CentOS Stream 8

2 x AMD EPYC 7763 64-Core testing with a AMD DAYTONA_X (RYM1001D BIOS) and llvmpipe on CentOS Stream 8 via the Phoronix Test Suite.

HTML result view exported from: https://openbenchmarking.org/result/2112063-TJ-CENTOSSTR24&grs.

ProcessorMotherboardChipsetMemoryDiskGraphicsMonitorNetworkOSKernelDesktopDisplay ServerOpenGLCompilerFile-SystemScreen ResolutionXeon Platinum 8380 2PEPYC 7763 2P CentOS Stream 8 CentOS Stream 82 x Intel Xeon Platinum 8380 @ 3.40GHz (80 Cores / 160 Threads)Intel M50CYP2SB2U (SE5C6200.86B.0022.D08.2103221623 BIOS)Intel Device 0998504GB7682GB INTEL SSDPF2KX076TZllvmpipeVE2282 x Intel X710 for 10GBASE-T + 2 x Intel E810-C for QSFPCentOS Stream 84.18.0-348.2.1.el8_5.x86_64 (x86_64)GNOME Shell 3.32.2X Server4.5 Mesa 21.3.0 (LLVM 12.0.1 256 bits)GCC 8.5.0 20210514xfs1920x10802 x AMD EPYC 7763 64-Core @ 2.45GHz (128 Cores / 256 Threads)AMD DAYTONA_X (RYM1001D BIOS)AMD Starship/Matisse2 x Mellanox MT27710OpenBenchmarking.orgKernel Details- Transparent Huge Pages: alwaysCompiler Details- --build=x86_64-redhat-linux --disable-libmpx --disable-libunwind-exceptions --enable-__cxa_atexit --enable-bootstrap --enable-cet --enable-checking=release --enable-gnu-indirect-function --enable-gnu-unique-object --enable-initfini-array --enable-languages=c,c++,fortran,lto --enable-multilib --enable-offload-targets=nvptx-none --enable-plugin --enable-shared --enable-threads=posix --mandir=/usr/share/man --with-arch_32=x86-64 --with-gcc-major-version-only --with-isl --with-linker-hash-style=gnu --with-tune=generic --without-cuda-driver Processor Details- Xeon Platinum 8380 2P: CentOS Stream 8: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0xd0002a0- EPYC 7763 2P: CentOS Stream 8: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa001119Java Details- OpenJDK Runtime Environment (build 1.8.0_312-b07)Python Details- Python 3.6.8Security Details- Xeon Platinum 8380 2P: CentOS Stream 8: SELinux + 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 Enhanced IBRS IBPB: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected - EPYC 7763 2P: CentOS Stream 8: SELinux + 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 IBRS_FW STIBP: always-on RSB filling + srbds: Not affected + tsx_async_abort: Not affected

