Big Threadripper

AMD Ryzen Threadripper 3970X 32-Core testing with a ASUS ROG ZENITH II EXTREME (0702 BIOS) and llvmpipe 63GB on Ubuntu 19.10 via the Phoronix Test Suite.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 2003108-PTS-BIGTHREA46
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

Limit displaying results to tests within:

Audio Encoding 3 Tests
AV1 4 Tests
Bioinformatics 4 Tests
BLAS (Basic Linear Algebra Sub-Routine) Tests 2 Tests
C++ Boost Tests 4 Tests
Chess Test Suite 7 Tests
Timed Code Compilation 10 Tests
C/C++ Compiler Tests 48 Tests
Compression Tests 6 Tests
CPU Massive 98 Tests
Creator Workloads 45 Tests
Cryptography 9 Tests
Database Test Suite 11 Tests
Disk Test Suite 7 Tests
Encoding 13 Tests
Fortran Tests 6 Tests
Game Development 3 Tests
Go Language Tests 2 Tests
HPC - High Performance Computing 35 Tests
Imaging 9 Tests
Java 5 Tests
Common Kernel Benchmarks 19 Tests
LAPACK (Linear Algebra Pack) Tests 3 Tests
Linear Algebra 4 Tests
Machine Learning 9 Tests
Memory Test Suite 6 Tests
Molecular Dynamics 8 Tests
MPI Benchmarks 11 Tests
Multi-Core 71 Tests
Networking Test Suite 3 Tests
Node.js + NPM Tests 2 Tests
NVIDIA GPU Compute 10 Tests
Intel oneAPI 4 Tests
OpenCL 3 Tests
OpenMPI Tests 15 Tests
Productivity 6 Tests
Programmer / Developer System Benchmarks 19 Tests
Python 7 Tests
Raytracing 5 Tests
Renderers 15 Tests
Rust Tests 2 Tests
Scientific Computing 20 Tests
Server 21 Tests
Server CPU Tests 51 Tests
Single-Threaded 46 Tests
Speech 2 Tests
Telephony 2 Tests
Video Encoding 10 Tests
Common Workstation Benchmarks 9 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
Disable Color Branding
Prefer Vertical Bar Graphs

Additional Graphs

Show Perf Per Core/Thread Calculation Graphs Where Applicable

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
Threadripper 3970X
March 06 2020
  2 Days, 1 Minute
16 Cores
March 08 2020
  1 Day, 14 Hours, 14 Minutes
Invert Hiding All Results Option
  1 Day, 19 Hours, 7 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):


Big ThreadripperProcessorMotherboardChipsetMemoryDiskGraphicsAudioNetworkOSKernelDesktopDisplay ServerDisplay DriverOpenGLCompilerFile-SystemScreen ResolutionThreadripper 3970X16 CoresAMD Ryzen Threadripper 3970X 32-Core @ 3.70GHz (32 Cores / 64 Threads)ASUS ROG ZENITH II EXTREME (0702 BIOS)AMD Starship/Matisse64GB1000GB Force MP600llvmpipe 63GBAMD Navi 10 HDMI AudioAquantia AQC107 NBase-T/IEEE + Intel I211 + Intel Wi-Fi 6 AX200Ubuntu 19.105.3.0-18-generic (x86_64)GNOME Shell 3.34.1X Server 1.20.5modesetting 1.20.53.3 Mesa 19.2.8 (LLVM 9.0 128 bits)GCC 9.2.1 20191008 + Clang 9.0.0-2ext43840x2160AMD Ryzen Threadripper 3970X 32-Core @ 3.70GHz (16 Cores)5.3.0-40-generic (x86_64)OpenBenchmarking.orgCompiler Details- --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,gm2 --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-offload-targets=nvptx-none,hsa --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 Disk Details- NONE / errors=remount-ro,relatime,rwProcessor Details- Scaling Governor: acpi-cpufreq ondemand - CPU Microcode: 0x8301025Java Details- OpenJDK Runtime Environment (build 11.0.6+10-post-Ubuntu-1ubuntu119.10.1)Python Details- Python 2.7.17 + Python 3.7.5Security Details- Threadripper 3970X: l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional STIBP: conditional RSB filling - 16 Cores: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional RSB filling + tsx_async_abort: Not affected

Threadripper 3970X vs. 16 Cores ComparisonPhoronix Test SuiteBaseline+920.9%+920.9%+1841.8%+1841.8%+2762.7%+2762.7%542.7%506%502%492.2%421.7%393.5%376.2%360.7%328.2%240.6%181.6%148.4%146.3%110.9%110.6%110%107.7%104.7%92.8%91.3%85.5%85.4%83.4%82.6%81.1%80.1%79.7%78.3%72.4%71.9%67.2%66.4%62.2%61.7%60.5%59.4%55.2%52.6%51.4%50.7%50.4%49.8%49.8%49.1%47.6%47.5%47.2%46%45.7%45.4%44.9%40.1%38.3%36.5%33.4%33.1%33%32.6%31.6%27.8%26.4%24%23.7%23.5%22.9%22.8%22.3%19.4%18.8%18.3%17.6%16.9%16.7%16.4%16.1%15.9%15.5%15.4%15.4%15.4%15.3%15.1%15%15%14.6%14.1%13.8%13.8%13%12.9%12.8%12.2%11.2%10.4%10.1%8.9%8.7%8.5%8.2%7.8%7.7%7.6%7.3%7.3%7.2%7.1%7.1%6.8%6.7%6.1%5.9%5.8%5.6%5.5%5.3%5.1%4.7%4.7%4.3%4%4%3.9%3.9%3.8%3.5%3.5%3.5%3.3%2.9%2.9%2.8%2.6%2.6%2.5%2.5%2.5%2.3%Context Switching841.5%Forking784.4%Reads543.1%Rand DeleteSeq FillOverwriteRand Fill64Hot ReadRand ReadLaunch Programs363.6%Seek RandStatic OMP SpeedupWrite326.3%deep326.2%Create Threads307.1%Writes3683.6%Fill Sync260%G-Ptrans250%Write TestNon-Exponential237.6%Elapsed Time223.8%Rand Fill Sync220.3%Lsearch214.5%Time To Solve207%Matrix Math204.2%Buffer Test - Normal Load - Read Only190.9%Tsearch190%Pathtracer - Crown184.7%LoaderBsearch179.3%Pathtracer ISPC - Crown177.1%NASA Streamlines - Path Tracer173.2%G.C.S.F172%XFrog Forest - SciVis170.8%San Miguel - SciVis170%G.Q.D.S169.8%Barbershop - CPU-Only169.4%Pabellon Barcelona - CPU-Only169.3%XFrog Forest - Path Tracer168%San Miguel - Path Tracer167.9%NASA Streamlines - SciVis166.7%Bedroom164.9%Classroom - CPU-Only163.5%CoreMark Size 666 - I.P.S161.9%Pabellon Barcelona - CPU-Only161.8%Read While Writing160.5%Classroom - CPU-Only159.5%Total Time159.2%Barbershop - CPU-Only158.1%1.H.M.2.D156.4%G.I.R.1.S155.9%BMW27 - CPU-Only151.4%Vector Math150.7%Fishy Cat - CPU-Only150.4%C.S.T150%Rand Read148.2%m7m148.1%M.R - SciVis147.9%12Trace Time143.6%Fishy Cat - CPU-Only143.3%CPU Stress142.1%BMW27 - CPU-Only141.8%V.P.M141.5%Volumetric Caustic140.3%Hsearch138.1%32 - Process136.9%Hair135.3%Large Scene135.3%Total Time133%S.F.P.R132.9%Supercar126.9%json125.3%CPU124.3%lbry118.8%Time To Compile116.6%V.Q.O - Bosphorus 1080p114%OpenMP CUTCP113.6%1.8.b.Y.T.H.V.E112.8%Socket Activity111.4%leblancbigR.N.N.T - f32Memory Copying110%Crypto109.6%skein109.3%sha256t109.1%Emily108.9%Disney Material108.6%Total Time - 4.1.R.P.P108.6%CPU107.9%12107.3%1.P.N.G104.8%sedovbigR.4.b.P104.5%Swirl102.8%Time To Compile101.1%8 - 81191.2%ATPase Simulation - 327,506 Atoms98.7%OpenMP LavaMD98.3%97.9%Create Processes96.6%EP.D96.4%Enc Mode 8 - 1080p94.3%P.N.T.T.2.0.094.3%httpPrependEP.C90.5%IMB-P2P PingPong90%Time To Compile89.2%Enhanced87.4%Sharpen85.7%Rand Write - IO_uring - No - Yes - 4KBRand Write - IO_uring - No - Yes - 4KBG.I.I.S84.5%TradebeansEP-STREAM TriadO.SR.R.BNoise-Gaussian79.8%HTTP - Bandwidth - 2Rand Fill78.5%ReplaceM.R - Path Tracer77.8%Time To Compile73%BLASAppendSetBLAS CPU66.6%EP-DGEMMTime To Compile66.3%Time To Compile66.1%Blowfish65.7%Seq Fill63.3%O.M.GAddWater Benchmark60.5%S.V.M.P60.5%21C.u.1.0.3.s.i.i.C.L.158.8%58.3%Summer Nature 4K58%H.2.V.E55.7%scikit_qdaMD555%Speed 8 Realtimegarbage52.3%Seq FillG-Rand Access50.8%Set - 64Time To Compile50.5%OverwriteInterpreterIP Batch All - f32H2Time To Compile48.2%Rand FillAdd - 64OpenMP Stencil47.3%I.M.D.SMemorial46.7%A.U.C.TOpenMP LBMH.2.H.T.N.DTime To Compile45%D.B.d - f32G-Ffte44.4%tConvolve OpenMP - Gridding43.2%IMB-MPI1 Exchange42.2%C.B.c - f3241.4%C.B.c - f3239.2%SP.B39.2%S.N.1scikit_icaBT.C36.5%Rhodopsin Protein35.5%1.F.1.S.N.S.F35.4%C.B.c - f3234.9%Material Tester34.3%unsharp-maskD.B.d - f32auto-levelsLatency Ping PongP.D.SInitial Create30.5%S.N.130.4%IMB-MPI1 Sendrecv30.1%C.B.c - f3229.7%Water Caustic29.1%SVG Files To PNGC.u.1.0.3.s.i.i.C.L.926.7%IMB-MPI1 SendrecvSeq Read - POSIX AIO - Yes - No - 4KB26.1%T.T.C.S.P.M25.2%C.S.T25%Rand Read - POSIX AIO - Yes - No - 4KB24.3%4Kb - 512MB - Write Performance24.2%Rand Read - POSIX AIO - Yes - No - 4KB24.1%5.F.1.S.4.TServer Rack - CPU-onlyOpenMP CFD Solver23.4%64Kb - 512MB - Write Performance23.2%G-HPLD.T.P22.9%Latency Under LoadCPU22.5%resizeResizing21.4%Masskrug - CPU-only21.3%Create Files21.2%LU.C21.2%L.E.H20%4Kb - 512MB - Write Performance19.4%DeleteBoat - CPU-onlySemaphoresApache Spark Bayes17.9%SmallScale - Integer17.4%Savina Reactors.IO17.3%Speed 5Fayalite-FIST Data1MB - 512MB - Write Performance16.4%build5201 - 10 Seconds - TCP - 6416.1%R.C.a.PD.B.d - f3215.7%Fill SyncCPUSeq Read - IO_uring - No - Yes - 4KBSeq Read - IO_uring - No - Yes - 4KBIMB-MPI1 ExchangeR.R.LGetBuffer Test - Normal Load - Read WritetConvolve MT - Gridding15%Windowed Gaussian14.8%ScaleIMB-MPI1 PingPongL.S.T.A.T.RRand Write - 16tConvolve MPI - Gridding13.3%S.W.P.S4.F.3.S.D.1.SH.2.1.V.ERand Write - 1612.5%Get - 64rotateFT.C10.5%myr-grMemory10.4%Apache Spark ALSSeq Write - IO_uring - No - Yes - 4KB9.5%Seq Write - IO_uring - No - Yes - 4KB9.5%DLSC9.1%Earthgecko Skyline9.1%9.1%GETF.F.T8.6%P.P.ATile GlassCifar108.1%tConvolve OpenMP - Degridding8.1%Kernel8.1%P.R.W.S.S.M8%tConvolve MT - Degridding7.9%Canvas Text SimplePod2html7.4%Rand Read - IO_uring - No - Yes - 4KBAntialiasRand Read - IO_uring - No - Yes - 4KBRelative EntropytConvolve MPI - Degridding7.1%C.G.C7%CropRead Test1.F.1.SP.P.SCopy5.8%HWB Color Space5.8%U.U.D.S - 128GET5.7%RotateTotal TimeTotal5.4%F.E.T5.4%LPUSHDhrystone 25.2%Rand Write - POSIX AIO - Yes - No - 4KBMG.C5%Userspace4.9%C.u.1.0.3.s.i.i.C.L.1Memory AllocationsTriad4.5%Scale4.4%Seq Write - POSIX AIO - Yes - No - 4KBUnnamed Pipe - 128SVG Files To PNGD.L.M.F4%T.B.T3.9%No - Inference - ResNet 50 - CPUIP Batch 1D - f323.9%IS.DT.T.T.1.V.L3.9%Add3.8%Wavelet BlurLPOP3.7%1TCP Socket - 128scikit_linearridgeregressionF.F.T3.5%Redis3.5%Server Room - CPU-only3.3%floatSETComposite2.9%Copy - Integer2502.7%JythonCartoonS.M.M2.6%TradesoapM.P.P.BC.C.4.S2.3%Monte Carlo2.2%D.L.M.F2.2%Stepanov Vector2.2%Composite2.2%T.T.C.C.G.C2.1%T.T.S.S2%Stress-NGStress-NGApache CassandraLevelDBLevelDBLevelDBLevelDBMariaDBLevelDBLevelDBOSBenchLevelDBCLOMPBlogBenchCpuminer-OptOSBenchApache CassandraLevelDBHPC ChallengeIORTungsten RendererN-QueensFacebook RocksDBStress-NGm-queensStress-NGPostgreSQL pgbenchStress-NGEmbreeSystemd Total Boot TimeStress-NGEmbreeOSPrayStress-NGOSPrayOSPrayStress-NGBlenderBlenderOSPrayOSPrayOSPrayIndigoBenchBlenderCoremarkBlenderFacebook RocksDBBlenderStockfishBlenderasmFishSmallptBlenderStress-NGBlender7-Zip CompressionQMCPACKFacebook RocksDBCpuminer-OptOSPrayDbenchPOV-RayBlenderStress-NGBlenderBRL-CADTungsten RendererStress-NGHackbenchTungsten Rendererrays1benchTachyonACES DGEMMIndigoBenchGo BenchmarksChaos Group V-RAYCpuminer-OptTimed MPlayer CompilationSVT-VP9ParboilSVT-HEVCStress-NGPennantMKL-DNN DNNLStress-NGR BenchmarkStress-NGCpuminer-OptCpuminer-OptAppleseedAppleseedC-RaySysbenchDbenchAircrack-ngPrimesievePennantOpenSSLGraphicsMagickTimed Linux Kernel CompilationSchbenchNAMDRodiniaebizzyOSBenchNAS Parallel BenchmarksSVT-AV1Rust Prime BenchmarkGo BenchmarksMemcached mcperfNAS Parallel BenchmarksIntel MPI BenchmarksTimed LLVM CompilationGraphicsMagickGraphicsMagickFlexible IO TesterFlexible IO TesterSunflow Rendering SystemDaCapo BenchmarkHPC ChallengeRodiniaHPC ChallengeGraphicsMagickEthrFacebook RocksDBMemcached mcperfOSPrayTimed FFmpeg CompilationLeelaChessZeroMemcached mcperfMemcached mcperfArrayFireHPC ChallengeTimed GCC CompilationTimed ImageMagick CompilationJohn The RipperFacebook RocksDBParboilMemcached mcperfGROMACSStress-NGt-test1t-test1Zstd CompressionXsbenchdav1dx264Mlpack BenchmarkJohn The RipperAOM AV1Go BenchmarksLevelDBHPC ChallengeMemcached mcperfTimed PHP CompilationLevelDBPerl BenchmarksMKL-DNN DNNLDaCapo BenchmarkTimed Apache CompilationLevelDBMemcached mcperfParboilRenaissanceIntel Open Image DenoiseRenaissanceParboilFFmpegBuild2MKL-DNN DNNLHPC ChallengeASKAPIntel MPI BenchmarksMKL-DNN DNNLKeyDBMKL-DNN DNNLNAS Parallel Benchmarkslibgav1Mlpack BenchmarkNAS Parallel BenchmarksLAMMPS Molecular Dynamics SimulatorFS-MarkMKL-DNN DNNLAppleseedGIMPMKL-DNN DNNLGIMPSockperfTimed HMMer SearchCompile Benchdav1dIntel MPI BenchmarksMKL-DNN DNNLTungsten RendererlibrsvgXZ CompressionIntel MPI BenchmarksFlexible IO TesterRust Mandelbrotctx_clockFlexible IO TesterIOzoneFlexible IO TesterFS-MarkDarktableGNU Octave BenchmarkRodiniaIOzoneHPC ChallengePostMarkSockperfNeatBenchGIMPGraphicsMagickDarktableOSBenchNAS Parallel BenchmarksCloverLeafIOzoneMemcached mcperfDarktableStress-NGRenaissanceminiFERAMspeed SMPRenaissanceVP9 libvpx EncodingCP2K Molecular DynamicsIOzoneGo BenchmarksiPerfLULESHLuxCoreRenderMKL-DNN DNNLLevelDBDeepSpeechFlexible IO TesterFlexible IO TesterIntel MPI BenchmarksHPC ChallengeMemcached mcperfPostgreSQL pgbenchASKAPNumenta Anomaly BenchmarkGEGLIntel MPI BenchmarksRAR CompressionApache HBaseASKAPApache BenchmarkFS-Markx265Apache HBaseMemcached mcperfGIMPNAS Parallel BenchmarksCpuminer-OptSysbenchRenaissanceFlexible IO TesterFlexible IO TesterLuxCoreRenderNumenta Anomaly BenchmarkNebular Empirical Analysis ToolRedisNode.js Express HTTP Load TestSciMarkTimed MrBayes AnalysisGEGLTensorflowASKAPSystemd Total Boot TimeTTSIOD 3D RendererASKAPOpenCV BenchmarkQmlbenchAlgebraic Multi-Grid BenchmarkPerl BenchmarksFlexible IO TesterGEGLFlexible IO TesterNumenta Anomaly BenchmarkASKAPArrayFireHigh Performance Conjugate GradientGEGLIORFS-MarkHimeno BenchmarkStreamGraphicsMagickIPC_benchmarkRedisGraphicsMagickSudokutSystemd Total Boot TimeBork File EncrypterRedisBYTE Unix BenchmarkFlexible IO TesterNAS Parallel BenchmarksSystemd Total Boot TimeZstd CompressionOSBenchStreamStreamFlexible IO TesterIPC_benchmarkInkscapeSciMarkRawTherapeePlaidMLMKL-DNN DNNLNAS Parallel BenchmarksLoopback TCP Network PerformanceStreamGEGLRedisSQLiteIPC_benchmarkMlpack BenchmarkLuaJITMemtier_benchmarkDarktablePyPerformanceRedisNode.js Octane BenchmarkSciMarkRAMspeed SMPApache SiegeDaCapo BenchmarkGEGLLuaJITDaCapo BenchmarkHPC ChallengeFhourstonesCython benchmarkBLAKE2LuaJITLuaJITCppPerformanceBenchmarksLuaJITGiteSpeak Speech EngineThreadripper 3970X16 Cores

