphics APIs. This test profile does a coding test of both compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMT/s, More Is BetterASTC Encoder 4.7Preset: MediumNoctua NH-L9a-AM5Noctua NH-L12Sx771326395265SE +/- 0.05, N = 3SE +/- 0.04, N = 357.3058.921. (CXX) g++ options: -O3 -flto -pthread

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 FastNoctua NH-L9a-AM5Noctua NH-L12Sx77714212835SE +/- 0.04, N = 3SE +/- 0.04, N = 329.2130.371. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

srsRAN Project

srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMbps, More Is BettersrsRAN Project 23.10.1-20240325Test: PUSCH Processor Benchmark, Throughput ThreadNoctua NH-L9a-AM5Noctua NH-L12Sx774080120160200SE +/- 0.00, N = 4SE +/- 0.02, N = 4183.4183.4MIN: 105.9MIN: 105.9 / MAX: 183.51. (CXX) g++ options: -O3 -march=native -mavx2 -mavx -msse4.1 -mfma -mavx512f -mavx512cd -mavx512bw -mavx512dq -fno-trapping-math -fno-math-errno -ldl

oneDNN

This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.4Harness: IP Shapes 1D - Engine: CPUNoctua NH-L9a-AM5Noctua NH-L12Sx770.59441.18881.78322.37762.972SE +/- 0.00064, N = 4SE +/- 0.00071, N = 42.641602.60748MIN: 2.57MIN: 2.561. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -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 4K - Video Preset: Ultra FastNoctua NH-L9a-AM5Noctua NH-L12Sx77918273645SE +/- 0.10, N = 4SE +/- 0.10, N = 439.6941.081. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

Primesieve

Primesieve generates prime numbers using a highly optimized sieve of Eratosthenes implementation. Primesieve primarily benchmarks the CPU's L1/L2 cache performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterPrimesieve 12.1Length: 1e12Noctua NH-L9a-AM5Noctua NH-L12Sx7748121620SE +/- 0.01, N = 3SE +/- 0.01, N = 417.3816.651. (CXX) g++ options: -O3

7-Zip Compression

This is a test of 7-Zip compression/decompression with its integrated benchmark feature. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 22.01Test: Decompression RatingNoctua NH-L9a-AM5Noctua NH-L12Sx7715K30K45K60K75KSE +/- 55.98, N = 3SE +/- 9.53, N = 363782683481. (CXX) g++ options: -lpthread -ldl -O2 -fPIC

OpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 22.01Test: Compression RatingNoctua NH-L9a-AM5Noctua NH-L12Sx7720K40K60K80K100KSE +/- 43.19, N = 3SE +/- 120.13, N = 379144813911. (CXX) g++ options: -lpthread -ldl -O2 -fPIC

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 1.11Video Input: Bosphorus 1080p - Video Preset: FasterNoctua NH-L9a-AM5Noctua NH-L12Sx77918273645SE +/- 0.12, N = 3SE +/- 0.03, N = 436.2237.881. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects

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: SlowNoctua NH-L9a-AM5Noctua NH-L12Sx771020304050SE +/- 0.17, N = 4SE +/- 0.07, N = 443.9945.661. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 8 - Input: Bosphorus 4KNoctua NH-L9a-AM5Noctua NH-L12Sx77918273645SE +/- 0.11, N = 3SE +/- 0.07, N = 336.2137.191. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

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: MediumNoctua NH-L9a-AM5Noctua NH-L12Sx771122334455SE +/- 0.08, N = 4SE +/- 0.04, N = 445.5347.131. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

AOM AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.9Encoder Mode: Speed 6 Two-Pass - Input: Bosphorus 1080pNoctua NH-L9a-AM5Noctua NH-L12Sx771530456075SE +/- 0.14, N = 4SE +/- 0.09, N = 563.0565.901. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 4 - Input: Bosphorus 1080pNoctua NH-L9a-AM5Noctua NH-L12Sx7748121620SE +/- 0.04, N = 5SE +/- 0.03, N = 516.3416.671. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