onednn: Recurrent Neural Network Training - u8s8f32 - CPUonednn: Matrix Multiply Batch Shapes Transformer - u8s8f32 - CPUonednn: Deconvolution Batch shapes_3d - u8s8f32 - CPUaskap: tConvolve MPI - Degriddingrenaissance: Akka Unbalanced Cobwebbed Treeonednn: Deconvolution Batch shapes_1d - u8s8f32 - CPUaskap: tConvolve MPI - Griddingstress-ng: Cryptostress-ng: CPU Stressnginx: 1000onednn: IP Shapes 3D - u8s8f32 - CPUnginx: 500stress-ng: Mallocstress-ng: MEMFDstress-ng: Vector Mathcryptsetup: AES-XTS 512b Decryptionnginx: 200cryptsetup: AES-XTS 512b Encryptioncompress-7zip: Decompression Ratingospray: Magnetic Reconnection - SciViscryptsetup: AES-XTS 256b Encryptioncryptsetup: AES-XTS 256b Decryptionstress-ng: Forkingstress-ng: Glibc C String Functionsappleseed: Material Testertensorflow-lite: NASNet Mobilerenaissance: In-Memory Database Shootoutblender: Pabellon Barcelona - CPU-Onlyembree: Pathtracer - Crownstress-ng: SENDFILEblender: Classroom - CPU-Onlycompress-7zip: Compression Ratingstockfish: Total Timerbenchmark: blender: BMW27 - CPU-Onlynpb: LU.Copenssl: openssl: redis: GETstress-ng: Semaphoresluxcorerender: Orange Juice - CPUembree: Pathtracer ISPC - Asian Dragonnpb: BT.Cnpb: SP.Cprimesieve: 1e12 Prime Number Generationredis: LPUSH and LPOP: lpopcompress-zstd: 19, Long Mode - Compression Speedstress-ng: Glibc Qsort Data Sortingcompress-zstd: 8, Long Mode - Compression Speedstress-ng: Memory Copyingapache: 500embree: Pathtracer ISPC - Crownmnn: inception-v3ospray: NASA Streamlines - Path Tracernpb: IS.Drav1e: 10stress-ng: Matrix Mathbuild-llvm: Ninjacompress-zstd: 19 - Decompression Speedluxcorerender: Danish Mood - CPUnpb: SP.Bcompress-zstd: 8, Long Mode - Decompression Speedcompress-zstd: 8 - Compression Speedcryptsetup: Serpent-XTS 256b Decryptioncryptsetup: Serpent-XTS 512b Decryptionbuild-nodejs: Time To Compileappleseed: Emilyospray: San Miguel - SciViscompress-zstd: 19, Long Mode - Decompression Speedsockperf: Latency Ping Pongmocassin: Dust 2D tau100.0compress-zstd: 8 - Decompression Speedappleseed: Disney Materialstress-ng: MMAPplaidml: No - Inference - VGG16 - CPUcryptsetup: PBKDF2-sha512stress-ng: NUMAcryptsetup: Serpent-XTS 256b Encryptioncryptsetup: Serpent-XTS 512b Encryptionpyperformance: python_startupnamd: ATPase Simulation - 327,506 Atomsx265: Bosphorus 4Kapache: 200octave-benchmark: rav1e: 6rav1e: 5ospray: NASA Streamlines - SciVisoidn: RT.ldr_alb_nrm.3840x2160renaissance: Apache Spark PageRankplaidml: No - Inference - VGG19 - CPUospray: San Miguel - Path Tracerembree: Pathtracer - Asian Dragonoidn: RT.hdr_alb_nrm.3840x2160phpbench: PHP Benchmark Suitebuild-linux-kernel: Time To Compiletensorflow-lite: Inception V4oidn: RTLightmap.hdr.4096x4096hpcg: dacapobench: Jythonpyperformance: nbodynpb: CG.Cdacapobench: Eclipsesvt-hevc: 7 - Bosphorus 1080pstress-ng: System V Message Passingsvt-vp9: PSNR/SSIM Optimized - Bosphorus 1080psvt-av1: Preset 8 - Bosphorus 4Kincompact3d: X3D-benchmarking input.i3dbuild-gcc: Time To Compilepyperformance: chaosbuild-llvm: Unix Makefilesgromacs: MPI CPU - water_GMX50_barecryptsetup: PBKDF2-whirlpoolapache: 1000pyperformance: 2to3svt-av1: Preset 4 - Bosphorus 4Ksockperf: Throughputpyperformance: floatgpaw: Carbon Nanotubecompress-zstd: 19 - Compression Speedcryptsetup: Twofish-XTS 256b Decryptioncryptsetup: Twofish-XTS 512b Decryptionsvt-hevc: 10 - Bosphorus 1080psvt-vp9: Visual Quality Optimized - Bosphorus 1080ppyperformance: json_loadsplaidml: No - Inference - ResNet 50 - CPUpyperformance: pickle_pure_pythonpyperformance: pathlibpyperformance: regex_compilecryptsetup: Twofish-XTS 512b Encryptionmt-dgemm: Sustained Floating-Point Ratecryptsetup: Twofish-XTS 256b Encryptionpyperformance: django_templatesvt-vp9: VMAF Optimized - Bosphorus 1080ppyperformance: crypto_pyaespybench: Total For Average Test Timespyperformance: raytraceincompact3d: input.i3d 193 Cells Per Directionpyperformance: golammps: 20k Atomsopencv: DNN - Deep Neural Networkstress-ng: RdRandospray: Magnetic Reconnection - Path Traceropencv: Object Detectionopencv: Features 2Donnx: super-resolution-10 - CPUonnx: shufflenet-v2-10 - CPUonnx: fcn-resnet101-11 - CPUonnx: yolov4 - CPUblender: Barbershop - CPU-Onlyncnn: CPU - regnety_400mncnn: CPU - squeezenet_ssdncnn: CPU - yolov4-tinyncnn: CPU - resnet50ncnn: CPU - alexnetncnn: CPU - resnet18ncnn: CPU - vgg16ncnn: CPU - googlenetncnn: CPU - blazefacencnn: CPU - efficientnet-b0ncnn: CPU - mnasnetncnn: CPU - shufflenet-v2ncnn: CPU-v3-v3 - mobilenet-v3ncnn: CPU-v2-v2 - mobilenet-v2ncnn: CPU - mobilenetmnn: mobilenet-v1-1.0mnn: MobileNetV2_224mnn: SqueezeNetV1.0mnn: resnet-v2-50mnn: squeezenetv1.1mnn: mobilenetV3redis: SETredis: MIXredis: LPUSH and LPOP: lpushstress-ng: Context Switchingstress-ng: Socket Activitystress-ng: CPU Cacheredis: SETredis: GETpgbench: 1000 - 500 - Read Write - Average Latencypgbench: 1000 - 500 - Read Writepgbench: 1000 - 250 - Read Write - Average Latencypgbench: 1000 - 250 - Read Writepgbench: 1000 - 500 - Read Only - Average Latencypgbench: 1000 - 500 - Read Onlypgbench: 1000 - 250 - Read Only - Average Latencypgbench: 1000 - 250 - Read Onlytensorflow-lite: Inception ResNet V2tensorflow-lite: Mobilenet Quanttensorflow-lite: Mobilenet Floattensorflow-lite: SqueezeNetaskap: Hogbom Clean OpenMPaskap: tConvolve OpenMP - Degriddingaskap: tConvolve OpenMP - Griddingaskap: tConvolve MT - Degriddingaskap: tConvolve MT - Griddingonednn: Recurrent Neural Network Inference - u8s8f32 - CPUonednn: Convolution Batch Shapes Auto - u8s8f32 - CPUonednn: IP Shapes 1D - u8s8f32 - CPUluxcorerender: Rainbow Colors and Prism - CPUluxcorerender: LuxCore Benchmark - CPUluxcorerender: DLSC - CPUdacapobench: Tradebeansdacapobench: H2lammps: Rhodopsin Proteinincompact3d: input.i3d 129 Cells Per Directionnpb: MG.Cnpb: FT.Cnpb: EP.Dnpb: EP.Csockperf: Latency Under LoadXeon Platinum 8380 2PEPYC 7763 2P CentOS Stream 8 CentOS Stream 8664.1140.2261580.18926617972.914749.80.36379419830.212276.06112721.70198662.900.435150196548.341446451541.162826.87225795.623223.8185945.293198.43686591253651.83637.263152.487757894.79192.41409574916.43801.377.5663.82271034645.1961.534779741800990730.394923.2339569.9517870.91182907.4782586.5612394981.2914.31108.231049340.4124243.783.498527747.4418.3978.75393.88486.63146927.6073.534722.90228.572029.927.347333297.17122.3452521.56.5123554.853001.22169.8560.8562.284.857122.7150361002674.35.229190284558.1349373281.9762.60153525010.00602.4602.610.50.2751927.30142666.946.6972.7972.078142.862.963035.352.2410.5383.57962.9363942921.2356725851.4040.6329361312322885.7516229299.174971692.52465.6858.515319.572357654.209127180.0108.654602629113997.003554.808327763120359.41067.1363.0363.3574.56359.8926.96.0155320.2208368.144.409275367.978.5476.33114116762112.073682530135.3619101669474.48500729931499131080410328474671246.8338.6924.4126.6127.0910.1114.8331.7323.196.8814.5011.1111.9611.7211.8220.731.9252.6694.61919.4343.8282.148309851.24464386.541126290.2611298849.843240.0417.631829369.292357157.421709.89125426810.23521541182.8609566548.13738227258089034025.332947.847883.01086.45118173.011254.812604.48372.03442.2071.196801.5014712.896.719.053367384333.4232.9735707444883.5138431.532597.652611.8117.3713033.040.7000870.54140245968.336819.30.82553644990.127068.88245332.5093923.860.90880294578.252941712927.331438.30439985.851668.296705.091666.569097267.201971.71968.034192.0614162266.61347.5198041352236735.144.71110.10861779916.2436.157922642970635990.240514.4863413.6928503.61851481.5500099.8419337664.1621.9471.032374487.1435030.132.449753633.0012.91346.44287.611617.46107388.17100.422531.18438.462701.615.527440217.9293.3223267.98.4130167.433816.02754.8711.2712.167.679153.4412311253340.94.2362343494.647.4897554001.0251.5118635998.25725.6724.812.60.2296622.97120100.535.6412.3681.768166.672.543519.245.0812.2072.15212.5355217018.7376044901.2636.7376399513625292.4817900329.564524150.60424.8763.948345.578715706.430137167.7999.282644885106659.273755.066345024126344.53469.8377.4376.5591.35350.6727.65.8656720.7213376.845.449912376.077.1467.87112115061312.190645530335.49250041104485193327180.32116.1473.1250.2798.4315.4736.4141.4465.9119.9058.6546.9435.1648.6650.9392.944.4605.1208.08935.9299.5066.876209135.79334361.81803762.1470077496.403235.8939.741463413.611763052.442048.089233341052.54719378184.337146223121.6687483865565050910.251660.259188.8257.2466070.395689.7013224.08717.811338.151.578921.5542317.178.6614.315456526429.4763.4498578548659.6239242.313175.993244.2216.892OpenBenchmarking.org

oneDNN

Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPU

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPUCentOS Stream 87001400210028003500SE +/- 2.02, N = 3SE +/- 32.78, N = 4664.113033.041. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

oneDNN

Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPU

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPUCentOS Stream 80.15750.3150.47250.630.7875SE +/- 0.001114, N = 3SE +/- 0.007446, N = 50.2261580.7000871. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

oneDNN

Harness: Deconvolution Batch shapes_3d - Data Type: u8s8f32 - Engine: CPU

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Deconvolution Batch shapes_3d - Data Type: u8s8f32 - Engine: CPUCentOS Stream 80.12180.24360.36540.48720.609SE +/- 0.000855, N = 3SE +/- 0.002376, N = 30.1892660.5414021. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