Big Threadripperpennant: sedovbigpennant: leblancbighpcc: G-HPLdbench: 12dbench: 12cpp-perf-bench: Rand Numbersmcperf: Set - 64mcperf: Add - 64build-gcc: Time To Compilemcperf: Get - 64blogbench: Readmysqlslap: 64mkl-dnn: Convolution Batch conv_all - f32qmcpack: pgbench: Buffer Test - Normal Load - Read Writemkl-dnn: Deconvolution Batch deconv_all - f32amg: cryptopp: Keyed Algorithmsblender: Barbershop - CPU-Onlyfftw: Float + SSE - 2D FFT Size 4096blender: Barbershop - CPU-Onlyplaidml: No - Inference - ResNet 50 - CPUblender: Pabellon Barcelona - CPU-Onlyhint: FLOATdacapobench: Tradesoapcpp-perf-bench: Math Librarycassandra: Readsbrl-cad: VGR Performance Metricblender: Pabellon Barcelona - CPU-Onlycompress-lzma: 256MB File Compressionblender: Classroom - CPU-Onlykeydb: tinymembench: Standard Memsettinymembench: Standard Memcpyapache-siege: 250blender: Classroom - CPU-Onlynumpy: radiance: Serialrocksdb: Rand Fill Syncluxcorerender: DLSCluxcorerender: Rainbow Colors and Prismhpcg: fftw: Stock - 2D FFT Size 4096compress-zstd: Compressing ubuntu-16.04.3-server-i386.img, Compression Level 19fs-mark: 5000 Files, 1MB Size, 4 Threadsospray: XFrog Forest - Path Tracerfio: Seq Read - IO_uring - No - Yes - 4KB - Default Test Directoryfio: Seq Read - IO_uring - No - Yes - 4KB - Default Test Directorycp2k: Fayalite-FIST Dataappleseed: Emilyospray: San Miguel - Path Tracergmpbench: Total Timemkl-dnn: Convolution Batch conv_googlenet_v3 - f32mlpack: scikit_linearridgeregressionblender: Fishy Cat - CPU-Onlycachebench: Writecachebench: Readcachebench: Read / Modify / Writebyte: Dhrystone 2asmfish: 1024 Hash Memory, 26 Depthblender: Fishy Cat - CPU-Onlynamd: ATPase Simulation - 327,506 Atomsfhourstones: Complex Connect-4 Solvingcassandra: Writesminion: Quasigroupleveldb: Hot Readappleseed: Material Testerrocksdb: Read While Writingmcperf: Appendfio: Seq Write - IO_uring - No - Yes - 4KB - Default Test Directoryfio: Seq Write - IO_uring - No - Yes - 4KB - Default Test Directoryhbase: Rand Write - 16hbase: Rand Write - 16mlpack: scikit_qdacpuminer-opt: lbryfio: Rand Read - IO_uring - No - Yes - 4KB - Default Test Directoryfio: Rand Read - IO_uring - No - Yes - 4KB - Default Test Directorystress-ng: Semaphoresstress-ng: Lsearchstress-ng: Context Switchingstress-ng: System V Message Passingtensorflow: Cifar10aobench: 2048 x 2048 - Total Timegegl: Cartoonpgbench: Buffer Test - Normal Load - Read Onlyrocksdb: Seq Filllulesh: blender: BMW27 - CPU-Onlyospray: San Miguel - SciVismemtier-benchmark: Redisbuild-gdb: Time To Compilemkl-dnn: Recurrent Neural Network Training - f32intel-mpi: IMB-MPI1 Sendrecvintel-mpi: IMB-MPI1 Sendrecvleveldb: Rand Readcpp-perf-bench: Stepanov Vectorcpuminer-opt: deepcpuminer-opt: myr-grfio: Rand Write - IO_uring - No - Yes - 4KB - Default Test Directoryfio: Rand Write - IO_uring - No - Yes - 4KB - Default Test Directoryfio: Rand Write - POSIX AIO - Yes - No - 4KB - Default Test Directorydacapobench: Tradebeansgromacs: Water Benchmarkschbench: 8 - 8leveldb: Seq Fillleveldb: Seq Fillleveldb: Rand Deleteperl-benchmark: Pod2htmlrenaissance: Savina Reactors.IOnumenta-nab: Earthgecko Skylinemrbayes: Primate Phylogeny Analysisappleseed: Disney Materialttsiod-renderer: Phong Rendering With Soft-Shadow Mappingclomp: Static OMP Speedupcompress-7zip: Compress Speed Testnpb: EP.Dblender: BMW27 - CPU-Onlyjohn-the-ripper: MD5sockperf: Latency Under Loadospray: XFrog Forest - SciVistiobench: Write - 32MB - 4askap: tConvolve MT - Degriddingaskap: tConvolve MT - Griddingv-ray: CPUcompress-rar: Linux Source Tree Archiving To RARopencv-bench: mcperf: Setindigobench: Bedroomgegl: Wavelet Blurindigobench: Supercarfio: Rand Write - POSIX AIO - Yes - No - 4KB - Default Test Directoryminion: Solitairebuild2: Time To Compilerocksdb: Rand Fillgraphics-magick: Sharpengraphics-magick: Noise-Gaussiangraphics-magick: Resizingrocksdb: Rand Readgraphics-magick: Swirlgraphics-magick: Rotategraphics-magick: Enhancedgraphics-magick: HWB Color Spacemkl-dnn: Deconvolution Batch deconv_1d - f32himeno: Poisson Pressure Solverbuild-llvm: Time To Compilecpp-perf-bench: Atolsqlite-speedtest: Timed Time - Size 1,000radiance: SMP Parallelgegl: Color Enhanceparboil: OpenMP LBMnginx: Static Web Page Servingstockfish: Total Timeredis: SADDglibc-bench: sincosmcperf: Addebizzy: mcperf: Deletehackbench: 32 - Processperl-benchmark: Interpretermcperf: Getmpcbench: Multi-Precision Benchmarkmcperf: Replacestress-ng: Matrix Mathmcperf: Prependparboil: OpenMP MRI Griddingintel-mpi: IMB-MPI1 PingPonglibreoffice: 20 Documents To PDFintel-mpi: IMB-MPI1 Exchangeintel-mpi: IMB-MPI1 Exchangeaskap: tConvolve OpenMP - Degriddingaskap: tConvolve OpenMP - Griddingredis: LPUSHrenaissance: Rand Forestmkl-dnn: IP Batch All - f32redis: GETnpb: BT.Cospray: NASA Streamlines - Path Tracerjava-scimark2: Compositemkl-dnn: IP Batch 1D - f32mlpack: scikit_icabuild-linux-kernel: Time To Compilerenaissance: Apache Spark Bayesaskap: tConvolve MPI - Degriddingaskap: tConvolve MPI - Griddingcoremark: CoreMark Size 666 - Iterations Per Secondrawtherapee: Total Benchmark Timedeepspeech: CPUtachyon: Total Timeleveldb: Seek Randlibgav1: Summer Nature 1080pminion: Gracefulsockperf: Latency Ping Pongcython-bench: build-php: Time To Compiledcraw: RAW To PPM Image Conversionmkl-dnn: Convolution Batch conv_3d - f32redis: LPOPlzbench: XZ 0 - Decompressionlzbench: XZ 0 - Compressioncpuminer-opt: m7mgegl: Rotate 90 Degreesgegl: Antialiasospray: Magnetic Reconnection - SciVism-queens: Time To Solverust-mandel: Time To Complete Serial/Parallel Mandelbrotpostmark: Disk Transaction Performancemt-dgemm: Sustained Floating-Point Ratenpb: LU.Cstress-ng: Glibc Qsort Data Sortingrenaissance: Apache Spark PageRankqmlbench: Canvas Text Simplecompress-gzip: Linux Source Tree Archiving To .tar.gzmbw: Memory Copy, Fixed Block Size - 4096 MiBiperf: 5201 - 10 Seconds - TCP - 64embree: Pathtracer ISPC - Crowncpp-perf-bench: Ctyperedis: SETembree: Pathtracer - Crownredis: GETcpuminer-opt: skeinnode-octane: cpuminer-opt: sha256tintel-mpi: IMB-P2P PingPongpyperformance: floatfio: Rand Read - POSIX AIO - Yes - No - 4KB - Default Test Directoryfio: Rand Read - POSIX AIO - Yes - No - 4KB - Default Test Directoryfio: Seq Write - POSIX AIO - Yes - No - 4KB - Default Test Directoryfio: Seq Write - POSIX AIO - Yes - No - 4KB - Default Test Directoryfio: Seq Write - POSIX AIO - Yes - No - 4KB - Default Test Directoryfio: Rand Write - POSIX AIO - Yes - No - 4KB - Default Test Directoryfio: Seq Read - POSIX AIO - Yes - No - 4KB - Default Test Directoryfio: Seq Read - POSIX AIO - Yes - No - 4KB - Default Test Directoryfio: Rand Read - POSIX AIO - Yes - No - 4KB - Default Test Directoryfio: Seq Read - POSIX AIO - Yes - No - 4KB - Default Test Directoryglibc-bench: expstress-ng: CPU Stressjohn-the-ripper: Blowfishaircrack-ng: stress-ng: Tsearchstress-ng: Vector Mathstress-ng: Bsearchstress-ng: Cryptostress-ng: Socket Activitystress-ng: Memory Copyingstress-ng: Hsearchstress-ng: Forkingstress-ng: Glibc C String Functionsbuild-ffmpeg: Time To Compilegegl: Tile Glassgegl: Reflectminife: Smalllzbench: Zstd 8 - Decompressionlzbench: Zstd 8 - Compressionlzbench: Crush 0 - Decompressionlzbench: Crush 0 - Compressioncpp-perf-bench: Stepanov Abstractionespeak: Text-To-Speech Synthesisnpb: IS.Dsunflow: Global Illumination + Image Synthesisluajit: Compositestream: Copyapache: Static Web Page Servingscimark2: Compositerbenchmark: renaissance: Akka Unbalanced Cobwebbed Treeethr: HTTP - Bandwidth - 2inkscape: SVG Files To PNGc-ray: Total Time - 4K, 16 Rays Per Pixellzbench: Brotli 2 - Decompressionlzbench: Brotli 2 - Compressionpovray: Trace Timenumenta-nab: Bayesian Changepointtesseract-ocr: Time To OCR 7 Imagesmkl-dnn: Convolution Batch conv_alexnet - f32vpxenc: Speed 5botan: AES-256build-apache: Time To Compilelzbench: Brotli 0 - Decompressionlzbench: Brotli 0 - Compressiontungsten: Water Causticlzbench: Libdeflate 1 - Decompressionlzbench: Libdeflate 1 - Compressioncrafty: Elapsed Timepybench: Total For Average Test Timeslzbench: Zstd 1 - Decompressionlzbench: Zstd 1 - Compressionmutex: Shared Mutex Lock Sharedrsvg: SVG Files To PNGcompress-xz: Compressing ubuntu-16.04.3-server-i386.img, Compression Level 9t-test1: 1mbw: Memory Copy - 4096 MiBglibc-bench: singlibc-bench: cosbotan: Blowfishbotan: Twofishbotan: CAST-256botan: KASUMIopenssl: RSA 4096-bit Performancepolyhedron: acdarktable: Boat - CPU-onlyswet: Averagego-benchmark: buildgnupg: Linux 4.3 Package File Encryptionbuild-mplayer: Time To Compilecompilebench: Initial Createbuild-imagemagick: Time To Compilegimp: unsharp-maskrenaissance: In-Memory Database Shootoutarrayfire: BLAS CPUrenaissance: Apache Spark ALSospray: NASA Streamlines - SciVisrodinia: OpenMP Streamclusterphpbench: PHP Benchmark Suiteredis: SETxsbench: leveldb: Overwriteleveldb: Overwriteleveldb: Rand Fillleveldb: Rand Fillrust-prime: Prime Number Test To 200,000,000mlpack: scikit_svmdav1d: Summer Nature 4Kcpp-perf-bench: Function Objectsgimp: auto-levelssqlite: 1tungsten: Hairnpb: CG.Cramspeed: Scale - Integersockperf: Throughputneat: ipc-benchmark: FIFO Named Pipe - 128npb: FT.Ciperf: 5201 - 10 Seconds - UDP - 64npb: SP.Bencode-flac: WAV To FLACaom-av1: Speed 8 Realtimesudokut: Total Timemutex: Semaphore Release And Acquireosbench: Create Processesrodinia: OpenMP LavaMDramspeed: Copy - Integernumenta-nab: Relative Entropyoctave-benchmark: gimp: rotatego-benchmark: httpgnupg: 2GB File Encryptioncompress-zstd: Compressing