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, LosslessNoctua NH-L9a-AM5Noctua NH-L12Sx773691215SE +/- 0.03, N = 5SE +/- 0.01, N = 510.4410.041. (CXX) g++ options: -O3 -fPIC -lm

uvg266

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.8.0Video Input: Bosphorus 1080p - Video Preset: Very FastNoctua NH-L9a-AM5Noctua NH-L12Sx771326395265SE +/- 0.04, N = 5SE +/- 0.07, N = 558.1260.16

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 2.16.1Device: CPU - Batch Size: 1 - Model: ResNet-50Noctua NH-L9a-AM5Noctua NH-L12Sx7748121620SE +/- 0.02, N = 5SE +/- 0.01, N = 515.1515.38

Speedb

Speedb is a next-generation key value storage engine that is RocksDB compatible and aiming for stability, efficiency, and performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Sequential FillNoctua NH-L9a-AM5Noctua NH-L12Sx77200K400K600K800K1000KSE +/- 2511.52, N = 4SE +/- 2100.55, N = 497931710214211. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti

AOM AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.9Encoder Mode: Speed 11 Realtime - Input: Bosphorus 4KNoctua NH-L9a-AM5Noctua NH-L12Sx7720406080100SE +/- 0.45, N = 6SE +/- 0.66, N = 691.6992.321. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm

uvg266

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.8.0Video Input: Bosphorus 1080p - Video Preset: Super FastNoctua NH-L9a-AM5Noctua NH-L12Sx771428425670SE +/- 0.03, N = 5SE +/- 0.06, N = 560.7862.84

WebP Image Encode

OpenBenchmarking.orgMP/s, More Is BetterWebP Image Encode 1.4Encode Settings: Quality 100, LosslessNoctua NH-L9a-AM5Noctua NH-L12Sx770.46130.92261.38391.84522.3065SE +/- 0.00, N = 4SE +/- 0.01, N = 42.042.051. (CC) gcc options: -fvisibility=hidden -O2 -lm

oneDNN

This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.4Harness: IP Shapes 3D - Engine: CPUNoctua NH-L9a-AM5Noctua NH-L12Sx770.97141.94282.91423.88564.857SE +/- 0.00208, N = 5SE +/- 0.00089, N = 54.317484.23669MIN: 4.2MIN: 4.191. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

Parallel BZIP2 Compression

This test measures the time needed to compress a file (FreeBSD-13.0-RELEASE-amd64-memstick.img) using Parallel BZIP2 compression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterParallel BZIP2 Compression 1.1.13FreeBSD-13.0-RELEASE-amd64-memstick.img CompressionNoctua NH-L9a-AM5Noctua NH-L12Sx77246810SE +/- 0.020097, N = 6SE +/- 0.044895, N = 67.8875977.6315781. (CXX) g++ options: -O2 -pthread -lbz2 -lpthread

RNNoise

RNNoise is a recurrent neural network for audio noise reduction developed by Mozilla and Xiph.Org. This test profile is a single-threaded test measuring the time to denoise a sample 26 minute long 16-bit RAW audio file using this recurrent neural network noise suppression library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterRNNoise 0.2Input: 26 Minute Long Talking SampleNoctua NH-L9a-AM5Noctua NH-L12Sx77246810SE +/- 0.031, N = 6SE +/- 0.024, N = 67.7697.7281. (CC) gcc options: -O2 -pedantic -fvisibility=hidden

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 3.6Video Input: Bosphorus 1080pNoctua NH-L9a-AM5Noctua NH-L12Sx7720406080100SE +/- 0.28, N = 6SE +/- 0.58, N = 677.5879.651. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