ASKAP

Test: tConvolve MPI - Degridding

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMpix/sec, More Is BetterASKAP 1.0Test: tConvolve MPI - DegriddingCentOS Stream 810K20K30K40K50KSE +/- 71.10, N = 3SE +/- 167.17, N = 317972.945968.31. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp

Renaissance

Test: Akka Unbalanced Cobwebbed Tree

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.12Test: Akka Unbalanced Cobwebbed TreeCentOS Stream 88K16K24K32K40KSE +/- 127.73, N = 9SE +/- 133.92, N = 314749.836819.3

oneDNN

Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPU

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPUCentOS Stream 80.18570.37140.55710.74280.9285SE +/- 0.000501, N = 3SE +/- 0.008999, N = 150.3637940.8255361. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

ASKAP

Test: tConvolve MPI - Gridding

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMpix/sec, More Is BetterASKAP 1.0Test: tConvolve MPI - GriddingCentOS Stream 810K20K30K40K50KSE +/- 131.68, N = 3SE +/- 423.01, N = 319830.244990.11. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp

Stress-NG

Test: Crypto

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: CryptoCentOS Stream 86K12K18K24K30KSE +/- 1.08, N = 3SE +/- 42.88, N = 312276.0627068.881. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc

Stress-NG

Test: CPU Stress

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: CPU StressCentOS Stream 850K100K150K200K250KSE +/- 149.97, N = 3SE +/- 376.22, N = 3112721.70245332.501. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc

nginx

Concurrent Requests: 1000

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.21.1Concurrent Requests: 1000CentOS Stream 840K80K120K160K200KSE +/- 415.61, N = 3SE +/- 437.18, N = 3198662.9093923.861. (CC) gcc options: -ldl -lpthread -lcrypt -lz -O3 -march=native

oneDNN

Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPU

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPUCentOS Stream 80.20450.4090.61350.8181.0225SE +/- 0.004759, N = 3SE +/- 0.007250, N = 150.4351500.9088021. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

nginx

Concurrent Requests: 500

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.21.1Concurrent Requests: 500CentOS Stream 840K80K120K160K200KSE +/- 392.84, N = 3SE +/- 533.88, N = 3196548.3494578.251. (CC) gcc options: -ldl -lpthread -lcrypt -lz -O3 -march=native

Stress-NG

Test: Malloc

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: MallocCentOS Stream 8600M1200M1800M2400M3000MSE +/- 3355046.59, N = 3SE +/- 2008786.61, N = 31446451541.162941712927.331. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc

Stress-NG

Test: MEMFD

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: MEMFDCentOS Stream 86001200180024003000SE +/- 2.47, N = 3SE +/- 2.31, N = 32826.871438.301. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc

Stress-NG

Test: Vector Math

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: Vector MathCentOS Stream 890K180K270K360K450KSE +/- 54.64, N = 3SE +/- 154.22, N = 3225795.62439985.851. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc

Cryptsetup

AES-XTS 512b Decryption

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMiB/s, More Is BetterCryptsetupAES-XTS 512b DecryptionCentOS Stream 87001400210028003500SE +/- 8.37, N = 3SE +/- 5.57, N = 33223.81668.2

nginx

Concurrent Requests: 200

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgRequests Per Second, More Is Betternginx 1.21.1Concurrent Requests: 200CentOS Stream 840K80K120K160K200KSE +/- 1210.93, N = 3SE +/- 211.37, N = 3185945.2996705.091. (CC) gcc options: -ldl -lpthread -lcrypt -lz -O3 -march=native

Cryptsetup

AES-XTS 512b Encryption

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMiB/s, More Is BetterCryptsetupAES-XTS 512b EncryptionCentOS Stream 87001400210028003500SE +/- 15.69, N = 3SE +/- 5.02, N = 33198.41666.5

7-Zip Compression

Test: Decompression Rating

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 21.06Test: Decompression RatingCentOS Stream 8150K300K450K600K750KSE +/- 441.97, N = 3SE +/- 7232.75, N = 43686596909721. (CXX) g++ options: -lpthread -ldl -O2 -fPIC

OSPray

Demo: Magnetic Reconnection - Renderer: SciVis

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgFPS, More Is BetterOSPray 1.8.5Demo: Magnetic Reconnection - Renderer: SciVisCentOS Stream 8306090120150SE +/- 0.53, N = 9125.0067.20

Cryptsetup

AES-XTS 256b Encryption

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMiB/s, More Is BetterCryptsetupAES-XTS 256b EncryptionCentOS Stream 88001600240032004000SE +/- 13.26, N = 3SE +/- 4.85, N = 33651.81971.7

Cryptsetup

AES-XTS 256b Decryption

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMiB/s, More Is BetterCryptsetupAES-XTS 256b DecryptionCentOS Stream 88001600240032004000SE +/- 22.87, N = 3SE +/- 6.88, N = 33637.21968.0

Stress-NG

Test: Forking

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: ForkingCentOS Stream 814K28K42K56K70KSE +/- 29.17, N = 3SE +/- 411.00, N = 463152.4834192.061. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc

Stress-NG

Test: Glibc C String Functions

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: Glibc C String FunctionsCentOS Stream 83M6M9M12M15MSE +/- 52776.39, N = 3SE +/- 40185.46, N = 37757894.7914162266.611. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc

Appleseed

Scene: Material Tester

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgSeconds, Fewer Is BetterAppleseed 2.0 BetaScene: Material TesterCentOS Stream 880160240320400192.41347.52

TensorFlow Lite

Model: NASNet Mobile

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: NASNet MobileCentOS Stream 830K60K90K120K150KSE +/- 155.05, N = 3SE +/- 1721.19, N = 1274916.4135223.0

Renaissance

Test: In-Memory Database Shootout

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.12Test: In-Memory Database ShootoutCentOS Stream 814002800420056007000SE +/- 38.74, N = 15SE +/- 50.27, N = 103801.36735.1

Blender

Blend File: Pabellon Barcelona - Compute: CPU-Only

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.0Blend File: Pabellon Barcelona - Compute: CPU-OnlyCentOS Stream 820406080100SE +/- 0.13, N = 3SE +/- 0.08, N = 377.5644.71

Embree

Binary: Pathtracer - Model: Crown

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.13Binary: Pathtracer - Model: CrownCentOS Stream 820406080100SE +/- 0.10, N = 3SE +/- 0.41, N = 363.82110.11

Stress-NG

Test: SENDFILE

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: SENDFILECentOS Stream 8400K800K1200K1600K2000KSE +/- 5479.93, N = 3SE +/- 12384.02, N = 31034645.191779916.241. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc

Blender

Blend File: Classroom - Compute: CPU-Only

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.0Blend File: Classroom - Compute: CPU-OnlyCentOS Stream 81428425670SE +/- 0.13, N = 3SE +/- 0.06, N = 361.5336.15

7-Zip Compression

Test: Compression Rating

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 21.06Test: Compression RatingCentOS Stream 8200K400K600K800K1000KSE +/- 2452.32, N = 3SE +/- 8841.92, N = 44779747922641. (CXX) g++ options: -lpthread -ldl -O2 -fPIC