ubuntu-16.04.3-server-i386.img, Compression Level 19rodinia: OpenMP CFD Solverlczero: BLASnetwork-loopback: Time To Transfer 10GB Via Loopbackdacapobench: H2glibc-bench: pthread_onceglibc-bench: atanhglibc-bench: tanhglibc-bench: sqrtglibc-bench: sinhglibc-bench: modfglibc-bench: log2glibc-bench: ffsllglibc-bench: asinhglibc-bench: ffssysbench: Memorysysbench: CPUrenaissance: Scala Dottyprimesieve: 1e12 Prime Number Generationnode-express-loadtest: go-benchmark: jsongo-benchmark: garbagex265: H.265 1080p Video Encodingbork: File Encryption Timex264: H.264 Video Encodingsvt-av1: Enc Mode 8 - 1080pparboil: OpenMP Stencilscikit-learn: rays1bench: Large Scenegegl: Cropgimp: resizemkl-dnn: Deconvolution Batch deconv_3d - f32neatbench: CPUn-queens: Elapsed Timeencode-mp3: WAV To MP3t-test1: 2dacapobench: Jythonfs-mark: 4000 Files, 32 Sub Dirs, 1MB Sizeunpack-linux: linux-4.15.tar.xznumenta-nab: Windowed Gaussiansvt-vp9: Visual Quality Optimized - Bosphorus 1080pnpb: MG.Ctjbench: Decompression Throughputffmpeg: H.264 HD To NTSC DVtungsten: Volumetric Caustictungsten: Non-Exponentialior: Read Testior: Write Testdav1d: Summer Nature 1080pipc-benchmark: Unnamed Pipe - 128oidn: Memorialsvt-hevc: 1080p 8-bit YUV To HEVC Video Encodeosbench: Create Filesgegl: Scaleencode-ogg: WAV To Ogglammps: Rhodopsin Proteinsmallpt: Global Illumination Renderer; 128 Samplesosbench: Memory Allocationsosbench: Launch Programsosbench: Create Threadsgit: Time To Complete Common Git Commandsnpb: EP.Cospray: Magnetic Reconnection - Path Tracerarrayfire: Conjugate Gradient CPUhmmer: Pfam Database Searchdarktable: Masskrug - CPU-onlybullet: Raytestsmultichase: 1GB Array, 256 Byte Stride, 4 Threadsipc-benchmark: Unnamed Unix Domain Socket - 128ipc-benchmark: TCP Socket - 128optcarrot: Optimized Benchmarkdarktable: Server Room - CPU-onlycloverleaf: Lagrangian-Eulerian Hydrodynamicsffte: N=256, 1D Complex FFT Routinepolybench-c: 3 Matrix Multiplicationspolybench-c: Covariance Computationpolybench-c: Correlation Computationtscp: AI Chess Performanceleveldb: Fill Syncleveldb: Fill Syncfs-mark: 1000 Files, 1MB Sizedarktable: Server Rack - CPU-onlyiozone: 4Kb - 512MB - Write Performanceiozone: 1MB - 512MB - Write Performanceparboil: OpenMP CUTCPiozone: 4Kb - 512MB - Write Performancectx-clock: Context Switch Timeblake2: iozone: 64Kb - 512MB - Write Performancesystem-libxml2: 1 MBfs-mark: 1000 Files, 1MB Size, No Sync/FSyncsystemd-boot-total: Totalsystemd-boot-total: Userspacesystemd-boot-total: Firmwaresystemd-boot-total: Loadersystemd-boot-total: Kernelbullet: Convex Trimeshbullet: Prim Trimeshbullet: 136 Ragdollsbullet: 1000 Convexbullet: 1000 Stackbullet: 3000 Fallscimark2: Jacobi Successive Over-Relaxationscimark2: Dense LU Matrix Factorizationscimark2: Sparse Matrix Multiplyscimark2: Fast Fourier Transformscimark2: Monte Carloluajit: Jacobi Successive Over-Relaxationluajit: Dense LU Matrix Factorizationluajit: Sparse Matrix Multiplyluajit: Fast Fourier Transformluajit: Monte Carlojava-scimark2: Jacobi Successive Over-Relaxationjava-scimark2: Dense LU Matrix Factorizationjava-scimark2: Sparse Matrix Multiplyjava-scimark2: Fast Fourier Transformjava-scimark2: Monte Carlohpcc: Max Ping Pong Bandwidthhpcc: Rand Ring Bandwidthhpcc: Rand Ring Latencyhpcc: G-Rand Accesshpcc: EP-STREAM Triadhpcc: G-Ptranshpcc: EP-DGEMMhpcc: G-Fftestream: Addstream: Triadstream: Scaleblogbench: WriteThreadripper 3970X16 Cores8885.0378623.65173.19643665.371595.2611110.91148126.048669.6589.10978526.01260198207968.7172581.711196.8029821713.9024786.28696.162775194.1721610159.937.40153.88386120425.421173301273.23220360507345146.02221.246119.47119630.1820042.412225.455626.42109.39432.11564.243291672.512.0710.67076875.5170.775485.94.693445331346494.653138.4849573.675794.656.58001.4771.2831832.5041834923035.0361162.93826863544416335.29591619767.040.6235714702.3293003107.52962144.286131.522285620407649940.526920010527241803353.028985329433311513379188.34434.8121267958.3126168429.6942.0231.70890.949792360.45605911479759.390898747.57502351667.5883.115299.85882.614955.0044.35377.194293231392318063370626366744963.2883587256.32827.6256.4360.1141024613551.88069.66375.41370.304869667.3390.711961862907.5840.40252075013.2348.9910.5503738.112651.594584168.08966.35545117.54.90662.3959.317103060.64904549.57610832254686312243180626444159069366914992.786884607.312790122.76658.05758.588173.31756.06952.96923636707.66957287871851271.3512.262245047.0254237274147.431.3890.0013662377680.2905047267.1234235.7646495.960.1751875106.636.663149.286258.073764.474771.161417209.043840.45917.32972005988.5074815.9913.583009.123.4165549.2024.5422677.5133795.212682.401331439.75223540.81056.6845525.035368.59276.4641.3361152.93140.85832.03739.6556.828462175519.08131411624.3038.16539.00332.6217.94532.093781210.86853465854.97489.8116363.844358.833.91412649.3085484035.545732.1661674234.1337.65722229021.421027035012522362317789641.37681210356.214400307000120045933317944.9931712151.3143255112289.3001111.0488021.7220222.039207.8522217.768324.57253747.4297017.283406391.4621.96931.03329.9839322.75165610750012128.62227.3531464.450.5171659.5152075.836286.78718.030.285812746.6141844.5826.27116.37773921613.23524.89524.084137.78524.485902.14118.58463354718.68351245266889278895815035431923.87018.65225.60021587.37742.036642.2514475.104391.425154.147100.1029307.45.916.0108114425482006767710112.38411.581716.2213.43020.1655898.137802.0114708.53966.6720.4016417561687462.136079554263.93826.8259.84027.310.27211.50324.6614.86816.98914.4088.1821710815.5942704.7453424012.818338832434599.4967.133710.707.76140.8812.9041127.9637188.33741870.312.41713.06712.437782511.4278.5449.7842105.1110.40438861.7613310.091410.68412.220277.796962.234425.882181.758978.418351.758709119242.226668765.37785168.0666.12310908156416858971264.908.121205.3169.7415.2821417.973225.768.0038.4704.9123937.63.4967.1888.7434023560.74.9846.111368.4130218.91218.8990177.3313.563822.642891047.58182.76726.09355889321.96174.7511.4467565.6615.22517.2742.89470.31838149.30257811.3296514.9142831.86444.4425.184.8423.0882.10647574.56124162133578760103.981.9911.059483.922.1392.1412.12013368771938.1673.6576.30.1201529.571554.840.9933421594.311488.351677.554242724.8370553134520002997657100.9520420.7967872.1735803.7872014.0948883.3942501298.02983.93793.80356.71157.712133.373996.691352.77287.97526.751957.926674.412683.182043.321686.7921719.6472.208140.469620.171381.903487.4411929.8030716.1230756819.257321.851667.1498744341.3434088.41489.992531382.091466.071115.81672538.571799.0979.61688109.7128348310801369.921039.412881.1796211182.6126671.34693.034360501.0721534430.817.69414.42381206340.365153219275.7293166210066382.22220.226314.82167550.1020120.612317.754186.82283.82430.36564.7891062.302.409.972766841.8163.052602.31.753976001553423.751289.3059871.375767.176.35411.42178.4631631.7919813493024.6861009.39469054042231623.537415509163.131.2390115067.47744108.5163358.974176.691013238136085868.92457509618147567334.164106331573312343998333.22138.272258922.9316301159.7345.4431.73388.677272392.97447670280910.902115119.5818.522272891.4282.193142.35063.496263.739.31378.8906879.9115369.09335000130927700024522.0484631542.29941.839.8970.1225662915896.11875.99269.520146.682849617.8463.04784881480.1097.69162600010.7753.3210.4093463.502306.502043359.83761.53575424.01.85260.1144.106108261.09467871.88360697625235118477277439778473235714173.224354880.015106232.3358.98758.173171.48955.42936.35483736762.59369327421874999.4012.405372837.7128437888516.674.3530.0009118689370.8901384273.876991.0488961.437.0924185828.126.541104.957217.493482.693330.881492623.123826.80311.56872184449.1154825.134.973005.073.5504236.0549.3573157.2553544.172368.48508354.39941742.41449.1047258.319814.888105.7641.6720442.21039.93948.21439.4289.503722097030.0013041654.6237.69136.36113.1655.08440.17363574.66695954335.46181.5516238.527386.434.07112748.6084721612.826132.3781723424.5913.22842108260.5349073515821069509362495.739130599.745.21160032033312513193332733333643331424116003690004.951895019.092611154172.197383.1835109.177239.754392.3610510.353963.81106584.8010970.251252570.2638.01728.68729.41110962.7165010949812028.86727.9081521.340.9541624.1449216.841019.20697.800.13618732.8413314.79225.26234.15773621532.24624.67123.823178.66628.625858.02927.54262853924.12601244267886984695514935431918.67623.63416.05921862.42642.278842.8141471.104385.677154.99299.8494550.95.895.0608155974661724446290512.46525.090548.7522.30215.1164006.599481.3684278.4052511.2666413801696335.08384040443.84640.343.87440.319.95811.43205.5014.97812.77213.91519.255810888.1936372.4453208913.981335394631324.2067.124225.647.77662.3712.2321154.98329816.53443057.6911.59010.57711.188405911.47013.56512.0773630.2410.80626061.7759410.201010.71652.239797.829572.254155.897991.778008.537111.787548262426.982933077.16625197.59712.53811858352358589793673.218.556131.8835.8907.7820157.87595.947.4916.9233.6916630.711.3217.1825.4493921633.24.9647.013172.1328788.06218.3427055.0418.564208.922151117.52622.53556.99370194914.9782.1313.8789434.9405.25312.7477.40767.155997228.57586546.1204855.0181486.4425023.523.6783.7462.12722673.27025554313705204102.752.0571.262.1352.1202.11413193721677.7071611.70.0971280.551335.372.1222031283.8743489581858.541361.614242012.9390483287820153354261700.9615520.8049272.1952783.8227674.1321333.4281561275.97946.17782.85328.58155.442097.743910.401319.00278.18515.371951.486637.132716.602040.931679.2122271.5423.976860.407880.113623.476142.1262949.5807011.1667354720.554847.949498.811700OpenBenchmarking.org