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 2.16.1Device: CPU - Batch Size: 1 - Model: AlexNetNoctua NH-L9a-AM5Noctua NH-L12Sx7748121620SE +/- 0.01, N = 5SE +/- 0.01, N = 514.3614.45

oneDNN

This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.4Harness: Convolution Batch Shapes Auto - Engine: CPUNoctua NH-L9a-AM5Noctua NH-L12Sx77246810SE +/- 0.01217, N = 7SE +/- 0.06044, N = 78.252198.22246MIN: 7.92MIN: 7.91. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

srsRAN Project

srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMbps, More Is BettersrsRAN Project 23.10.1-20240325Test: PUSCH Processor Benchmark, Throughput TotalNoctua NH-L9a-AM5Noctua NH-L12Sx7730060090012001500SE +/- 0.29, N = 6SE +/- 0.16, N = 61205.71206.5MIN: 763.5 / MAX: 1206.5MIN: 763.6 / MAX: 12071. (CXX) g++ options: -O3 -march=native -mavx2 -mavx -msse4.1 -mfma -mavx512f -mavx512cd -mavx512bw -mavx512dq -fno-trapping-math -fno-math-errno -ldl

uvg266

OpenBenchmarking.orgFrames Per Second, More Is Betteruvg266 0.8.0Video Input: Bosphorus 1080p - Video Preset: Ultra FastNoctua NH-L9a-AM5Noctua NH-L12Sx771632486480SE +/- 0.10, N = 5SE +/- 0.06, N = 569.1871.42

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 FastNoctua NH-L9a-AM5Noctua NH-L12Sx7720406080100SE +/- 0.09, N = 7SE +/- 0.04, N = 798.06100.941. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 12 - Input: Bosphorus 4KNoctua NH-L9a-AM5Noctua NH-L12Sx7720406080100SE +/- 0.14, N = 6SE +/- 0.09, N = 696.7599.881. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

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: 6Noctua NH-L9a-AM5Noctua NH-L12Sx77246810SE +/- 0.013, N = 6SE +/- 0.007, N = 67.1116.8441. (CXX) g++ options: -O3 -fPIC -lm

Google Draco

Draco is a library developed by Google for compressing/decompressing 3D geometric meshes and point clouds. This test profile uses some Artec3D PLY models as the sample 3D model input formats for Draco compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterGoogle Draco 1.5.6Model: Church FacadeNoctua NH-L9a-AM5Noctua NH-L12Sx7711002200330044005500SE +/- 8.39, N = 6SE +/- 11.53, N = 7512350001. (CXX) g++ options: -O3

srsRAN Project

srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMbps, More Is BettersrsRAN Project 23.10.1-20240325Test: PDSCH Processor Benchmark, Throughput TotalNoctua NH-L9a-AM5Noctua NH-L12Sx7712002400360048006000SE +/- 7.44, N = 7SE +/- 5.04, N = 75443.45500.61. (CXX) g++ options: -O3 -march=native -mavx2 -mavx -msse4.1 -mfma -mavx512f -mavx512cd -mavx512bw -mavx512dq -fno-trapping-math -fno-math-errno -ldl

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 8 - Input: Bosphorus 1080pNoctua NH-L9a-AM5Noctua NH-L12Sx77306090120150SE +/- 0.13, N = 7SE +/- 0.14, N = 7117.54120.331. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

AOM AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.9Encoder Mode: Speed 8 Realtime - Input: Bosphorus 1080pNoctua NH-L9a-AM5Noctua NH-L12Sx7760120180240300SE +/- 3.26, N = 15SE +/- 1.78, N = 15253.33259.221. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.9Encoder Mode: Speed 11 Realtime - Input: Bosphorus 1080pNoctua NH-L9a-AM5Noctua NH-L12Sx7760120180240300SE +/- 7.60, N = 15SE +/- 7.12, N = 15270.70275.241. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.9Encoder Mode: Speed 9 Realtime - Input: Bosphorus 1080pNoctua NH-L12Sx77Noctua NH-L9a-AM560120180240300SE +/- 7.27, N = 15SE +/- 6.57, N = 15263.51274.231. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm

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 FastNoctua NH-L9a-AM5Noctua NH-L12Sx77306090120150SE +/- 0.09, N = 7SE +/- 0.06, N = 8121.49124.241. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