Stockfish

Total Time

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgNodes Per Second, More Is BetterStockfish 13Total TimeCentOS Stream 860M120M180M240M300MSE +/- 1573780.07, N = 3SE +/- 2759095.27, N = 31800990732970635991. (CXX) g++ options: -lgcov -m64 -lpthread -fno-exceptions -std=c++17 -fprofile-use -fno-peel-loops -fno-tracer -pedantic -O3 -msse -msse3 -mpopcnt -mavx2 -msse4.1 -mssse3 -msse2 -flto -flto=jobserver

R Benchmark

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgSeconds, Fewer Is BetterR BenchmarkCentOS Stream 80.08890.17780.26670.35560.4445SE +/- 0.0015, N = 3SE +/- 0.0010, N = 30.39490.24051. R scripting front-end version 4.1.2 (2021-11-01)

Blender

Blend File: BMW27 - Compute: CPU-Only

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.0Blend File: BMW27 - Compute: CPU-OnlyCentOS Stream 8612182430SE +/- 0.07, N = 3SE +/- 0.04, N = 323.2314.48

NAS Parallel Benchmarks

Test / Class: LU.C

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: LU.CCentOS Stream 814K28K42K56K70KSE +/- 84.61, N = 3SE +/- 52.36, N = 339569.9563413.691. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi

OpenSSL

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgsign/s, More Is BetterOpenSSLCentOS Stream 86K12K18K24K30KSE +/- 48.50, N = 3SE +/- 15.51, N = 317870.928503.61. OpenSSL 1.1.1k FIPS 25 Mar 2021

OpenSSL

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgverify/s, More Is BetterOpenSSLCentOS Stream 8400K800K1200K1600K2000KSE +/- 728.21, N = 3SE +/- 1553.72, N = 31182907.41851481.51. OpenSSL 1.1.1k FIPS 25 Mar 2021

Redis Memtier / Redis Benchmark

Test: GET

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgOperations Per Second, More Is BetterRedis Memtier / Redis BenchmarkTest: GETCentOS Stream 8200K400K600K800K1000KSE +/- 1984.40, N = 3SE +/- 1838.84, N = 3782586.56500099.841. (CXX) g++ options: -O2 -levent -levent_openssl -lcrypto -lssl -lpthread -lz -lpcre2. Redis server v=5.0.3 sha=00000000:0 malloc=jemalloc-5.1.0 bits=64 build=28849dbea6f07cc8

Stress-NG

Test: Semaphores

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: SemaphoresCentOS Stream 84M8M12M16M20MSE +/- 10851.57, N = 3SE +/- 116082.15, N = 312394981.2919337664.161. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc

LuxCoreRender

Scene: Orange Juice - Acceleration: CPU

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: Orange Juice - Acceleration: CPUCentOS Stream 8510152025SE +/- 0.11, N = 3SE +/- 0.11, N = 314.3121.94

Embree

Binary: Pathtracer ISPC - Model: Asian Dragon

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.13Binary: Pathtracer ISPC - Model: Asian DragonCentOS Stream 820406080100SE +/- 0.25, N = 3SE +/- 0.80, N = 4108.2371.03

NAS Parallel Benchmarks

Test / Class: BT.C

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: BT.CCentOS Stream 816K32K48K64K80KSE +/- 24.48, N = 3SE +/- 90.74, N = 349340.4174487.141. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi

NAS Parallel Benchmarks

Test / Class: SP.C

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: SP.CCentOS Stream 88K16K24K32K40KSE +/- 14.36, N = 3SE +/- 77.88, N = 324243.7835030.131. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi

Primesieve

1e12 Prime Number Generation

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgSeconds, Fewer Is BetterPrimesieve 7.71e12 Prime Number GenerationCentOS Stream 80.78711.57422.36133.14843.9355SE +/- 0.015, N = 3SE +/- 0.014, N = 33.4982.4491. (CXX) g++ options: -O3 -lpthread

Redis Memtier / Redis Benchmark

Test: LPUSH and LPOP: lpop

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgRequests Per Second, More Is BetterRedis Memtier / Redis BenchmarkTest: LPUSH and LPOP: lpopCentOS Stream 8160K320K480K640K800KSE +/- 1237.45, N = 3SE +/- 3605.89, N = 3527747.44753633.001. (CXX) g++ options: -O2 -levent -levent_openssl -lcrypto -lssl -lpthread -lz -lpcre2. Redis server v=5.0.3 sha=00000000:0 malloc=jemalloc-5.1.0 bits=64 build=28849dbea6f07cc8

Zstd Compression

Compression Level: 19, Long Mode - Compression Speed

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMB/s, More Is BetterZstd CompressionCompression Level: 19, Long Mode - Compression SpeedCentOS Stream 8510152025SE +/- 0.17, N = 7SE +/- 0.22, N = 1218.312.91. *** zstd command line interface 64-bits v1.4.4, by Yann Collet ***

Stress-NG

Test: Glibc Qsort Data Sorting

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: Glibc Qsort Data SortingCentOS Stream 830060090012001500SE +/- 0.88, N = 3SE +/- 0.90, N = 3978.751346.441. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc

Zstd Compression

Compression Level: 8, Long Mode - Compression Speed

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMB/s, More Is BetterZstd CompressionCompression Level: 8, Long Mode - Compression SpeedCentOS Stream 890180270360450SE +/- 0.98, N = 3SE +/- 3.28, N = 15393.8287.61. *** zstd command line interface 64-bits v1.4.4, by Yann Collet ***

Stress-NG

Test: Memory Copying

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: Memory CopyingCentOS Stream 82K4K6K8K10KSE +/- 12.87, N = 3SE +/- 24.77, N = 38486.6311617.461. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc

Apache HTTP Server

Concurrent Requests: 500

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgRequests Per Second, More Is BetterApache HTTP Server 2.4.48Concurrent Requests: 500CentOS Stream 830K60K90K120K150KSE +/- 884.81, N = 3SE +/- 495.48, N = 3146927.60107388.171. (CC) gcc options: -shared -fPIC -O2 -pthread

Embree

Binary: Pathtracer ISPC - Model: Crown

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.13Binary: Pathtracer ISPC - Model: CrownCentOS Stream 820406080100SE +/- 0.26, N = 3SE +/- 0.44, N = 373.53100.42

Mobile Neural Network

Model: inception-v3

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 1.2Model: inception-v3CentOS Stream 8714212835SE +/- 0.07, N = 15SE +/- 0.20, N = 1222.9031.181. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

OSPray

Demo: NASA Streamlines - Renderer: Path Tracer

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgFPS, More Is BetterOSPray 1.8.5Demo: NASA Streamlines - Renderer: Path TracerCentOS Stream 8918273645SE +/- 0.00, N = 3SE +/- 0.00, N = 328.5738.46

NAS Parallel Benchmarks

Test / Class: IS.D

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: IS.DCentOS Stream 86001200180024003000SE +/- 5.09, N = 3SE +/- 4.62, N = 32029.922701.611. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi

rav1e

Speed: 10

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgFrames Per Second, More Is Betterrav1e 0.5Speed: 10CentOS Stream 8246810SE +/- 0.099, N = 3SE +/- 0.076, N = 37.3475.527

Stress-NG

Test: Matrix Math

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: Matrix MathCentOS Stream 890K180K270K360K450KSE +/- 1114.79, N = 3SE +/- 287.18, N = 3333297.17440217.921. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc

Timed LLVM Compilation

Build System: Ninja

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgSeconds, Fewer Is BetterTimed LLVM Compilation 13.0Build System: NinjaCentOS Stream 8306090120150SE +/- 0.28, N = 3SE +/- 0.29, N = 3122.3593.32

Zstd Compression

Compression Level: 19 - Decompression Speed

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMB/s, More Is BetterZstd CompressionCompression Level: 19 - Decompression SpeedCentOS Stream 87001400210028003500SE +/- 15.70, N = 3SE +/- 8.99, N = 32521.53267.91. *** zstd command line interface 64-bits v1.4.4, by Yann Collet ***

LuxCoreRender

Scene: Danish Mood - Acceleration: CPU

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: Danish Mood - Acceleration: CPUCentOS Stream 8246810SE +/- 0.05, N = 15SE +/- 0.13, N = 156.518.41

NAS Parallel Benchmarks

Test / Class: SP.B

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: SP.BCentOS Stream 86K12K18K24K30KSE +/- 11.94, N = 3SE +/- 174.51, N = 323554.8530167.431. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi

Zstd Compression

Compression Level: 8, Long Mode - Decompression Speed

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMB/s, More Is BetterZstd CompressionCompression Level: 8, Long Mode - Decompression SpeedCentOS Stream 88001600240032004000SE +/- 6.95, N = 3SE +/- 3.85, N = 153001.23816.01. *** zstd command line interface 64-bits v1.4.4, by Yann Collet ***

Zstd Compression

Compression Level: 8 - Compression Speed

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMB/s, More Is BetterZstd CompressionCompression Level: 8 - Compression SpeedCentOS Stream 86001200180024003000SE +/- 11.63, N = 3SE +/- 32.84, N = 42169.82754.81. *** zstd command line interface 64-bits v1.4.4, by Yann Collet ***

Cryptsetup

Serpent-XTS 256b Decryption

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMiB/s, More Is BetterCryptsetupSerpent-XTS 256b DecryptionCentOS Stream 8150300450600750SE +/- 0.12, N = 3SE +/- 1.05, N = 3560.8711.2

Cryptsetup

Serpent-XTS 512b Decryption

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMiB/s, More Is BetterCryptsetupSerpent-XTS 512b DecryptionCentOS Stream 8150300450600750SE +/- 0.55, N = 3SE +/- 0.12, N = 3562.2712.1

Timed Node.js Compilation

Time To Compile

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgSeconds, Fewer Is BetterTimed Node.js Compilation 15.11Time To CompileCentOS Stream 820406080100SE +/- 0.30, N = 3SE +/- 0.63, N = 384.8667.68

Appleseed

Scene: Emily

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgSeconds, Fewer Is BetterAppleseed 2.0 BetaScene: EmilyCentOS Stream 8306090120150122.72153.44

OSPray

Demo: San Miguel - Renderer: SciVis

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgFPS, More Is BetterOSPray 1.8.5Demo: San Miguel - Renderer: SciVisCentOS Stream 8306090120150100125

Zstd Compression

Compression Level: 19, Long Mode - Decompression Speed

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMB/s, More Is BetterZstd CompressionCompression Level: 19, Long Mode - Decompression SpeedCentOS Stream 87001400210028003500SE +/- 9.70, N = 7SE +/- 7.85, N = 122674.33340.91. *** zstd command line interface 64-bits v1.4.4, by Yann Collet ***

Sockperf

Test: Latency Ping Pong

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgusec, Fewer Is BetterSockperf 3.7Test: Latency Ping PongCentOS Stream 81.17652.3533.52954.7065.8825SE +/- 0.009, N = 5SE +/- 0.019, N = 55.2294.2361. (CXX) g++ options: --param -O3 -rdynamic -ldl -lpthread

Monte Carlo Simulations of Ionised Nebulae

Input: Dust 2D tau100.0

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgSeconds, Fewer Is BetterMonte Carlo Simulations of Ionised Nebulae 2019-03-24Input: Dust 2D tau100.0CentOS Stream 850100150200250SE +/- 2.65, N = 91902341. (F9X) gfortran options: -cpp -Jsource/ -ffree-line-length-0 -lm -std=legacy -O3 -O2 -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi

Zstd Compression

Compression Level: 8 - Decompression Speed

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMB/s, More Is BetterZstd CompressionCompression Level: 8 - Decompression SpeedCentOS Stream 87001400210028003500SE +/- 10.76, N = 42845.03494.61. *** zstd command line interface 64-bits v1.4.4, by Yann Collet ***

Appleseed

Scene: Disney Material

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgSeconds, Fewer Is BetterAppleseed 2.0 BetaScene: Disney MaterialCentOS Stream 8132639526558.1347.49

Stress-NG

Test: MMAP

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: MMAPCentOS Stream 89001800270036004500SE +/- 3.28, N = 3SE +/- 13.65, N = 33281.974001.021. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc

PlaidML

FP16: No - Mode: Inference - Network: VGG16 - Device: CPU

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgFPS, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: VGG16 - Device: CPUCentOS Stream 81428425670SE +/- 0.27, N = 3SE +/- 0.21, N = 362.6051.51

Cryptsetup

PBKDF2-sha512

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgIterations Per Second, More Is BetterCryptsetupPBKDF2-sha512CentOS Stream 8400K800K1200K1600K2000KSE +/- 3974.38, N = 315352501863599

Stress-NG

Test: NUMA

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: NUMACentOS Stream 83691215SE +/- 0.13, N = 3SE +/- 0.02, N = 310.008.251. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc

Cryptsetup

Serpent-XTS 256b Encryption

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMiB/s, More Is BetterCryptsetupSerpent-XTS 256b EncryptionCentOS Stream 8160320480640800SE +/- 0.33, N = 3SE +/- 1.07, N = 3602.4725.6

Cryptsetup

Serpent-XTS 512b Encryption

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMiB/s, More Is BetterCryptsetupSerpent-XTS 512b EncryptionCentOS Stream 8160320480640800SE +/- 0.85, N = 2SE +/- 2.17, N = 3602.6724.8

PyPerformance

Benchmark: python_startup

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: python_startupCentOS Stream 83691215SE +/- 0.00, N = 3SE +/- 0.00, N = 310.512.6

NAMD

ATPase Simulation - 327,506 Atoms

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgdays/ns, Fewer Is BetterNAMD 2.14ATPase Simulation - 327,506 AtomsCentOS Stream 80.06190.12380.18570.24760.3095SE +/- 0.00054, N = 3SE +/- 0.00044, N = 30.275190.22966

x265

Video Input: Bosphorus 4K

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgFrames Per Second, More Is Betterx265 3.4Video Input: Bosphorus 4KCentOS Stream 8612182430SE +/- 0.40, N = 15SE +/- 0.26, N = 527.3022.971. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl

Apache HTTP Server

Concurrent Requests: 200

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgRequests Per Second, More Is BetterApache HTTP Server 2.4.48Concurrent Requests: 200CentOS Stream 830K60K90K120K150KSE +/- 925.34, N = 3SE +/- 752.69, N = 3142666.94120100.531. (CC) gcc options: -shared -fPIC -O2 -pthread