Pennant

Pennant is an application focused on hydrodynamics on general unstructured meshes in 2D. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgHydro Cycle Time - Seconds, Fewer Is BetterPennant 1.0.1Test: sedovbigThreadripper 3970X16 Cores2K4K6K8K10KSE +/- 5.38, N = 3SE +/- 4.90, N = 38885.044341.341. (CXX) g++ options: -fopenmp -pthread -lmpi_cxx -lmpi
OpenBenchmarking.orgHydro Cycle Time - Seconds, Fewer Is BetterPennant 1.0.1Test: sedovbigThreadripper 3970X16 Cores15003000450060007500Min: 8879.51 / Avg: 8885.04 / Max: 8895.8Min: 4336.37 / Avg: 4341.34 / Max: 4351.141. (CXX) g++ options: -fopenmp -pthread -lmpi_cxx -lmpi

OpenBenchmarking.orgHydro Cycle Time - Seconds, Fewer Is BetterPennant 1.0.1Test: leblancbigThreadripper 3970X16 Cores2K4K6K8K10KSE +/- 2.82, N = 3SE +/- 1.62, N = 38623.654088.411. (CXX) g++ options: -fopenmp -pthread -lmpi_cxx -lmpi
OpenBenchmarking.orgHydro Cycle Time - Seconds, Fewer Is BetterPennant 1.0.1Test: leblancbigThreadripper 3970X16 Cores15003000450060007500Min: 8618.4 / Avg: 8623.65 / Max: 8628.08Min: 4085.18 / Avg: 4088.41 / Max: 4090.081. (CXX) g++ options: -fopenmp -pthread -lmpi_cxx -lmpi

HPC Challenge

HPC Challenge (HPCC) is a cluster-focused benchmark consisting of the HPL Linpack TPP benchmark, DGEMM, STREAM, PTRANS, RandomAccess, FFT, and communication bandwidth and latency. This HPC Challenge test profile attempts to ship with standard yet versatile configuration/input files though they can be modified. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOPS, More Is BetterHPC Challenge 1.5.0Test / Class: G-HPLThreadripper 3970X16 Cores20406080100SE +/- 0.17, N = 3SE +/- 0.19, N = 373.2089.991. (CC) gcc options: -lblas -lm -pthread -lmpi -fomit-frame-pointer -funroll-loops2. ATLAS + Open MPI 3.1.3
OpenBenchmarking.orgGFLOPS, More Is BetterHPC Challenge 1.5.0Test / Class: G-HPLThreadripper 3970X16 Cores20406080100Min: 72.98 / Avg: 73.2 / Max: 73.52Min: 89.62 / Avg: 89.99 / Max: 90.21. (CC) gcc options: -lblas -lm -pthread -lmpi -fomit-frame-pointer -funroll-loops2. ATLAS + Open MPI 3.1.3

Dbench

Dbench is a benchmark designed by the Samba project as a free alternative to netbench, but dbench contains only file-system calls for testing the disk performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterDbench 4.0Client Count: 12Threadripper 3970X16 Cores30060090012001500SE +/- 26.66, N = 9SE +/- 46.78, N = 9665.371382.09
OpenBenchmarking.orgMB/s, More Is BetterDbench 4.0Client Count: 12Threadripper 3970X16 Cores2004006008001000Min: 567.65 / Avg: 665.37 / Max: 795.66Min: 1089.61 / Avg: 1382.09 / Max: 1500.29

OpenBenchmarking.orgMB/s, More Is BetterDbench 4.0Client Count: 12Threadripper 3970X16 Cores30060090012001500SE +/- 13.11, N = 9SE +/- 14.76, N = 8595.261466.071. (CC) gcc options: -lpopt -O2
OpenBenchmarking.orgMB/s, More Is BetterDbench 4.0Client Count: 12Threadripper 3970X16 Cores30060090012001500Min: 555.46 / Avg: 595.26 / Max: 654.89Min: 1384.36 / Avg: 1466.07 / Max: 1500.261. (CC) gcc options: -lpopt -O2

CppPerformanceBenchmarks

CppPerformanceBenchmarks is a set of C++ compiler performance benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterCppPerformanceBenchmarks 9Test: Random NumbersThreadripper 3970X16 Cores2004006008001000SE +/- 0.90, N = 3SE +/- 0.28, N = 31110.911115.821. (CXX) g++ options: -std=c++11 -O3
OpenBenchmarking.orgSeconds, Fewer Is BetterCppPerformanceBenchmarks 9Test: Random NumbersThreadripper 3970X16 Cores2004006008001000Min: 1109.55 / Avg: 1110.91 / Max: 1112.62Min: 1115.28 / Avg: 1115.82 / Max: 1116.251. (CXX) g++ options: -std=c++11 -O3

Memcached mcperf

This is a test of twmperf/mcperf with memcached, a distributed memory object caching system. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOperations Per Second, More Is BetterMemcached mcperf 1.5.22Method: Set - Connections: 64Threadripper 3970X16 Cores16K32K48K64K80KSE +/- 524.58, N = 3SE +/- 722.58, N = 348126.072538.51. (CC) gcc options: -O2 -lm -rdynamic
OpenBenchmarking.orgOperations Per Second, More Is BetterMemcached mcperf 1.5.22Method: Set - Connections: 64Threadripper 3970X16 Cores13K26K39K52K65KMin: 47130.3 / Avg: 48126 / Max: 48910.2Min: 71347.4 / Avg: 72538.5 / Max: 73842.81. (CC) gcc options: -O2 -lm -rdynamic

OpenBenchmarking.orgOperations Per Second, More Is BetterMemcached mcperf 1.5.22Method: Add - Connections: 64Threadripper 3970X16 Cores15K30K45K60K75KSE +/- 803.43, N = 3SE +/- 347.35, N = 348669.671799.01. (CC) gcc options: -O2 -lm -rdynamic
OpenBenchmarking.orgOperations Per Second, More Is BetterMemcached mcperf 1.5.22Method: Add - Connections: 64Threadripper 3970X16 Cores12K24K36K48K60KMin: 47065.1 / Avg: 48669.63 / Max: 49546.7Min: 71397.1 / Avg: 71799.03 / Max: 72490.71. (CC) gcc options: -O2 -lm -rdynamic

Timed GCC Compilation

This test times how long it takes to build the GNU Compiler Collection (GCC). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed GCC Compilation 8.2Time To CompileThreadripper 3970X16 Cores2004006008001000SE +/- 0.83, N = 3SE +/- 1.05, N = 3589.11979.62
OpenBenchmarking.orgSeconds, Fewer Is BetterTimed GCC Compilation 8.2Time To CompileThreadripper 3970X16 Cores2004006008001000Min: 587.98 / Avg: 589.11 / Max: 590.73Min: 978.43 / Avg: 979.62 / Max: 981.7

Memcached mcperf

This is a test of twmperf/mcperf with memcached, a distributed memory object caching system. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOperations Per Second, More Is BetterMemcached mcperf 1.5.22Method: Get - Connections: 64Threadripper 3970X16 Cores20K40K60K80K100KSE +/- 810.66, N = 3SE +/- 238.31, N = 378526.088109.71. (CC) gcc options: -O2 -lm -rdynamic
OpenBenchmarking.orgOperations Per Second, More Is BetterMemcached mcperf 1.5.22Method: Get - Connections: 64Threadripper 3970X16 Cores15K30K45K60K75KMin: 76905.4 / Avg: 78526.03 / Max: 79377Min: 87727.5 / Avg: 88109.67 / Max: 88547.41. (CC) gcc options: -O2 -lm -rdynamic

BlogBench

BlogBench is designed to replicate the load of a real-world busy file server by stressing the file-system with multiple threads of random reads, writes, and rewrites. The behavior is mimicked of that of a blog by creating blogs with content and pictures, modifying blog posts, adding comments to these blogs, and then reading the content of the blogs. All of these blogs generated are created locally with fake content and pictures. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFinal Score, More Is BetterBlogBench 1.1Test: ReadThreadripper 3970X16 Cores300K600K900K1200K1500KSE +/- 12676.46, N = 8SE +/- 3455.51, N = 3126019812834831. (CC) gcc options: -O2 -pthread
OpenBenchmarking.orgFinal Score, More Is BetterBlogBench 1.1Test: ReadThreadripper 3970X16 Cores200K400K600K800K1000KMin: 1202576 / Avg: 1260197.75 / Max: 1296307Min: 1277229 / Avg: 1283483 / Max: 12891571. (CC) gcc options: -O2 -pthread

MariaDB

This is a MariaDB MySQL database server benchmark making use of mysqlslap. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 10.3.8Clients: 64Threadripper 3970X16 Cores2004006008001000SE +/- 2.66, N = 3SE +/- 1.35, N = 320710801. (CXX) g++ options: -pie -fPIC -fstack-protector -fno-rtti -O2 -lpthread -llzma -lbz2 -laio -lnuma -lz -lm -lpcre -lcrypt -lssl -lcrypto -ldl
OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 10.3.8Clients: 64Threadripper 3970X16 Cores2004006008001000Min: 202.68 / Avg: 206.84 / Max: 211.77Min: 1077.59 / Avg: 1079.92 / Max: 1082.251. (CXX) g++ options: -pie -fPIC -fstack-protector -fno-rtti -O2 -lpthread -llzma -lbz2 -laio -lnuma -lz -lm -lpcre -lcrypt -lssl -lcrypto -ldl

MKL-DNN DNNL

This is a test of the Intel MKL-DNN (DNNL / Deep Neural Network Library) 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. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN DNNL 1.1Harness: Convolution Batch conv_all - Data Type: f32Threadripper 3970X16 Cores30060090012001500SE +/- 4.86, N = 3SE +/- 2.68, N = 3968.721369.92MIN: 791.36MIN: 1354.971. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -fopenmp -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN DNNL 1.1Harness: Convolution Batch conv_all - Data Type: f32Threadripper 3970X16 Cores2004006008001000Min: 960.52 / Avg: 968.72 / Max: 977.34Min: 1365.64 / Avg: 1369.92 / Max: 1374.841. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -fopenmp -pie -lpthread -ldl

QMCPACK

QMCPACK is a modern high-performance open-source Quantum Monte Carlo (QMC) simulation code making use of MPI for this benchmark of the H20 example code. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTotal Execution Time - Seconds, Fewer Is BetterQMCPACK 3.8Threadripper 3970X16 Cores60012001800240030002581.71039.41. (CXX) g++ options: -fopenmp -fomit-frame-pointer -finline-limit=1000 -fstrict-aliasing -funroll-all-loops -march=native -O3 -ffast-math -lm

PostgreSQL pgbench

This is a simple benchmark of PostgreSQL using pgbench. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 12.0Scaling: Buffer Test - Test: Normal Load - Mode: Read WriteThreadripper 3970X16 Cores3K6K9K12K15KSE +/- 333.19, N = 15SE +/- 1319.04, N = 1511196.8012881.181. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lpthread -lrt -lcrypt -ldl -lm
OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 12.0Scaling: Buffer Test - Test: Normal Load - Mode: Read WriteThreadripper 3970X16 Cores2K4K6K8K10KMin: 7914.69 / Avg: 11196.8 / Max: 13328.07Min: 8973.46 / Avg: 12881.18 / Max: 27101.091. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lpthread -lrt -lcrypt -ldl -lm

MKL-DNN DNNL

This is a test of the Intel MKL-DNN (DNNL / Deep Neural Network Library) 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. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN DNNL 1.1Harness: Deconvolution Batch deconv_all - Data Type: f32Threadripper 3970X16 Cores400800120016002000SE +/- 2.16, N = 3SE +/- 0.54, N = 31713.901182.61MIN: 1362.26MIN: 1163.991. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -fopenmp -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN DNNL 1.1Harness: Deconvolution Batch deconv_all - Data Type: f32Threadripper 3970X16 Cores30060090012001500Min: 1709.78 / Avg: 1713.9 / Max: 1717.11Min: 1181.78 / Avg: 1182.61 / Max: 1183.611. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -fopenmp -pie -lpthread -ldl

Algebraic Multi-Grid Benchmark

AMG is a parallel algebraic multigrid solver for linear systems arising from problems on unstructured grids. The driver provided with AMG builds linear systems for various 3-dimensional problems. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFigure Of Merit, More Is BetterAlgebraic Multi-Grid BenchmarkThreadripper 3970X16 Cores6K12K18K24K30KSE +/- 14.75, N = 3SE +/- 24.90, N = 324786.2826671.341. (CC) gcc options: -lparcsr_ls -lparcsr_mv -lseq_mv -lIJ_mv -lkrylov -lHYPRE_utilities -lm -fopenmp -pthread -lmpi
OpenBenchmarking.orgFigure Of Merit, More Is BetterAlgebraic Multi-Grid BenchmarkThreadripper 3970X16 Cores5K10K15K20K25KMin: 24760.59 / Avg: 24786.28 / Max: 24811.69Min: 26640.06 / Avg: 26671.34 / Max: 26720.531. (CC) gcc options: -lparcsr_ls -lparcsr_mv -lseq_mv -lIJ_mv -lkrylov -lHYPRE_utilities -lm -fopenmp -pthread -lmpi

Crypto++

Crypto++ is a C++ class library of cryptographic algorithms. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/second, More Is BetterCrypto++ 8.2Test: Keyed AlgorithmsThreadripper 3970X16 Cores150300450600750SE +/- 1.50, N = 3SE +/- 0.14, N = 3696.16693.031. (CXX) g++ options: -g2 -O3 -fPIC -pthread -pipe
OpenBenchmarking.orgMiB/second, More Is BetterCrypto++ 8.2Test: Keyed AlgorithmsThreadripper 3970X16 Cores120240360480600Min: 694.57 / Avg: 696.16 / Max: 699.15Min: 692.88 / Avg: 693.03 / Max: 693.311. (CXX) g++ options: -g2 -O3 -fPIC -pthread -pipe

Blender

Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing via OpenCL or CUDA is supported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.82Blend File: Barbershop - Compute: CPU-OnlyThreadripper 3970X16 Cores110220330440550SE +/- 0.21, N = 3SE +/- 0.54, N = 3194.17501.07
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.82Blend File: Barbershop - Compute: CPU-OnlyThreadripper 3970X16 Cores90180270360450Min: 193.81 / Avg: 194.17 / Max: 194.55Min: 500.4 / Avg: 501.07 / Max: 502.14