Google Draco

Draco is a library developed by Google for compressing/decompressing 3D geometric meshes and point clouds. This test profile uses some Artec3D PLY models as the sample 3D model input formats for Draco compression/decompression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterGoogle Draco 1.5.6Model: LionNoctua NH-L9a-AM5Noctua NH-L12Sx778001600240032004000SE +/- 6.58, N = 7SE +/- 5.14, N = 7392038601. (CXX) g++ options: -O3

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, LosslessNoctua NH-L9a-AM5Noctua NH-L12Sx771.11472.22943.34414.45885.5735SE +/- 0.009, N = 7SE +/- 0.010, N = 84.9544.8041. (CXX) g++ options: -O3 -fPIC -lm

AOM AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.9Encoder Mode: Speed 10 Realtime - Input: Bosphorus 1080pNoctua NH-L9a-AM5Noctua NH-L12Sx7760120180240300SE +/- 9.00, N = 15SE +/- 8.00, N = 15270.01275.901. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm

WebP Image Encode

OpenBenchmarking.orgMP/s, More Is BetterWebP Image Encode 1.4Encode Settings: Quality 100, Highest CompressionNoctua NH-L9a-AM5Noctua NH-L12Sx771.12052.2413.36154.4825.6025SE +/- 0.00, N = 7SE +/- 0.00, N = 74.974.981. (CC) gcc options: -fvisibility=hidden -O2 -lm

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 2.16.1Device: CPU - Batch Size: 1 - Model: GoogLeNetNoctua NH-L12Sx77Noctua NH-L9a-AM51122334455SE +/- 0.09, N = 8SE +/- 0.15, N = 850.2150.42

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 FastNoctua NH-L9a-AM5Noctua NH-L12Sx774080120160200SE +/- 0.11, N = 8SE +/- 0.13, N = 9160.11163.471. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt

oneDNN

This is a test of the Intel oneDNN as an Intel-optimized library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Intel oneDNN was formerly known as DNNL (Deep Neural Network Library) and MKL-DNN before being rebranded as part of the Intel oneAPI toolkit. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetteroneDNN 3.4Harness: Deconvolution Batch shapes_3d - Engine: CPUNoctua NH-L9a-AM5Noctua NH-L12Sx771.29382.58763.88145.17526.469SE +/- 0.00259, N = 9SE +/- 0.00143, N = 95.750305.74921MIN: 5.53MIN: 5.541. (CXX) g++ options: -O3 -march=native -fopenmp -msse4.1 -fPIC -pie -ldl

AOM AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterAOM AV1 3.9Encoder Mode: Speed 6 Realtime - Input: Bosphorus 1080pNoctua NH-L9a-AM5Noctua NH-L12Sx7750100150200250SE +/- 0.82, N = 10SE +/- 1.72, N = 10239.80240.331. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm

srsRAN Project

srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMbps, More Is BettersrsRAN Project 23.10.1-20240325Test: PDSCH Processor Benchmark, Throughput ThreadNoctua NH-L9a-AM5Noctua NH-L12Sx772004006008001000SE +/- 2.68, N = 9SE +/- 4.03, N = 9900.3902.11. (CXX) g++ options: -O3 -march=native -mavx2 -mavx -msse4.1 -mfma -mavx512f -mavx512cd -mavx512bw -mavx512dq -fno-trapping-math -fno-math-errno -ldl

WebP Image Encode

OpenBenchmarking.orgMP/s, More Is BetterWebP Image Encode 1.4Encode Settings: Quality 100Noctua NH-L9a-AM5Noctua NH-L12Sx7748121620SE +/- 0.06, N = 15SE +/- 0.01, N = 1114.8414.861. (CC) gcc options: -fvisibility=hidden -O2 -lm

SVT-AV1

This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 12 - Input: Bosphorus 1080pNoctua NH-L9a-AM5Noctua NH-L12Sx7790180270360450SE +/- 5.23, N = 15SE +/- 0.64, N = 11414.60429.911. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 13 - Input: Bosphorus 1080pNoctua NH-L12Sx77Noctua NH-L9a-AM5110220330440550SE +/- 11.01, N = 15SE +/- 1.86, N = 15501.53503.421. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