GNU Octave Benchmark

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgSeconds, Fewer Is BetterGNU Octave Benchmark 5.2.0CentOS Stream 8246810SE +/- 0.006, N = 5SE +/- 0.013, N = 56.6975.641

rav1e

Speed: 6

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgFrames Per Second, More Is Betterrav1e 0.5Speed: 6CentOS Stream 80.62931.25861.88792.51723.1465SE +/- 0.013, N = 3SE +/- 0.028, N = 42.7972.368

rav1e

Speed: 5

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgFrames Per Second, More Is Betterrav1e 0.5Speed: 5CentOS Stream 80.46760.93521.40281.87042.338SE +/- 0.008, N = 3SE +/- 0.013, N = 152.0781.768

OSPray

Demo: NASA Streamlines - Renderer: SciVis

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgFPS, More Is BetterOSPray 1.8.5Demo: NASA Streamlines - Renderer: SciVisCentOS Stream 84080120160200SE +/- 0.00, N = 3SE +/- 0.00, N = 3142.86166.67

Intel Open Image Denoise

Run: RT.ldr_alb_nrm.3840x2160

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 1.4.0Run: RT.ldr_alb_nrm.3840x2160CentOS Stream 80.6661.3321.9982.6643.33SE +/- 0.01, N = 3SE +/- 0.01, N = 32.962.54

Renaissance

Test: Apache Spark PageRank

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.12Test: Apache Spark PageRankCentOS Stream 88001600240032004000SE +/- 28.17, N = 3SE +/- 42.84, N = 43035.33519.2

PlaidML

FP16: No - Mode: Inference - Network: VGG19 - Device: CPU

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgFPS, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: VGG19 - Device: CPUCentOS Stream 81224364860SE +/- 0.28, N = 3SE +/- 0.37, N = 352.2445.08

OSPray

Demo: San Miguel - Renderer: Path Tracer

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgFPS, More Is BetterOSPray 1.8.5Demo: San Miguel - Renderer: Path TracerCentOS Stream 83691215SE +/- 0.00, N = 3SE +/- 0.00, N = 310.5312.20

Embree

Binary: Pathtracer - Model: Asian Dragon

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgFrames Per Second, More Is BetterEmbree 3.13Binary: Pathtracer - Model: Asian DragonCentOS Stream 820406080100SE +/- 0.24, N = 3SE +/- 0.87, N = 483.5872.15

Intel Open Image Denoise

Run: RT.hdr_alb_nrm.3840x2160

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 1.4.0Run: RT.hdr_alb_nrm.3840x2160CentOS Stream 80.65931.31861.97792.63723.2965SE +/- 0.01, N = 3SE +/- 0.01, N = 32.932.53

PHPBench

PHP Benchmark Suite

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgScore, More Is BetterPHPBench 0.8.1PHP Benchmark SuiteCentOS Stream 8140K280K420K560K700KSE +/- 2276.07, N = 3SE +/- 1363.68, N = 3639429552170

Timed Linux Kernel Compilation

Time To Compile

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 5.14Time To CompileCentOS Stream 8510152025SE +/- 0.18, N = 8SE +/- 0.14, N = 1421.2418.74

TensorFlow Lite

Model: Inception V4

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: Inception V4CentOS Stream 8140K280K420K560K700KSE +/- 4325.52, N = 3SE +/- 9303.83, N = 15672585604490

Intel Open Image Denoise

Run: RTLightmap.hdr.4096x4096

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgImages / Sec, More Is BetterIntel Open Image Denoise 1.4.0Run: RTLightmap.hdr.4096x4096CentOS Stream 80.3150.630.9451.261.575SE +/- 0.02, N = 3SE +/- 0.00, N = 31.401.26

High Performance Conjugate Gradient

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgGFLOP/s, More Is BetterHigh Performance Conjugate Gradient 3.1CentOS Stream 8918273645SE +/- 0.02, N = 3SE +/- 0.03, N = 340.6336.741. (CXX) g++ options: -O3 -ffast-math -ftree-vectorize -pthread -lmpi_cxx -lmpi

DaCapo Benchmark

Java Test: Jython

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 9.12-MR1Java Test: JythonCentOS Stream 89001800270036004500SE +/- 15.96, N = 4SE +/- 31.02, N = 436133995

PyPerformance

Benchmark: nbody

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: nbodyCentOS Stream 8306090120150123136

NAS Parallel Benchmarks

Test / Class: CG.C

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: CG.CCentOS Stream 85K10K15K20K25KSE +/- 166.54, N = 15SE +/- 182.81, N = 322885.7525292.481. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi

DaCapo Benchmark

Java Test: Eclipse

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 9.12-MR1Java Test: EclipseCentOS Stream 84K8K12K16K20KSE +/- 134.99, N = 4SE +/- 109.08, N = 41622917900

SVT-HEVC

Tuning: 7 - Input: Bosphorus 1080p

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 7 - Input: Bosphorus 1080pCentOS Stream 870140210280350SE +/- 1.81, N = 3SE +/- 0.88, N = 3299.17329.561. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -lpthread -lrt

Stress-NG

Test: System V Message Passing

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: System V Message PassingCentOS Stream 81.1M2.2M3.3M4.4M5.5MSE +/- 5457.50, N = 3SE +/- 1927.81, N = 34971692.524524150.601. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc

SVT-VP9

Tuning: PSNR/SSIM Optimized - Input: Bosphorus 1080p

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: PSNR/SSIM Optimized - Input: Bosphorus 1080pCentOS Stream 8100200300400500SE +/- 4.13, N = 15SE +/- 4.45, N = 3465.68424.871. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

SVT-AV1

Encoder Mode: Preset 8 - Input: Bosphorus 4K

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 0.8.7Encoder Mode: Preset 8 - Input: Bosphorus 4KCentOS Stream 81428425670SE +/- 0.81, N = 3SE +/- 0.38, N = 358.5263.951. (CXX) g++ options: -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq -pie

Xcompact3d Incompact3d

Input: X3D-benchmarking input.i3d

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgSeconds, Fewer Is BetterXcompact3d Incompact3d 2021-03-11Input: X3D-benchmarking input.i3dCentOS Stream 880160240320400SE +/- 0.11, N = 3SE +/- 0.09, N = 3319.57345.581. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi

Timed GCC Compilation

Time To Compile

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgSeconds, Fewer Is BetterTimed GCC Compilation 11.2.0Time To CompileCentOS Stream 8150300450600750SE +/- 0.56, N = 3SE +/- 1.40, N = 3654.21706.43

PyPerformance

Benchmark: chaos

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: chaosCentOS Stream 8306090120150127137

Timed LLVM Compilation

Build System: Unix Makefiles

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgSeconds, Fewer Is BetterTimed LLVM Compilation 13.0Build System: Unix MakefilesCentOS Stream 84080120160200SE +/- 1.48, N = 3SE +/- 0.51, N = 3180.01167.80

GROMACS

Implementation: MPI CPU - Input: water_GMX50_bare

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgNs Per Day, More Is BetterGROMACS 2021.2Implementation: MPI CPU - Input: water_GMX50_bareCentOS Stream 83691215SE +/- 0.020, N = 3SE +/- 0.033, N = 38.6549.2821. (CXX) g++ options: -O3 -pthread

Cryptsetup

PBKDF2-whirlpool

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgIterations Per Second, More Is BetterCryptsetupPBKDF2-whirlpoolCentOS Stream 8140K280K420K560K700KSE +/- 1213.62, N = 3602629644885