FFTW

FFTW is a C subroutine library for computing the discrete Fourier transform (DFT) in one or more dimensions. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Float + SSE - Size: 2D FFT Size 4096Threadripper 3970X16 Cores5K10K15K20K25KSE +/- 102.34, N = 3SE +/- 48.64, N = 321610215341. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm
OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Float + SSE - Size: 2D FFT Size 4096Threadripper 3970X16 Cores4K8K12K16K20KMin: 21405 / Avg: 21609.67 / Max: 21714Min: 21446 / Avg: 21533.67 / Max: 216141. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

Blender

Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing is supported. This system/blender test profile makes use of the system-supplied Blender. Use pts/blender if wishing to stick to a fixed version of Blender. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.80Blend File: Barbershop - Compute: CPU-OnlyThreadripper 3970X16 Cores90180270360450SE +/- 0.33, N = 3SE +/- 0.65, N = 3159.93430.81
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.80Blend File: Barbershop - Compute: CPU-OnlyThreadripper 3970X16 Cores80160240320400Min: 159.6 / Avg: 159.93 / Max: 160.58Min: 429.52 / Avg: 430.81 / Max: 431.56

PlaidML

This test profile uses PlaidML deep learning framework developed by Intel for offering up various benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: ResNet 50 - Device: CPUThreadripper 3970X16 Cores246810SE +/- 0.05, N = 3SE +/- 0.03, N = 37.407.69
OpenBenchmarking.orgFPS, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: ResNet 50 - Device: CPUThreadripper 3970X16 Cores3691215Min: 7.31 / Avg: 7.4 / Max: 7.48Min: 7.63 / Avg: 7.69 / Max: 7.75

Blender

Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing via OpenCL or CUDA is supported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.82Blend File: Pabellon Barcelona - Compute: CPU-OnlyThreadripper 3970X16 Cores90180270360450SE +/- 0.59, N = 3SE +/- 0.44, N = 3153.88414.42
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.82Blend File: Pabellon Barcelona - Compute: CPU-OnlyThreadripper 3970X16 Cores70140210280350Min: 153.17 / Avg: 153.88 / Max: 155.04Min: 413.77 / Avg: 414.42 / Max: 415.25

Hierarchical INTegration

This test runs the U.S. Department of Energy's Ames Laboratory Hierarchical INTegration (HINT) benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQUIPs, More Is BetterHierarchical INTegration 1.0Test: FLOATThreadripper 3970X16 Cores80M160M240M320M400MSE +/- 452031.90, N = 3SE +/- 450887.49, N = 3386120425.42381206340.371. (CC) gcc options: -O3 -march=native -lm
OpenBenchmarking.orgQUIPs, More Is BetterHierarchical INTegration 1.0Test: FLOATThreadripper 3970X16 Cores70M140M210M280M350MMin: 385253858.06 / Avg: 386120425.42 / Max: 386776855.39Min: 380313827.37 / Avg: 381206340.37 / Max: 381764239.191. (CC) gcc options: -O3 -march=native -lm

DaCapo Benchmark

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

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 9.12-MR1Java Test: TradesoapThreadripper 3970X16 Cores7001400210028003500SE +/- 29.24, N = 11SE +/- 23.31, N = 2033013219
OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 9.12-MR1Java Test: TradesoapThreadripper 3970X16 Cores6001200180024003000Min: 3179 / Avg: 3301.45 / Max: 3527Min: 3001 / Avg: 3218.75 / Max: 3454

CppPerformanceBenchmarks

CppPerformanceBenchmarks is a set of C++ compiler performance benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterCppPerformanceBenchmarks 9Test: Math LibraryThreadripper 3970X16 Cores60120180240300SE +/- 1.09, N = 3SE +/- 0.39, N = 3273.23275.731. (CXX) g++ options: -std=c++11 -O3
OpenBenchmarking.orgSeconds, Fewer Is BetterCppPerformanceBenchmarks 9Test: Math LibraryThreadripper 3970X16 Cores50100150200250Min: 271.46 / Avg: 273.23 / Max: 275.22Min: 275.22 / Avg: 275.73 / Max: 276.51. (CXX) g++ options: -std=c++11 -O3

Apache Cassandra

This is a benchmark of the Apache Cassandra NoSQL database management system making use of cassandra-stress. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterApache Cassandra 3.11.4Test: ReadsThreadripper 3970X16 Cores4K8K12K16K20KSE +/- 7579.05, N = 9SE +/- 37.22, N = 3203603166
OpenBenchmarking.orgOp/s, More Is BetterApache Cassandra 3.11.4Test: ReadsThreadripper 3970X16 Cores4K8K12K16K20KMin: 4309 / Avg: 20360.11 / Max: 73306Min: 3125 / Avg: 3165.67 / Max: 3240

BRL-CAD

BRL-CAD 7.28.0 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 7.28.0VGR Performance MetricThreadripper 3970X16 Cores110K220K330K440K550K5073452100661. (CXX) g++ options: -std=c++98 -pipe -fno-strict-aliasing -fno-common -fexceptions -ftemplate-depth-128 -m64 -ggdb3 -O3 -fipa-pta -fstrength-reduce -finline-functions -flto -pedantic -rdynamic -lSM -lICE -lXi -lGLU -lGL -lX11 -lXext -lXrender -lpthread -ldl -luuid -lm

Blender

Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing is supported. This system/blender test profile makes use of the system-supplied Blender. Use pts/blender if wishing to stick to a fixed version of Blender. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.80Blend File: Pabellon Barcelona - Compute: CPU-OnlyThreadripper 3970X16 Cores80160240320400SE +/- 0.24, N = 3SE +/- 0.17, N = 3146.02382.22
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.80Blend File: Pabellon Barcelona - Compute: CPU-OnlyThreadripper 3970X16 Cores70140210280350Min: 145.56 / Avg: 146.02 / Max: 146.39Min: 381.93 / Avg: 382.22 / Max: 382.52

LZMA Compression

This test measures the time needed to compress a file using LZMA compression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterLZMA Compression256MB File CompressionThreadripper 3970X16 Cores50100150200250SE +/- 0.33, N = 3SE +/- 0.09, N = 3221.25220.231. (CXX) g++ options: -O2
OpenBenchmarking.orgSeconds, Fewer Is BetterLZMA Compression256MB File CompressionThreadripper 3970X16 Cores4080120160200Min: 220.76 / Avg: 221.25 / Max: 221.87Min: 220.05 / Avg: 220.23 / Max: 220.341. (CXX) g++ options: -O2

Blender

Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing via OpenCL or CUDA is supported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.82Blend File: Classroom - Compute: CPU-OnlyThreadripper 3970X16 Cores70140210280350SE +/- 0.31, N = 3SE +/- 0.36, N = 3119.47314.82
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.82Blend File: Classroom - Compute: CPU-OnlyThreadripper 3970X16 Cores60120180240300Min: 118.86 / Avg: 119.47 / Max: 119.85Min: 314.32 / Avg: 314.82 / Max: 315.51

KeyDB

A benchmark of KeyDB as a multi-threaded fork of the Redis server. The KeyDB benchmark is conducted using memtier-benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps/sec, More Is BetterKeyDB 5.3.1Threadripper 3970X16 Cores40K80K120K160K200KSE +/- 2224.40, N = 15SE +/- 1098.87, N = 3119630.18167550.101. (CXX) g++ options: -O2 -levent -lpthread -lz -lpcre
OpenBenchmarking.orgOps/sec, More Is BetterKeyDB 5.3.1Threadripper 3970X16 Cores30K60K90K120K150KMin: 112541.36 / Avg: 119630.18 / Max: 147500.03Min: 166041.98 / Avg: 167550.1 / Max: 169688.611. (CXX) g++ options: -O2 -levent -lpthread -lz -lpcre

Tinymembench

This benchmark tests the system memory (RAM) performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterTinymembench 2018-05-28Standard MemsetThreadripper 3970X16 Cores4K8K12K16K20KSE +/- 54.47, N = 3SE +/- 20.00, N = 320042.420120.61. (CC) gcc options: -O2 -lm
OpenBenchmarking.orgMB/s, More Is BetterTinymembench 2018-05-28Standard MemsetThreadripper 3970X16 Cores3K6K9K12K15KMin: 19977.3 / Avg: 20042.4 / Max: 20150.6Min: 20081.4 / Avg: 20120.57 / Max: 20147.21. (CC) gcc options: -O2 -lm

OpenBenchmarking.orgMB/s, More Is BetterTinymembench 2018-05-28Standard MemcpyThreadripper 3970X16 Cores3K6K9K12K15KSE +/- 13.44, N = 3SE +/- 3.80, N = 312225.412317.71. (CC) gcc options: -O2 -lm
OpenBenchmarking.orgMB/s, More Is BetterTinymembench 2018-05-28Standard MemcpyThreadripper 3970X16 Cores2K4K6K8K10KMin: 12208.7 / Avg: 12225.4 / Max: 12252Min: 12313 / Avg: 12317.67 / Max: 12325.21. (CC) gcc options: -O2 -lm

Apache Siege

This is a test of the Apache web server performance being facilitated by the Siege web serverb enchmark program. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTransactions Per Second, More Is BetterApache Siege 2.4.29Concurrent Users: 250Threadripper 3970X16 Cores12K24K36K48K60KSE +/- 1582.21, N = 15SE +/- 579.13, N = 1255626.4254186.821. (CC) gcc options: -O2 -lpthread -ldl -lssl -lcrypto
OpenBenchmarking.orgTransactions Per Second, More Is BetterApache Siege 2.4.29Concurrent Users: 250Threadripper 3970X16 Cores10K20K30K40K50KMin: 39613.38 / Avg: 55626.42 / Max: 59396.53Min: 47819.43 / Avg: 54186.82 / Max: 54884.741. (CC) gcc options: -O2 -lpthread -ldl -lssl -lcrypto

Blender

Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing is supported. This system/blender test profile makes use of the system-supplied Blender. Use pts/blender if wishing to stick to a fixed version of Blender. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.80Blend File: Classroom - Compute: CPU-OnlyThreadripper 3970X16 Cores60120180240300SE +/- 0.18, N = 3SE +/- 0.30, N = 3109.39283.82
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.80Blend File: Classroom - Compute: CPU-OnlyThreadripper 3970X16 Cores50100150200250Min: 109.04 / Avg: 109.39 / Max: 109.63Min: 283.37 / Avg: 283.82 / Max: 284.39

Numpy Benchmark

This is a test to obtain the general Numpy performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgScore, More Is BetterNumpy BenchmarkThreadripper 3970X16 Cores90180270360450SE +/- 1.07, N = 3SE +/- 0.33, N = 3432.11430.36
OpenBenchmarking.orgScore, More Is BetterNumpy BenchmarkThreadripper 3970X16 Cores80160240320400Min: 430.52 / Avg: 432.11 / Max: 434.15Min: 429.76 / Avg: 430.36 / Max: 430.89

Radiance Benchmark

This is a benchmark of NREL Radiance, a synthetic imaging system that is open-source and developed by the Lawrence Berkeley National Laboratory in California. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterRadiance Benchmark 5.0Test: SerialThreadripper 3970X16 Cores120240360480600564.24564.78

Facebook RocksDB

This is a benchmark of 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 BetterFacebook RocksDB 6.3.6Test: Random Fill SyncThreadripper 3970X16 Cores6K12K18K24K30KSE +/- 999.84, N = 15SE +/- 17.38, N = 32916791061. (CXX) g++ options: -O3 -march=native -std=c++11 -fno-builtin-memcmp -fno-rtti -rdynamic -lpthread
OpenBenchmarking.orgOp/s, More Is BetterFacebook RocksDB 6.3.6Test: Random Fill SyncThreadripper 3970X16 Cores5K10K15K20K25KMin: 22081 / Avg: 29167.33 / Max: 34029Min: 9082 / Avg: 9106.33 / Max: 91401. (CXX) g++ options: -O3 -march=native -std=c++11 -fno-builtin-memcmp -fno-rtti -rdynamic -lpthread

LuxCoreRender

LuxCoreRender is an open-source physically based renderer. This test profile is focused on running LuxCoreRender on the CPU as opposed to the OpenCL version. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.2Scene: DLSCThreadripper 3970X16 Cores0.56481.12961.69442.25922.824SE +/- 0.19, N = 15SE +/- 0.02, N = 32.512.30MIN: 1.3 / MAX: 3MIN: 2.18 / MAX: 2.41
OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.2Scene: DLSCThreadripper 3970X16 Cores246810Min: 1.33 / Avg: 2.51 / Max: 2.98Min: 2.27 / Avg: 2.3 / Max: 2.34

OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.2Scene: Rainbow Colors and PrismThreadripper 3970X16 Cores0.541.081.622.162.7SE +/- 0.17, N = 15SE +/- 0.02, N = 32.072.40MIN: 1.13 / MAX: 3.1MIN: 2.33 / MAX: 2.49
OpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.2Scene: Rainbow Colors and PrismThreadripper 3970X16 Cores246810Min: 1.22 / Avg: 2.07 / Max: 3.08Min: 2.38 / Avg: 2.4 / Max: 2.43

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 3.1Threadripper 3970X16 Cores3691215SE +/- 0.00239, N = 3SE +/- 0.00185, N = 310.670709.972761. (CXX) g++ options: -O3 -ffast-math -ftree-vectorize -pthread -lmpi_cxx -lmpi
OpenBenchmarking.orgGFLOP/s, More Is BetterHigh Performance Conjugate Gradient 3.1Threadripper 3970X16 Cores3691215Min: 10.67 / Avg: 10.67 / Max: 10.68Min: 9.97 / Avg: 9.97 / Max: 9.981. (CXX) g++ options: -O3 -ffast-math -ftree-vectorize -pthread -lmpi_cxx -lmpi

FFTW

FFTW is a C subroutine library for computing the discrete Fourier transform (DFT) in one or more dimensions. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Stock - Size: 2D FFT Size 4096Threadripper 3970X16 Cores15003000450060007500SE +/- 2.20, N = 3SE +/- 16.83, N = 36875.56841.81. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm
OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Stock - Size: 2D FFT Size 4096Threadripper 3970X16 Cores12002400360048006000Min: 6872.2 / Avg: 6875.53 / Max: 6879.7Min: 6808.3 / Avg: 6841.83 / Max: 6861.11. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