WebP Image Encode

OpenBenchmarking.orgMP/s, More Is BetterWebP Image Encode 1.4Encode Settings: DefaultNoctua NH-L12Sx77Noctua NH-L9a-AM5612182430SE +/- 0.01, N = 12SE +/- 0.04, N = 1223.1323.221. (CC) gcc options: -fvisibility=hidden -O2 -lm

CPU Temperature Monitor

OpenBenchmarking.orgCelsiusCPU Temperature MonitorPhoronix Test Suite System MonitoringNoctua NH-L9a-AM5Noctua NH-L12Sx7720406080100Min: 43.25 / Avg: 87.58 / Max: 96.75Min: 38.25 / Avg: 72.38 / Max: 88.25

CPU Power Consumption Monitor

OpenBenchmarking.orgWattsCPU Power Consumption MonitorPhoronix Test Suite System MonitoringNoctua NH-L12Sx77Noctua NH-L9a-AM520406080100Min: 6.49 / Avg: 68.8 / Max: 111.71Min: 7.84 / Avg: 63.43 / Max: 92.23

CPU Peak Freq (Highest CPU Core Frequency) Monitor

OpenBenchmarking.orgMegahertzCPU Peak Freq (Highest CPU Core Frequency) MonitorPhoronix Test Suite System MonitoringNoctua NH-L12Sx77Noctua NH-L9a-AM59001800270036004500Min: 545 / Avg: 4768.03 / Max: 5159Min: 545 / Avg: 4736.22 / Max: 5159

Geometric Mean Of All Test Results

OpenBenchmarking.orgGeometric Mean, More Is BetterGeometric Mean Of All Test ResultsResult Composite - Noctua NH-L12Sx77 HeatsinkNoctua NH-L9a-AM5Noctua NH-L12Sx77306090120150128.01131.70