Zstd Compression

This test measures the time needed to compress a sample file (an Ubuntu file-system image) using Zstd compression supplied by the system or otherwise externally of the test profile. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterZstd CompressionCompressing ubuntu-16.04.3-server-i386.img, Compression Level 19Threadripper 3970X16 Cores4080120160200SE +/- 1.03, N = 3SE +/- 0.45, N = 3170.78163.05
OpenBenchmarking.orgSeconds, Fewer Is BetterZstd CompressionCompressing ubuntu-16.04.3-server-i386.img, Compression Level 19Threadripper 3970X16 Cores306090120150Min: 168.74 / Avg: 170.78 / Max: 172.03Min: 162.35 / Avg: 163.05 / Max: 163.9

FS-Mark

FS_Mark is designed to test a system's file-system performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFiles/s, More Is BetterFS-Mark 3.3Test: 5000 Files, 1MB Size, 4 ThreadsThreadripper 3970X16 Cores130260390520650SE +/- 28.74, N = 12SE +/- 81.95, N = 12485.9602.31. (CC) gcc options: -static
OpenBenchmarking.orgFiles/s, More Is BetterFS-Mark 3.3Test: 5000 Files, 1MB Size, 4 ThreadsThreadripper 3970X16 Cores110220330440550Min: 338.4 / Avg: 485.88 / Max: 699.2Min: 417.6 / Avg: 602.26 / Max: 1172.71. (CC) gcc options: -static

OSPray

Intel OSPray is a portable ray-tracing engine for high-performance, high-fidenlity 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.orgFPS, More Is BetterOSPray 1.8.5Demo: XFrog Forest - Renderer: Path TracerThreadripper 3970X16 Cores1.05532.11063.16594.22125.2765SE +/- 0.00, N = 12SE +/- 0.00, N = 34.691.75MIN: 4.65 / MAX: 4.76MIN: 1.73 / MAX: 1.77
OpenBenchmarking.orgFPS, More Is BetterOSPray 1.8.5Demo: XFrog Forest - Renderer: Path TracerThreadripper 3970X16 Cores246810Min: 4.69 / Avg: 4.69 / Max: 4.69Min: 1.75 / Avg: 1.75 / Max: 1.75

Flexible IO Tester

Fio is an advanced disk benchmark that depends upon the kernel's AIO access library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIOPS, More Is BetterFlexible IO Tester 3.18Type: Sequential Read - Engine: IO_uring - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test DirectoryThreadripper 3970X16 Cores90K180K270K360K450KSE +/- 3928.97, N = 15SE +/- 7133.29, N = 153445333976001. (CC) gcc options: -rdynamic -std=gnu99 -ffast-math -include -O3 -U_FORTIFY_SOURCE -march=native -ll -lcurl -lssl -lcrypto -lnuma -libverbs -lrt -laio -lz -lpthread -lm -ldl
OpenBenchmarking.orgIOPS, More Is BetterFlexible IO Tester 3.18Type: Sequential Read - Engine: IO_uring - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test DirectoryThreadripper 3970X16 Cores70K140K210K280K350KMin: 324000 / Avg: 344533.33 / Max: 367000Min: 371000 / Avg: 397600 / Max: 4420001. (CC) gcc options: -rdynamic -std=gnu99 -ffast-math -include -O3 -U_FORTIFY_SOURCE -march=native -ll -lcurl -lssl -lcrypto -lnuma -libverbs -lrt -laio -lz -lpthread -lm -ldl

OpenBenchmarking.orgMB/s, More Is BetterFlexible IO Tester 3.18Type: Sequential Read - Engine: IO_uring - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test DirectoryThreadripper 3970X16 Cores30060090012001500SE +/- 15.41, N = 15SE +/- 27.78, N = 15134615531. (CC) gcc options: -rdynamic -std=gnu99 -ffast-math -include -O3 -U_FORTIFY_SOURCE -march=native -ll -lcurl -lssl -lcrypto -lnuma -libverbs -lrt -laio -lz -lpthread -lm -ldl
OpenBenchmarking.orgMB/s, More Is BetterFlexible IO Tester 3.18Type: Sequential Read - Engine: IO_uring - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test DirectoryThreadripper 3970X16 Cores30060090012001500Min: 1264 / Avg: 1345.6 / Max: 1433Min: 1448 / Avg: 1553 / Max: 17251. (CC) gcc options: -rdynamic -std=gnu99 -ffast-math -include -O3 -U_FORTIFY_SOURCE -march=native -ll -lcurl -lssl -lcrypto -lnuma -libverbs -lrt -laio -lz -lpthread -lm -ldl

CP2K Molecular Dynamics

CP2K is an open-source molecular dynamics software package focused on quantum chemistry and solid-state physics. This test profile currently makes use of the OpenMP implementation and using the Fayalite-FIST molecular dynamics run and measures the total time to complete. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterCP2K Molecular Dynamics 6.1Fayalite-FIST DataThreadripper 3970X16 Cores110220330440550494.65423.75

Appleseed

Appleseed is an open-source production renderer focused on physically-based global illumination rendering engine primarily designed for animation and visual effects. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterAppleseed 2.0 BetaScene: EmilyThreadripper 3970X16 Cores60120180240300138.48289.31

OSPray

Intel OSPray is a portable ray-tracing engine for high-performance, high-fidenlity 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.orgFPS, More Is BetterOSPray 1.8.5Demo: San Miguel - Renderer: Path TracerThreadripper 3970X16 Cores0.82581.65162.47743.30324.129SE +/- 0.00, N = 4SE +/- 0.00, N = 33.671.37MIN: 3.65 / MAX: 3.72MIN: 1.36 / MAX: 1.38
OpenBenchmarking.orgFPS, More Is BetterOSPray 1.8.5Demo: San Miguel - Renderer: Path TracerThreadripper 3970X16 Cores246810Min: 3.66 / Avg: 3.67 / Max: 3.68Min: 1.37 / Avg: 1.37 / Max: 1.37

GNU GMP GMPbench

GMPbench is a test of the GMP 6.1.2 math library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGMPbench Score, More Is BetterGNU GMP GMPbench 6.1.2Total TimeThreadripper 3970X16 Cores120024003600480060005794.65767.11. (CC) gcc options: -O3 -fomit-frame-pointer -lm

MKL-DNN DNNL

This is a test of the Intel MKL-DNN (DNNL / Deep Neural Network Library) 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. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN DNNL 1.1Harness: Convolution Batch conv_googlenet_v3 - Data Type: f32Threadripper 3970X16 Cores20406080100SE +/- 0.89, N = 3SE +/- 0.24, N = 356.5876.35MIN: 41.14MIN: 75.021. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -fopenmp -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN DNNL 1.1Harness: Convolution Batch conv_googlenet_v3 - Data Type: f32Threadripper 3970X16 Cores1530456075Min: 55.24 / Avg: 56.58 / Max: 58.25Min: 76.08 / Avg: 76.35 / Max: 76.841. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -fopenmp -pie -lpthread -ldl

Mlpack Benchmark

Mlpack benchmark scripts for machine learning libraries Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterMlpack BenchmarkBenchmark: scikit_linearridgeregressionThreadripper 3970X16 Cores0.33080.66160.99241.32321.654SE +/- 0.02, N = 3SE +/- 0.02, N = 121.471.42
OpenBenchmarking.orgSeconds, Fewer Is BetterMlpack BenchmarkBenchmark: scikit_linearridgeregressionThreadripper 3970X16 Cores246810Min: 1.45 / Avg: 1.47 / Max: 1.5Min: 1.22 / Avg: 1.42 / Max: 1.45

Blender

Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing via OpenCL or CUDA is supported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.82Blend File: Fishy Cat - Compute: CPU-OnlyThreadripper 3970X16 Cores4080120160200SE +/- 0.08, N = 3SE +/- 0.09, N = 371.28178.46
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.82Blend File: Fishy Cat - Compute: CPU-OnlyThreadripper 3970X16 Cores306090120150Min: 71.11 / Avg: 71.28 / Max: 71.38Min: 178.27 / Avg: 178.46 / Max: 178.55

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: WriteThreadripper 3970X16 Cores7K14K21K28K35KSE +/- 14.14, N = 3SE +/- 29.01, N = 331832.5031631.79MIN: 27271.39 / MAX: 33182.75MIN: 27804.64 / MAX: 32902.031. (CC) gcc options: -lrt
OpenBenchmarking.orgMB/s, More Is BetterCacheBenchTest: WriteThreadripper 3970X16 Cores6K12K18K24K30KMin: 31804.84 / Avg: 31832.5 / Max: 31851.43Min: 31577.81 / Avg: 31631.79 / Max: 31677.21. (CC) gcc options: -lrt

OpenBenchmarking.orgMB/s, More Is BetterCacheBenchTest: ReadThreadripper 3970X16 Cores7001400210028003500SE +/- 4.12, N = 3SE +/- 0.65, N = 33035.033024.68MIN: 3020.86 / MAX: 3044.34MIN: 3022.65 / MAX: 3028.81. (CC) gcc options: -lrt
OpenBenchmarking.orgMB/s, More Is BetterCacheBenchTest: ReadThreadripper 3970X16 Cores5001000150020002500Min: 3026.79 / Avg: 3035.03 / Max: 3039.5Min: 3023.77 / Avg: 3024.68 / Max: 3025.931. (CC) gcc options: -lrt

OpenBenchmarking.orgMB/s, More Is BetterCacheBenchTest: Read / Modify / WriteThreadripper 3970X16 Cores13K26K39K52K65KSE +/- 167.50, N = 3SE +/- 147.71, N = 361162.9461009.39MIN: 53743.85 / MAX: 65830.07MIN: 55206.66 / MAX: 65416.91. (CC) gcc options: -lrt
OpenBenchmarking.orgMB/s, More Is BetterCacheBenchTest: Read / Modify / WriteThreadripper 3970X16 Cores11K22K33K44K55KMin: 60896.67 / Avg: 61162.94 / Max: 61472.13Min: 60714.03 / Avg: 61009.39 / Max: 61162.521. (CC) gcc options: -lrt

BYTE Unix Benchmark

This is a test of BYTE. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgLPS, More Is BetterBYTE Unix Benchmark 3.6Computational Test: Dhrystone 2Threadripper 3970X16 Cores10M20M30M40M50MSE +/- 385242.29, N = 3SE +/- 411120.29, N = 344416335.242231623.5
OpenBenchmarking.orgLPS, More Is BetterBYTE Unix Benchmark 3.6Computational Test: Dhrystone 2Threadripper 3970X16 Cores8M16M24M32M40MMin: 43703405 / Avg: 44416335.17 / Max: 45025847.3Min: 41627812.3 / Avg: 42231623.53 / Max: 43016870.3

asmFish

This is a test of asmFish, an advanced chess benchmark written in Assembly. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNodes/second, More Is BetterasmFish 2018-07-231024 Hash Memory, 26 DepthThreadripper 3970X16 Cores20M40M60M80M100MSE +/- 334321.64, N = 3SE +/- 159073.11, N = 39591619737415509
OpenBenchmarking.orgNodes/second, More Is BetterasmFish 2018-07-231024 Hash Memory, 26 DepthThreadripper 3970X16 Cores17M34M51M68M85MMin: 95317633 / Avg: 95916197 / Max: 96473556Min: 37098054 / Avg: 37415509.33 / Max: 37592385

Blender

Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing is supported. This system/blender test profile makes use of the system-supplied Blender. Use pts/blender if wishing to stick to a fixed version of Blender. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.80Blend File: Fishy Cat - Compute: CPU-OnlyThreadripper 3970X16 Cores4080120160200SE +/- 0.05, N = 3SE +/- 0.04, N = 367.04163.13
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.80Blend File: Fishy Cat - Compute: CPU-OnlyThreadripper 3970X16 Cores306090120150Min: 66.98 / Avg: 67.04 / Max: 67.14Min: 163.07 / Avg: 163.13 / Max: 163.21

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.orgdays/ns, Fewer Is BetterNAMD 2.13ATPase Simulation - 327,506 AtomsThreadripper 3970X16 Cores0.27880.55760.83641.11521.394SE +/- 0.00773, N = 15SE +/- 0.00167, N = 30.623571.23901
OpenBenchmarking.orgdays/ns, Fewer Is BetterNAMD 2.13ATPase Simulation - 327,506 AtomsThreadripper 3970X16 Cores246810Min: 0.58 / Avg: 0.62 / Max: 0.7Min: 1.24 / Avg: 1.24 / Max: 1.24

Fhourstones

This integer benchmark solves positions in the game of Connect-4, as played on a vertical 7x6 board. By default, it uses a 64Mb transposition table with the twobig replacement strategy. Positions are represented as 64-bit bitboards, and the hash function is computed using a single 64-bit modulo operation, giving 64-bit machines a slight edge. The alpha-beta searcher sorts moves dynamically based on the history heuristic. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgKpos / sec, More Is BetterFhourstones 3.1Complex Connect-4 SolvingThreadripper 3970X16 Cores3K6K9K12K15KSE +/- 35.91, N = 3SE +/- 53.43, N = 314702.315067.41. (CC) gcc options: -O3
OpenBenchmarking.orgKpos / sec, More Is BetterFhourstones 3.1Complex Connect-4 SolvingThreadripper 3970X16 Cores3K6K9K12K15KMin: 14651.9 / Avg: 14702.3 / Max: 14771.8Min: 14962 / Avg: 15067.37 / Max: 15135.51. (CC) gcc options: -O3

Apache Cassandra

This is a benchmark of the Apache Cassandra NoSQL database management system making use of cassandra-stress. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterApache Cassandra 3.11.4Test: WritesThreadripper 3970X16 Cores60K120K180K240K300KSE +/- 3852.60, N = 5SE +/- 21.67, N = 32930037744
OpenBenchmarking.orgOp/s, More Is BetterApache Cassandra 3.11.4Test: WritesThreadripper 3970X16 Cores50K100K150K200K250KMin: 277748 / Avg: 293003.2 / Max: 298817Min: 7711 / Avg: 7744.33 / Max: 7785

Minion

Minion is an open-source constraint solver that is designed to be very scalable. This test profile uses Minion's integrated benchmarking problems to solve. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterMinion 1.8Benchmark: QuasigroupThreadripper 3970X16 Cores20406080100SE +/- 0.37, N = 3SE +/- 0.16, N = 3107.53108.521. (CXX) g++ options: -std=gnu++11 -O3 -fomit-frame-pointer -rdynamic
OpenBenchmarking.orgSeconds, Fewer Is BetterMinion 1.8Benchmark: QuasigroupThreadripper 3970X16 Cores20406080100Min: 106.79 / Avg: 107.53 / Max: 107.95Min: 108.26 / Avg: 108.52 / Max: 108.821. (CXX) g++ options: -std=gnu++11 -O3 -fomit-frame-pointer -rdynamic

LevelDB

LevelDB is a key-value storage library developed by Google that supports making use of Snappy for data compression and has other modern features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: Hot ReadThreadripper 3970X16 Cores1020304050SE +/- 0.383, N = 11SE +/- 0.106, N = 1544.2868.9741. (CXX) g++ options: -O3 -lsnappy -lpthread
OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: Hot ReadThreadripper 3970X16 Cores918273645Min: 40.66 / Avg: 44.29 / Max: 45.32Min: 8.44 / Avg: 8.97 / Max: 9.511. (CXX) g++ options: -O3 -lsnappy -lpthread

Appleseed

Appleseed is an open-source production renderer focused on physically-based global illumination rendering engine primarily designed for animation and visual effects. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterAppleseed 2.0 BetaScene: Material TesterThreadripper 3970X16 Cores4080120160200131.52176.69

Facebook RocksDB

This is a benchmark of 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 BetterFacebook RocksDB 6.3.6Test: Read While WritingThreadripper 3970X16 Cores1.3M2.6M3.9M5.2M6.5MSE +/- 71829.99, N = 6SE +/- 32658.12, N = 4620407623813601. (CXX) g++ options: -O3 -march=native -std=c++11 -fno-builtin-memcmp -fno-rtti -rdynamic -lpthread
OpenBenchmarking.orgOp/s, More Is BetterFacebook RocksDB 6.3.6Test: Read While WritingThreadripper 3970X16 Cores1.1M2.2M3.3M4.4M5.5MMin: 6016980 / Avg: 6204075.83 / Max: 6533182Min: 2316596 / Avg: 2381359.75 / Max: 24715551. (CXX) g++ options: -O3 -march=native -std=c++11 -fno-builtin-memcmp -fno-rtti -rdynamic -lpthread

Memcached mcperf

This is a test of twmperf/mcperf with memcached. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOperations Per Second, More Is BetterMemcached mcperf 1.5.10Method: AppendThreadripper 3970X16 Cores20K40K60K80K100KSE +/- 1452.18, N = 12SE +/- 4614.05, N = 1549940.585868.91. (CC) gcc options: -O2 -lm -rdynamic
OpenBenchmarking.orgOperations Per Second, More Is BetterMemcached mcperf 1.5.10Method: AppendThreadripper 3970X16 Cores15K30K45K60K75KMin: 45727.2 / Avg: 49940.47 / Max: 60306.5Min: 64815.8 / Avg: 85868.95 / Max: 128295.21. (CC) gcc options: -O2 -lm -rdynamic

Flexible IO Tester

Fio is an advanced disk benchmark that depends upon the kernel's AIO access library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIOPS, More Is BetterFlexible IO Tester 3.18Type: Sequential Write - Engine: IO_uring - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test DirectoryThreadripper 3970X16 Cores60K120K180K240K300KSE +/- 10149.45, N = 15SE +/- 3424.79, N = 42692002457501. (CC) gcc options: -rdynamic -std=gnu99 -ffast-math -include -O3 -U_FORTIFY_SOURCE -march=native -ll -lcurl -lssl -lcrypto -lnuma -libverbs -lrt -laio -lz -lpthread -lm -ldl
OpenBenchmarking.orgIOPS, More Is BetterFlexible IO Tester 3.18Type: Sequential Write - Engine: IO_uring - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test DirectoryThreadripper 3970X16 Cores50K100K150K200K250KMin: 206000 / Avg: 269200 / Max: 352000Min: 236000 / Avg: 245750 / Max: 2520001. (CC) gcc options: -rdynamic -std=gnu99 -ffast-math -include -O3 -U_FORTIFY_SOURCE -march=native -ll -lcurl -lssl -lcrypto -lnuma -libverbs -lrt -laio -lz -lpthread -lm -ldl

OpenBenchmarking.orgMB/s, More Is BetterFlexible IO Tester 3.18Type: Sequential Write - Engine: IO_uring - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test DirectoryThreadripper 3970X16 Cores2004006008001000SE +/- 12.94, N = 410529611. (CC) gcc options: -rdynamic -std=gnu99 -ffast-math -include -O3 -U_FORTIFY_SOURCE -march=native -ll -lcurl -lssl -lcrypto -lnuma -libverbs -lrt -laio -lz -lpthread -lm -ldl
OpenBenchmarking.orgMB/s, More Is BetterFlexible IO Tester 3.18Type: Sequential Write - Engine: IO_uring - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test DirectoryThreadripper 3970X16 Cores2004006008001000Min: 924 / Avg: 960.5 / Max: 9851. (CC) gcc options: -rdynamic -std=gnu99 -ffast-math -include -O3 -U_FORTIFY_SOURCE -march=native -ll -lcurl -lssl -lcrypto -lnuma -libverbs -lrt -laio -lz -lpthread -lm -ldl

Apache HBase

This is a benchmark of the Apache HBase non-relational distributed database system inspired from Google's Bigtable. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMicroseconds - Average Latency, Fewer Is BetterApache HBase 2.2.3Test: Random Write - Clients: 16Threadripper 3970X16 Cores20406080100SE +/- 20.61, N = 15SE +/- 29.55, N = 157281
OpenBenchmarking.orgMicroseconds - Average Latency, Fewer Is BetterApache HBase 2.2.3Test: Random Write - Clients: 16Threadripper 3970X16 Cores1530456075Min: 21 / Avg: 72.2 / Max: 266Min: 25 / Avg: 80.73 / Max: 369

OpenBenchmarking.orgRows Per Second, More Is BetterApache HBase 2.2.3Test: Random Write - Clients: 16Threadripper 3970X16 Cores100K200K300K400K500KSE +/- 64718.62, N = 15SE +/- 57800.53, N = 15418033475673
OpenBenchmarking.orgRows Per Second, More Is BetterApache HBase 2.2.3Test: Random Write - Clients: 16Threadripper 3970X16 Cores80K160K240K320K400KMin: 59893 / Avg: 418033.47 / Max: 710472Min: 43338 / Avg: 475673 / Max: 636195

Mlpack Benchmark

Mlpack benchmark scripts for machine learning libraries Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterMlpack BenchmarkBenchmark: scikit_qdaThreadripper 3970X16 Cores1224364860SE +/- 0.42, N = 3SE +/- 0.21, N = 353.0234.16
OpenBenchmarking.orgSeconds, Fewer Is BetterMlpack BenchmarkBenchmark: scikit_qdaThreadripper 3970X16 Cores1122334455Min: 52.37 / Avg: 53.02 / Max: 53.81Min: 33.87 / Avg: 34.16 / Max: 34.56

Cpuminer-Opt

Cpuminer benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgkH/s - Hash Speed, More Is BetterCpuminer-Opt 3.8.8.1Algorithm: lbryThreadripper 3970X16 Cores20K40K60K80K100KSE +/- 464.63, N = 3SE +/- 330.05, N = 1589853410631. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp
OpenBenchmarking.orgkH/s - Hash Speed, More Is BetterCpuminer-Opt 3.8.8.1Algorithm: lbryThreadripper 3970X16 Cores16K32K48K64K80KMin: 89330 / Avg: 89853.33 / Max: 90780Min: 39950 / Avg: 41063.33 / Max: 446401. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

Flexible IO Tester

Fio is an advanced disk benchmark that depends upon the kernel's AIO access library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIOPS, More Is BetterFlexible IO Tester 3.18Type: Random Read - Engine: IO_uring - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test DirectoryThreadripper 3970X16 Cores70K140K210K280K350KSE +/- 881.92, N = 3SE +/- 2820.78, N = 152943333157331. (CC) gcc options: -rdynamic -std=gnu99 -ffast-math -include -O3 -U_FORTIFY_SOURCE -march=native -ll -lcurl -lssl -lcrypto -lnuma -libverbs -lrt -laio -lz -lpthread -lm -ldl
OpenBenchmarking.orgIOPS, More Is BetterFlexible IO Tester 3.18Type: Random Read - Engine: IO_uring - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test DirectoryThreadripper 3970X16 Cores50K100K150K200K250KMin: 293000 / Avg: 294333.33 / Max: 296000Min: 296000 / Avg: 315733.33 / Max: 3310001. (CC) gcc options: -rdynamic -std=gnu99 -ffast-math -include -O3 -U_FORTIFY_SOURCE -march=native -ll -lcurl -lssl -lcrypto -lnuma -libverbs -lrt -laio -lz -lpthread -lm -ldl

OpenBenchmarking.orgMB/s, More Is BetterFlexible IO Tester 3.18Type: Random Read - Engine: IO_uring - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test DirectoryThreadripper 3970X16 Cores30060090012001500SE +/- 2.91, N = 3SE +/- 10.95, N = 15115112341. (CC) gcc options: -rdynamic -std=gnu99 -ffast-math -include -O3 -U_FORTIFY_SOURCE -march=native -ll -lcurl -lssl -lcrypto -lnuma -libverbs -lrt -laio -lz -lpthread -lm -ldl
OpenBenchmarking.orgMB/s, More Is BetterFlexible IO Tester 3.18Type: Random Read - Engine: IO_uring - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test DirectoryThreadripper 3970X16 Cores2004006008001000Min: 1146 / Avg: 1150.67 / Max: 1156Min: 1157 / Avg: 1233.67 / Max: 12921. (CC) gcc options: -rdynamic -std=gnu99 -ffast-math -include -O3 -U_FORTIFY_SOURCE -march=native -ll -lcurl -lssl -lcrypto -lnuma -libverbs -lrt -laio -lz -lpthread -lm -ldl

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.07.26Test: SemaphoresThreadripper 3970X16 Cores900K1800K2700K3600K4500KSE +/- 11364.25, N = 3SE +/- 35197.46, N = 153379188.343998333.221. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.07.26Test: SemaphoresThreadripper 3970X16 Cores700K1400K2100K2800K3500KMin: 3359880.53 / Avg: 3379188.34 / Max: 3399227.07Min: 3858228.77 / Avg: 3998333.22 / Max: 4276272.991. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.07.26Test: LsearchThreadripper 3970X16 Cores90180270360450SE +/- 3.61, N = 3SE +/- 2.93, N = 15434.81138.271. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.07.26Test: LsearchThreadripper 3970X16 Cores80160240320400Min: 431.03 / Avg: 434.81 / Max: 442.03Min: 123.82 / Avg: 138.27 / Max: 150.191. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.07.26Test: Context SwitchingThreadripper 3970X16 Cores5M10M15M20M25MSE +/- 434072.66, N = 15SE +/- 9547.22, N = 321267958.312258922.931. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.07.26Test: Context SwitchingThreadripper 3970X16 Cores4M8M12M16M20MMin: 18229405.93 / Avg: 21267958.31 / Max: 24679470.18Min: 2246856.31 / Avg: 2258922.93 / Max: 2277772.091. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.07.26Test: System V Message PassingThreadripper 3970X16 Cores6M12M18M24M30MSE +/- 271951.30, N = 15SE +/- 23221.11, N = 326168429.6916301159.731. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc
OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.07.26Test: System V Message PassingThreadripper 3970X16 Cores5M10M15M20M25MMin: 24619139.59 / Avg: 26168429.69 / Max: 27792584.97Min: 16274161.55 / Avg: 16301159.73 / Max: 16347384.711. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc

Tensorflow

This is a benchmark of the Tensorflow deep learning framework using the CIFAR10 data set. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTensorflowBuild: Cifar10Threadripper 3970X16 Cores1020304050SE +/- 0.53, N = 5SE +/- 0.07, N = 342.0245.44
OpenBenchmarking.orgSeconds, Fewer Is BetterTensorflowBuild: Cifar10Threadripper 3970X16 Cores918273645Min: 40.53 / Avg: 42.02 / Max: 42.96Min: 45.29 / Avg: 45.44 / Max: 45.52

AOBench

AOBench is a lightweight ambient occlusion renderer, written in C. The test profile is using a size of 2048 x 2048. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterAOBenchSize: 2048 x 2048 - Total TimeThreadripper 3970X16 Cores714212835SE +/- 0.27, N = 14SE +/- 0.05, N = 331.7131.731. (CC) gcc options: -lm -O3
OpenBenchmarking.orgSeconds, Fewer Is BetterAOBenchSize: 2048 x 2048 - Total TimeThreadripper 3970X16 Cores714212835Min: 30.87 / Avg: 31.71 / Max: 35.16Min: 31.64 / Avg: 31.73 / Max: 31.811. (CC) gcc options: -lm -O3

GEGL

GEGL is the Generic Graphics Library and is the library/framework used by GIMP and other applications like GNOME Photos. This test profile times how long it takes to complete various GEGL operations on a static set of sample JPEG images. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterGEGLOperation: CartoonThreadripper 3970X16 Cores20406080100SE +/- 0.05, N = 3SE +/- 0.08, N = 390.9588.68
OpenBenchmarking.orgSeconds, Fewer Is BetterGEGLOperation: CartoonThreadripper 3970X16 Cores20406080100Min: 90.84 / Avg: 90.95 / Max: 91.01Min: 88.56 / Avg: 88.68 / Max: 88.82

PostgreSQL pgbench

This is a simple benchmark of PostgreSQL using pgbench. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 12.0Scaling: Buffer Test - Test: Normal Load - Mode: Read OnlyThreadripper 3970X16 Cores200K400K600K800K1000KSE +/- 3547.45, N = 3SE +/- 1543.39, N = 3792360.46272392.971. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lpthread -lrt -lcrypt -ldl -lm
OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 12.0Scaling: Buffer Test - Test: Normal Load - Mode: Read OnlyThreadripper 3970X16 Cores140K280K420K560K700KMin: 785641.93 / Avg: 792360.46 / Max: 797694.36Min: 269319.33 / Avg: 272392.97 / Max: 274176.151. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lpthread -lrt -lcrypt -ldl -lm