235 Results Shown

Timed Linux Kernel Compilation
Blender
Timed Node.js Compilation
Xmrig
OSPRay Studio
Quicksilver
Blender
Timed Godot Game Engine Compilation
OSPRay Studio:
  2 - 4K - 32 - Path Tracer - CPU
  1 - 4K - 32 - Path Tracer - CPU
Blender
OpenSSL:
  RSA4096:
    verify/s
    sign/s
C-Ray
TensorFlow
ASTC Encoder:
  Very Thorough
  Exhaustive
OSPRay Studio
BRL-CAD
Primesieve
Llamafile
SPECFEM3D
OSPRay Studio
FFmpeg
OSPRay Studio
Quicksilver
AOM AV1
Blender
Speedb
OpenSSL:
  AES-256-GCM
  ChaCha20-Poly1305
  ChaCha20
  AES-128-GCM
  SHA512
  SHA256
OSPRay
Blender
SPECFEM3D
OSPRay
GROMACS
FFmpeg:
  libx265 - Platform
  libx265 - Video On Demand
TensorFlow
Blender
FFmpeg
libavif avifenc
OSPRay
VVenC
FFmpeg:
  libx264 - Platform
  libx264 - Video On Demand
Timed Linux Kernel Compilation
NAMD
CacheBench:
  Read / Modify / Write
  Write
  Read
uvg266
Llamafile
OSPRay Studio
uvg266
High Performance Conjugate Gradient
OSPRay Studio:
  2 - 1080p - 32 - Path Tracer - CPU
  1 - 1080p - 32 - Path Tracer - CPU
AOM AV1
TensorFlow
OSPRay Studio
Quicksilver
Stockfish
QuantLib
Intel Open Image Denoise
oneDNN
OSPRay
Speedb
SPECFEM3D
OSPRay Studio
oneDNN
OSPRay Studio
Liquid-DSP:
  1 - 256 - 512
  4 - 256 - 57
OSPRay Studio
C-Ray
Chaos Group V-RAY
Embree
Liquid-DSP
Embree
ASTC Encoder
OSPRay Studio
Kvazaar
libavif avifenc
Timed PHP Compilation
SPECFEM3D:
  Mount St. Helens
  Tomographic Model
Timed FFmpeg Compilation
Kvazaar
Embree
GraphicsMagick:
  Sharpen
  Enhanced
JPEG-XL libjxl
Embree
GraphicsMagick
JPEG-XL libjxl
Speedb
GraphicsMagick:
  Swirl
  Rotate
Speedb
GraphicsMagick
RocksDB:
  Overwrite
  Read While Writing
Speedb
GraphicsMagick
Speedb
RocksDB:
  Update Rand
  Read Rand Write Rand
  Rand Read
OSPRay Studio:
  2 - 1080p - 16 - Path Tracer - CPU
  2 - 4K - 1 - Path Tracer - CPU
Embree
OSPRay Studio
VVenC
OSPRay Studio
Embree
OSPRay:
  gravity_spheres_volume/dim_512/ao/real_time
  gravity_spheres_volume/dim_512/scivis/real_time
Liquid-DSP
TensorFlow:
  CPU - 32 - GoogLeNet
  CPU - 64 - AlexNet
FFmpeg
uvg266:
  Bosphorus 4K - Very Fast
  Bosphorus 4K - Super Fast
NAMD
AOM AV1:
  Speed 6 Realtime - Bosphorus 4K
  Speed 6 Two-Pass - Bosphorus 4K
JPEG-XL libjxl:
  PNG - 90
  JPEG - 90
VVenC
JPEG-XL libjxl:
  JPEG - 100
  PNG - 100
uvg266
QuantLib
Timed Mesa Compilation
SVT-AV1:
  Preset 4 - Bosphorus 4K
  Preset 13 - Bosphorus 4K
x265
AOM AV1
Y-Cruncher
JPEG-XL Decoding libjxl
AOM AV1
Llamafile
FFmpeg
AOM AV1
ASTC Encoder
WebP Image Encode
Liquid-DSP:
  12 - 256 - 512
  2 - 256 - 512
  2 - 256 - 57
  8 - 256 - 512
  4 - 256 - 512
  1 - 256 - 32
  2 - 256 - 32
Stress-NG
Liquid-DSP:
  12 - 256 - 57
  12 - 256 - 32
  4 - 256 - 32
  8 - 256 - 32
TensorFlow
uvg266
AOM AV1:
  Speed 9 Realtime - Bosphorus 4K
  Speed 10 Realtime - Bosphorus 4K
Kvazaar
LZ4 Compression:
  9 - Decompression Speed
  9 - Compression Speed
uvg266
PyTorch
LZ4 Compression:
  3 - Decompression Speed
  3 - Compression Speed
  1 - Decompression Speed
  1 - Compression Speed
JPEG-XL Decoding libjxl
oneDNN
Y-Cruncher
ASTC Encoder
Kvazaar
srsRAN Project
oneDNN
Kvazaar
Primesieve
7-Zip Compression:
  Decompression Rating
  Compression Rating
VVenC
Kvazaar
SVT-AV1
Kvazaar
AOM AV1
SVT-AV1
libavif avifenc
uvg266
TensorFlow
Speedb
AOM AV1
uvg266
WebP Image Encode
oneDNN
Parallel BZIP2 Compression
RNNoise
x265
TensorFlow
oneDNN
srsRAN Project
uvg266
Kvazaar
SVT-AV1
libavif avifenc
Google Draco
srsRAN Project
SVT-AV1
AOM AV1:
  Speed 8 Realtime - Bosphorus 1080p
  Speed 11 Realtime - Bosphorus 1080p
  Speed 9 Realtime - Bosphorus 1080p
Kvazaar
Google Draco
libavif avifenc
AOM AV1
WebP Image Encode
TensorFlow
Kvazaar
oneDNN
AOM AV1
srsRAN Project
WebP Image Encode
SVT-AV1:
  Preset 12 - Bosphorus 1080p
  Preset 13 - Bosphorus 1080p
WebP Image Encode
CPU Temperature Monitor:
  Phoronix Test Suite System Monitoring:
    Celsius
    Watts
    Megahertz
  Result Composite - Noctua NH-L12Sx77 Heatsink:
    Geometric Mean