Facebook RocksDB

This is a benchmark of 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 BetterFacebook RocksDB 6.3.6Test: Sequential FillThreadripper 3970X16 Cores200K400K600K800K1000KSE +/- 3258.34, N = 3SE +/- 8372.80, N = 1511479757028091. (CXX) g++ options: -O3 -march=native -std=c++11 -fno-builtin-memcmp -fno-rtti -rdynamic -lpthread
OpenBenchmarking.orgOp/s, More Is BetterFacebook RocksDB 6.3.6Test: Sequential FillThreadripper 3970X16 Cores200K400K600K800K1000KMin: 1142352 / Avg: 1147975.33 / Max: 1153639Min: 648879 / Avg: 702809.33 / Max: 7672051. (CXX) g++ options: -O3 -march=native -std=c++11 -fno-builtin-memcmp -fno-rtti -rdynamic -lpthread

LULESH

LULESH is the Livermore Unstructured Lagrangian Explicit Shock Hydrodynamics. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgz/s, More Is BetterLULESH 2.0.3Threadripper 3970X16 Cores3691215SE +/- 0.0012891, N = 3SE +/- 0.0215340, N = 39.390898710.90211501. (CXX) g++ options: -O3 -fopenmp -lm -pthread -lmpi_cxx -lmpi
OpenBenchmarking.orgz/s, More Is BetterLULESH 2.0.3Threadripper 3970X16 Cores3691215Min: 9.39 / Avg: 9.39 / Max: 9.39Min: 10.86 / Avg: 10.9 / Max: 10.931. (CXX) g++ options: -O3 -fopenmp -lm -pthread -lmpi_cxx -lmpi

Blender

Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing via OpenCL or CUDA is supported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.82Blend File: BMW27 - Compute: CPU-OnlyThreadripper 3970X16 Cores306090120150SE +/- 0.09, N = 3SE +/- 0.17, N = 347.57119.58
OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.82Blend File: BMW27 - Compute: CPU-OnlyThreadripper 3970X16 Cores20406080100Min: 47.47 / Avg: 47.57 / Max: 47.76Min: 119.35 / Avg: 119.58 / Max: 119.92

OSPray

Intel OSPray is a portable ray-tracing engine for high-performance, high-fidenlity 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.orgFPS, More Is BetterOSPray 1.8.5Demo: San Miguel - Renderer: SciVisThreadripper 3970X16 Cores1122334455SE +/- 0.00, N = 1250.0018.52MIN: 43.48 / MAX: 52.63MIN: 17.86 / MAX: 19.61
OpenBenchmarking.orgFPS, More Is BetterOSPray 1.8.5Demo: San Miguel - Renderer: SciVisThreadripper 3970X16 Cores1020304050Min: 18.52 / Avg: 18.52 / Max: 18.52

Memtier_benchmark

Memtier_benchmark is a NoSQL Redis/Memcache traffic generation plus benchmarking tool. This current test profile currently just stresses the Redis protocol and basic options exposed wotj a 1:1 Set/Get ratio, 30 pipeline, 100 clients per thread, and thread count equal to the number of CPU cores/threads present. Patches to extend the test are welcome as always. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps/sec, More Is BetterMemtier_benchmark 1.2.17Protocol: RedisThreadripper 3970X16 Cores500K1000K1500K2000K2500KSE +/- 34708.07, N = 4SE +/- 8570.47, N = 32351667.582272891.421. (CXX) g++ options: -O2 -levent -lpthread -lz -lpcre
OpenBenchmarking.orgOps/sec, More Is BetterMemtier_benchmark 1.2.17Protocol: RedisThreadripper 3970X16 Cores400K800K1200K1600K2000KMin: 2298344.7 / Avg: 2351667.58 / Max: 2451345.17Min: 2260131.18 / Avg: 2272891.42 / Max: 2289183.21. (CXX) g++ options: -O2 -levent -lpthread -lz -lpcre

Timed GDB GNU Debugger Compilation

This test times how long it takes to build the GNU Debugger (GDB) in a default configuration. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed GDB GNU Debugger Compilation 9.1Time To CompileThreadripper 3970X16 Cores20406080100SE +/- 0.06, N = 3SE +/- 0.03, N = 383.1282.19
OpenBenchmarking.orgSeconds, Fewer Is BetterTimed GDB GNU Debugger Compilation 9.1Time To CompileThreadripper 3970X16 Cores1632486480Min: 83 / Avg: 83.12 / Max: 83.2Min: 82.15 / Avg: 82.19 / Max: 82.25

MKL-DNN DNNL

This is a test of the Intel MKL-DNN (DNNL / Deep Neural Network Library) 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. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN DNNL 1.1Harness: Recurrent Neural Network Training - Data Type: f32Threadripper 3970X16 Cores70140210280350SE +/- 2.80, N = 10SE +/- 0.28, N = 3299.86142.35MIN: 188.07MIN: 140.831. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -fopenmp -pie -lpthread -ldl
OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN DNNL 1.1Harness: Recurrent Neural Network Training - Data Type: f32Threadripper 3970X16 Cores50100150200250Min: 278.71 / Avg: 299.86 / Max: 309.85Min: 141.93 / Avg: 142.35 / Max: 142.871. (CXX) g++ options: -O3 -march=native -std=c++11 -msse4.1 -fPIC -fopenmp -pie -lpthread -ldl

Intel MPI Benchmarks

Intel MPI Benchmarks for stressing MPI implementations. At this point the test profile aggregates results for some common MPI functionality. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgAverage usec, More Is BetterIntel MPI Benchmarks 2019.3Test: IMB-MPI1 SendrecvThreadripper 3970X16 Cores20406080100SE +/- 0.48, N = 3SE +/- 0.45, N = 1582.6163.49MIN: 0.38 / MAX: 2576.08MIN: 0.19 / MAX: 1638.641. (CXX) g++ options: -O0 -pedantic -fopenmp -pthread -lmpi_cxx -lmpi
OpenBenchmarking.orgAverage usec, More Is BetterIntel MPI Benchmarks 2019.3Test: IMB-MPI1 SendrecvThreadripper 3970X16 Cores1632486480Min: 81.7 / Avg: 82.61 / Max: 83.32Min: 60.37 / Avg: 63.49 / Max: 67.71. (CXX) g++ options: -O0 -pedantic -fopenmp -pthread -lmpi_cxx -lmpi

OpenBenchmarking.orgAverage Mbytes/sec, More Is BetterIntel MPI Benchmarks 2019.3Test: IMB-MPI1 SendrecvThreadripper 3970X16 Cores13002600390052006500SE +/- 64.08, N = 3SE +/- 114.95, N = 154955.006263.73MAX: 30545.39MAX: 41562.161. (CXX) g++ options: -O0 -pedantic -fopenmp -pthread -lmpi_cxx -lmpi
OpenBenchmarking.orgAverage Mbytes/sec, More Is BetterIntel MPI Benchmarks 2019.3Test: IMB-MPI1 SendrecvThreadripper 3970X16 Cores11002200330044005500Min: 4833.79 / Avg: 4955 / Max: 5051.66Min: 5590.91 / Avg: 6263.73 / Max: 6875.151. (CXX) g++ options: -O0 -pedantic -fopenmp -pthread -lmpi_cxx -lmpi

LevelDB

LevelDB is a key-value storage library developed by Google that supports making use of Snappy for data compression and has other modern features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: Random ReadThreadripper 3970X16 Cores1020304050SE +/- 0.404, N = 10SE +/- 0.093, N = 344.3539.3131. (CXX) g++ options: -O3 -lsnappy -lpthread
OpenBenchmarking.orgMicroseconds Per Op, Fewer Is BetterLevelDB 1.22Benchmark: Random ReadThreadripper 3970X16 Cores918273645Min: 40.8 / Avg: 44.35 / Max: 45.32Min: 9.18 / Avg: 9.31 / Max: 9.491. (CXX) g++ options: -O3 -lsnappy -lpthread

CppPerformanceBenchmarks

CppPerformanceBenchmarks is a set of C++ compiler performance benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterCppPerformanceBenchmarks 9Test: Stepanov VectorThreadripper 3970X16 Cores20406080100SE +/- 0.36, N = 3SE +/- 0.37, N = 377.1978.891. (CXX) g++ options: -std=c++11 -O3
OpenBenchmarking.orgSeconds, Fewer Is BetterCppPerformanceBenchmarks 9Test: Stepanov VectorThreadripper 3970X16 Cores1530456075Min: 76.49 / Avg: 77.19 / Max: 77.66Min: 78.15 / Avg: 78.89 / Max: 79.311. (CXX) g++ options: -std=c++11 -O3

Cpuminer-Opt

Cpuminer benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgkH/s - Hash Speed, More Is BetterCpuminer-Opt 3.8.8.1Algorithm: deepThreadripper 3970X16 Cores6K12K18K24K30KSE +/- 31.80, N = 3SE +/- 1488.63, N = 1529323.006879.911. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp
OpenBenchmarking.orgkH/s - Hash Speed, More Is BetterCpuminer-Opt 3.8.8.1Algorithm: deepThreadripper 3970X16 Cores5K10K15K20K25KMin: 29260 / Avg: 29323.33 / Max: 29360Min: 44.04 / Avg: 6879.91 / Max: 118401. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

OpenBenchmarking.orgkH/s - Hash Speed, More Is BetterCpuminer-Opt 3.8.8.1Algorithm: myr-grThreadripper 3970X16 Cores3K6K9K12K15KSE +/- 71.26, N = 3SE +/- 1707.65, N = 1513923.0015369.091. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp
OpenBenchmarking.orgkH/s - Hash Speed, More Is BetterCpuminer-Opt 3.8.8.1Algorithm: myr-grThreadripper 3970X16 Cores3K6K9K12K15KMin: 13820 / Avg: 13923.33 / Max: 14060Min: 2621.41 / Avg: 15369.09 / Max: 215901. (CXX) g++ options: -O2 -lcurl -lz -lpthread -lssl -lcrypto -lgmp

Flexible IO Tester

Fio is an advanced disk benchmark that depends upon the kernel's AIO access library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIOPS, More Is BetterFlexible IO Tester 3.18Type: Random Write - Engine: IO_uring - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test DirectoryThreadripper 3970X16 Cores70K140K210K280K350KSE +/- 26330.47, N = 121806333350001. (CC) gcc options: -rdynamic -std=gnu99 -ffast-math -include -O3 -U_FORTIFY_SOURCE -march=native -ll -lcurl -lssl -lcrypto -lnuma -libverbs -lrt -laio -lz -lpthread -lm -ldl
OpenBenchmarking.orgIOPS, More Is BetterFlexible IO Tester 3.18Type: Random Write - Engine: IO_uring - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test DirectoryThreadripper 3970X16 Cores60K120K180K240K300KMin: 34300 / Avg: 180633.33 / Max: 2950001. (CC) gcc options: -rdynamic -std=gnu99 -ffast-math -include -O3 -U_FORTIFY_SOURCE -march=native -ll -lcurl -lssl -lcrypto -lnuma -libverbs -lrt -laio -lz -lpthread -lm -ldl

OpenBenchmarking.orgMB/s, More Is BetterFlexible IO Tester 3.18Type: Random Write - Engine: IO_uring - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test DirectoryThreadripper 3970X16 Cores30060090012001500SE +/- 102.96, N = 1270613091. (CC) gcc options: -rdynamic -std=gnu99 -ffast-math -include -O3 -U_FORTIFY_SOURCE -march=native -ll -lcurl -lssl -lcrypto -lnuma -libverbs -lrt -laio -lz -lpthread -lm -ldl
OpenBenchmarking.orgMB/s, More Is BetterFlexible IO Tester 3.18Type: Random Write - Engine: IO_uring - Buffered: No - Direct: Yes - Block Size: 4KB - Disk Target: Default Test DirectoryThreadripper 3970X16 Cores2004006008001000Min: 134 / Avg: 705.75 / Max: 11541. (CC) gcc options: -rdynamic -std=gnu99 -ffast-math -include -O3 -U_FORTIFY_SOURCE -march=native -ll -lcurl -lssl -lcrypto -lnuma -libverbs -lrt -laio -lz -lpthread -lm -ldl

OpenBenchmarking.orgIOPS, More Is BetterFlexible IO TesterType: Random Write - IO Engine: POSIX AIO - Buffered: Yes - Direct: No - Block Size: 4KB - Disk Target: Default Test DirectoryThreadripper 3970X16 Cores60K120K180K240K300KSE +/- 13593.30, N = 12SE +/- 5033.22, N = 3263667277000
OpenBenchmarking.orgIOPS, More Is BetterFlexible IO TesterType: Random Write - IO Engine: POSIX AIO - Buffered: Yes - Direct: No - Block Size: 4KB - Disk Target: Default Test DirectoryThreadripper 3970X16 Cores50K100K150K200K250KMin: 117000 / Avg: 263666.67 / Max: 290000Min: 271000 / Avg: 277000 / Max: 287000

DaCapo Benchmark

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

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 9.12-MR1Java Test: TradebeansThreadripper 3970X16 Cores10002000300040005000SE +/- 10.90, N = 4SE +/- 14.45, N = 444962452
OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 9.12-MR1Java Test: TradebeansThreadripper 3970X16 Cores8001600240032004000Min: 4465 / Avg: 4496 / Max: 4515Min: 2413 / Avg: 2451.5 / Max: 2483

GROMACS

The GROMACS (GROningen MAchine for Chemical Simulations) molecular dynamics package testing on the CPU with the water_GMX50 data. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNs Per Day, More Is BetterGROMACS 2020Water BenchmarkThreadripper 3970X16 Cores0.73981.47962.21942.95923.699SE +/- 0.004, N = 3SE +/- 0.002, N = 33.2882.0481. (CXX) g++ options: -O3 -pthread -lrt -lpthread -lm
OpenBenchmarking.orgNs Per Day, More Is BetterGROMACS 2020Water BenchmarkThreadripper 3970X16 Cores246810Min: 3.28 / Avg: 3.29 / Max: 3.3Min: 2.05 / Avg: 2.05 / Max: 2.051. (CXX) g++ options: -O3 -pthread -lrt -lpthread -lm

Schbench

This is a benchmark of Schbench, a Linux kernel scheduler benchmark developed by Facebook. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgusec, 99.9th Latency Percentile, Fewer Is BetterSchbenchMessage Threads: 8 - Workers Per Message Thread: 8Threadripper 3970X16 Cores10K20K30K40K50KSE +/- 468.91, N = 12SE +/- 740.24, N = 33587463151. (CC) gcc options: -O2 -lpthread
OpenBenchmarking.orgusec, 99.9th Latency Percentile, Fewer Is BetterSchbenchMessage Threads: 8 - Workers Per Message Thread: 8Threadripper 3970X16 Cores8K16K24K32K40KMin: 2260 / Avg: 3586.67 / Max: 8400Min: 44864 / Avg: 46314.67 / Max: 47296