Apache HTTP Server

Concurrent Requests: 1000

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgRequests Per Second, More Is BetterApache HTTP Server 2.4.48Concurrent Requests: 1000CentOS Stream 820K40K60K80K100KSE +/- 382.73, N = 3SE +/- 710.10, N = 3113997.00106659.271. (CC) gcc options: -shared -fPIC -O2 -pthread

PyPerformance

Benchmark: 2to3

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: 2to3CentOS Stream 880160240320400SE +/- 0.58, N = 3355375

SVT-AV1

Encoder Mode: Preset 4 - Input: Bosphorus 4K

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 0.8.7Encoder Mode: Preset 4 - Input: Bosphorus 4KCentOS Stream 81.13992.27983.41974.55965.6995SE +/- 0.032, N = 3SE +/- 0.012, N = 34.8085.0661. (CXX) g++ options: -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq -pie

Sockperf

Test: Throughput

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMessages Per Second, More Is BetterSockperf 3.7Test: ThroughputCentOS Stream 870K140K210K280K350KSE +/- 4047.87, N = 23SE +/- 2067.12, N = 253277633450241. (CXX) g++ options: --param -O3 -rdynamic -ldl -lpthread

PyPerformance

Benchmark: float

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: floatCentOS Stream 8306090120150SE +/- 0.33, N = 3120126

GPAW

Input: Carbon Nanotube

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgSeconds, Fewer Is BetterGPAW 20.1Input: Carbon NanotubeCentOS Stream 880160240320400SE +/- 0.67, N = 3SE +/- 0.69, N = 3359.41344.531. (CC) gcc options: -pthread -shared -lxc -lblas -lpython3.6m -lmpi -O2 -pipe -fexceptions -m64 -mtune=generic -fcf-protection -fPIC -fwrapv -UNDEBUG -std=c99

Zstd Compression

Compression Level: 19 - Compression Speed

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMB/s, More Is BetterZstd CompressionCompression Level: 19 - Compression SpeedCentOS Stream 81632486480SE +/- 0.77, N = 3SE +/- 0.74, N = 367.169.81. *** zstd command line interface 64-bits v1.4.4, by Yann Collet ***

Cryptsetup

Twofish-XTS 256b Decryption

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMiB/s, More Is BetterCryptsetupTwofish-XTS 256b DecryptionCentOS Stream 880160240320400SE +/- 0.23, N = 3SE +/- 0.43, N = 3363.0377.4

Cryptsetup

Twofish-XTS 512b Decryption

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMiB/s, More Is BetterCryptsetupTwofish-XTS 512b DecryptionCentOS Stream 880160240320400SE +/- 0.22, N = 3SE +/- 0.59, N = 3363.3376.5

SVT-HEVC

Tuning: 10 - Input: Bosphorus 1080p

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 1.5.0Tuning: 10 - Input: Bosphorus 1080pCentOS Stream 8130260390520650SE +/- 2.87, N = 3SE +/- 2.53, N = 3574.56591.351. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -lpthread -lrt

SVT-VP9

Tuning: Visual Quality Optimized - Input: Bosphorus 1080p

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: Visual Quality Optimized - Input: Bosphorus 1080pCentOS Stream 880160240320400SE +/- 4.61, N = 3SE +/- 2.18, N = 3359.89350.671. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

PyPerformance

Benchmark: json_loads

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: json_loadsCentOS Stream 8612182430SE +/- 0.03, N = 3SE +/- 0.09, N = 326.927.6

PlaidML

FP16: No - Mode: Inference - Network: ResNet 50 - Device: CPU

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgFPS, More Is BetterPlaidMLFP16: No - Mode: Inference - Network: ResNet 50 - Device: CPUCentOS Stream 8246810SE +/- 0.02, N = 3SE +/- 0.04, N = 36.015.86

PyPerformance

Benchmark: pickle_pure_python

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: pickle_pure_pythonCentOS Stream 8120240360480600SE +/- 0.67, N = 3553567

PyPerformance

Benchmark: pathlib

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: pathlibCentOS Stream 8510152025SE +/- 0.00, N = 3SE +/- 0.03, N = 320.220.7

PyPerformance

Benchmark: regex_compile

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: regex_compileCentOS Stream 850100150200250SE +/- 0.33, N = 3208213

Cryptsetup

Twofish-XTS 512b Encryption

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMiB/s, More Is BetterCryptsetupTwofish-XTS 512b EncryptionCentOS Stream 880160240320400SE +/- 0.22, N = 3SE +/- 0.15, N = 3368.1376.8

ACES DGEMM

Sustained Floating-Point Rate

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgGFLOP/s, More Is BetterACES DGEMM 1.0Sustained Floating-Point RateCentOS Stream 81020304050SE +/- 0.49, N = 15SE +/- 0.55, N = 444.4145.451. (CC) gcc options: -O3 -march=native -fopenmp

Cryptsetup

Twofish-XTS 256b Encryption

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMiB/s, More Is BetterCryptsetupTwofish-XTS 256b EncryptionCentOS Stream 880160240320400SE +/- 0.23, N = 3SE +/- 1.87, N = 3367.9376.0

PyPerformance

Benchmark: django_template

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: django_templateCentOS Stream 820406080100SE +/- 0.12, N = 3SE +/- 0.12, N = 378.577.1

SVT-VP9

Tuning: VMAF Optimized - Input: Bosphorus 1080p

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 0.3Tuning: VMAF Optimized - Input: Bosphorus 1080pCentOS Stream 8100200300400500SE +/- 5.38, N = 4SE +/- 5.40, N = 4476.33467.871. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm

PyPerformance

Benchmark: crypto_pyaes

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: crypto_pyaesCentOS Stream 8306090120150114112

PyBench

Total For Average Test Times

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMilliseconds, Fewer Is BetterPyBench 2018-02-16Total For Average Test TimesCentOS Stream 830060090012001500SE +/- 2.19, N = 3SE +/- 1.86, N = 311671150

PyPerformance

Benchmark: raytrace

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: raytraceCentOS Stream 8130260390520650SE +/- 1.76, N = 3SE +/- 0.67, N = 3621613

Xcompact3d Incompact3d

Input: input.i3d 193 Cells Per Direction

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgSeconds, Fewer Is BetterXcompact3d Incompact3d 2021-03-11Input: input.i3d 193 Cells Per DirectionCentOS Stream 83691215SE +/- 0.08, N = 3SE +/- 0.16, N = 312.0712.191. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi

PyPerformance

Benchmark: go

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMilliseconds, Fewer Is BetterPyPerformance 1.0.0Benchmark: goCentOS Stream 870140210280350SE +/- 0.58, N = 3301303

LAMMPS Molecular Dynamics Simulator

Model: 20k Atoms

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgns/day, More Is BetterLAMMPS Molecular Dynamics Simulator 29Oct2020Model: 20k AtomsCentOS Stream 8816243240SE +/- 0.02, N = 3SE +/- 0.14, N = 335.3635.491. (CXX) g++ options: -O3 -pthread -lm

OpenCV

Test: DNN - Deep Neural Network

OpenBenchmarking.orgms, Fewer Is BetterOpenCV 4.5.4Test: DNN - Deep Neural NetworkXeon Platinum 8380 2P: CentOS Stream 82K4K6K8K10KSE +/- 122.38, N = 391011. (CXX) g++ options: -fPIC -fsigned-char -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -O3 -shared

Stress-NG

Test: RdRand

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: RdRandCentOS Stream 8140K280K420K560K700KSE +/- 53.75, N = 3669474.481. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc

OSPray

Demo: Magnetic Reconnection - Renderer: Path Tracer

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgFPS, More Is BetterOSPray 1.8.5Demo: Magnetic Reconnection - Renderer: Path TracerCentOS Stream 8110220330440550500500

OpenCV

Test: Object Detection

OpenBenchmarking.orgms, Fewer Is BetterOpenCV 4.5.4Test: Object DetectionXeon Platinum 8380 2P: CentOS Stream 816K32K48K64K80KSE +/- 1456.11, N = 12729931. (CXX) g++ options: -fPIC -fsigned-char -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -O3 -shared

OpenCV

Test: Features 2D

OpenBenchmarking.orgms, Fewer Is BetterOpenCV 4.5.4Test: Features 2DXeon Platinum 8380 2P: CentOS Stream 830K60K90K120K150KSE +/- 4393.54, N = 121499131. (CXX) g++ options: -fPIC -fsigned-char -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -O3 -shared

ONNX Runtime

Model: super-resolution-10 - Device: CPU

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.10Model: super-resolution-10 - Device: CPUCentOS Stream 82K4K6K8K10KSE +/- 8.65, N = 3SE +/- 295.99, N = 91080441101. (CXX) g++ options: -ffunction-sections -fdata-sections -march=native -mtune=native -O3 -flto -fno-fat-lto-objects -ldl -lrt -pthread -lpthread

ONNX Runtime

Model: shufflenet-v2-10 - Device: CPU

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.10Model: shufflenet-v2-10 - Device: CPUCentOS Stream 82K4K6K8K10KSE +/- 172.85, N = 12SE +/- 303.89, N = 91032844851. (CXX) g++ options: -ffunction-sections -fdata-sections -march=native -mtune=native -O3 -flto -fno-fat-lto-objects -ldl -lrt -pthread -lpthread

ONNX Runtime

Model: fcn-resnet101-11 - Device: CPU

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.10Model: fcn-resnet101-11 - Device: CPUCentOS Stream 8100200300400500SE +/- 15.75, N = 12SE +/- 2.15, N = 124741931. (CXX) g++ options: -ffunction-sections -fdata-sections -march=native -mtune=native -O3 -flto -fno-fat-lto-objects -ldl -lrt -pthread -lpthread

ONNX Runtime

Model: yolov4 - Device: CPU

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgInferences Per Minute, More Is BetterONNX Runtime 1.10Model: yolov4 - Device: CPUCentOS Stream 8140280420560700SE +/- 7.12, N = 3SE +/- 6.47, N = 126713271. (CXX) g++ options: -ffunction-sections -fdata-sections -march=native -mtune=native -O3 -flto -fno-fat-lto-objects -ldl -lrt -pthread -lpthread

Blender

Blend File: Barbershop - Compute: CPU-Only

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.0Blend File: Barbershop - Compute: CPU-OnlyCentOS Stream 850100150200250SE +/- 0.29, N = 3SE +/- 25.57, N = 9246.83180.32

NCNN

Target: CPU - Model: regnety_400m

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: regnety_400mCentOS Stream 8306090120150SE +/- 0.49, N = 4SE +/- 2.41, N = 938.69116.141. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread

NCNN

Target: CPU - Model: squeezenet_ssd

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: squeezenet_ssdCentOS Stream 81632486480SE +/- 0.25, N = 4SE +/- 7.24, N = 924.4173.121. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread

NCNN

Target: CPU - Model: yolov4-tiny

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: yolov4-tinyCentOS Stream 81122334455SE +/- 0.17, N = 4SE +/- 3.11, N = 926.6150.271. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread

NCNN

Target: CPU - Model: resnet50

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: resnet50CentOS Stream 820406080100SE +/- 0.43, N = 4SE +/- 14.06, N = 927.0998.431. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread

NCNN

Target: CPU - Model: alexnet

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: alexnetCentOS Stream 848121620SE +/- 0.39, N = 4SE +/- 0.80, N = 910.1115.471. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread

NCNN

Target: CPU - Model: resnet18

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: resnet18CentOS Stream 8816243240SE +/- 0.47, N = 4SE +/- 1.55, N = 914.8336.411. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread

NCNN

Target: CPU - Model: vgg16

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: vgg16CentOS Stream 8918273645SE +/- 0.71, N = 4SE +/- 1.42, N = 931.7341.441. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread

NCNN

Target: CPU - Model: googlenet

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: googlenetCentOS Stream 81530456075SE +/- 0.82, N = 4SE +/- 3.47, N = 923.1965.911. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread

NCNN

Target: CPU - Model: blazeface

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: blazefaceCentOS Stream 8510152025SE +/- 0.11, N = 4SE +/- 0.54, N = 96.8819.901. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread

NCNN

Target: CPU - Model: efficientnet-b0

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: efficientnet-b0CentOS Stream 81326395265SE +/- 0.27, N = 4SE +/- 1.83, N = 914.5058.651. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread

NCNN

Target: CPU - Model: mnasnet

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: mnasnetCentOS Stream 81122334455SE +/- 0.12, N = 4SE +/- 0.99, N = 911.1146.941. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread

NCNN

Target: CPU - Model: shufflenet-v2

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: shufflenet-v2CentOS Stream 8816243240SE +/- 0.30, N = 4SE +/- 0.72, N = 911.9635.161. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread

NCNN

Target: CPU-v3-v3 - Model: mobilenet-v3

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU-v3-v3 - Model: mobilenet-v3CentOS Stream 81122334455SE +/- 0.07, N = 4SE +/- 3.07, N = 911.7248.661. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread

NCNN

Target: CPU-v2-v2 - Model: mobilenet-v2

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU-v2-v2 - Model: mobilenet-v2CentOS Stream 81122334455SE +/- 0.14, N = 4SE +/- 2.60, N = 911.8250.931. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread

NCNN

Target: CPU - Model: mobilenet

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterNCNN 20210720Target: CPU - Model: mobilenetCentOS Stream 820406080100SE +/- 0.23, N = 4SE +/- 5.90, N = 920.7392.941. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread

Mobile Neural Network

Model: mobilenet-v1-1.0

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 1.2Model: mobilenet-v1-1.0CentOS Stream 81.00352.0073.01054.0145.0175SE +/- 0.008, N = 15SE +/- 0.132, N = 121.9254.4601. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

Mobile Neural Network

Model: MobileNetV2_224

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 1.2Model: MobileNetV2_224CentOS Stream 81.1522.3043.4564.6085.76SE +/- 0.035, N = 15SE +/- 0.151, N = 122.6695.1201. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

Mobile Neural Network

Model: SqueezeNetV1.0

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 1.2Model: SqueezeNetV1.0CentOS Stream 8246810SE +/- 0.141, N = 15SE +/- 0.119, N = 124.6198.0891. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

Mobile Neural Network

Model: resnet-v2-50

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 1.2Model: resnet-v2-50CentOS Stream 8816243240SE +/- 0.05, N = 15SE +/- 4.72, N = 1219.4335.931. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

Mobile Neural Network

Model: squeezenetv1.1

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 1.2Model: squeezenetv1.1CentOS Stream 83691215SE +/- 0.052, N = 15SE +/- 1.075, N = 123.8289.5061. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

Mobile Neural Network

Model: mobilenetV3

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterMobile Neural Network 1.2Model: mobilenetV3CentOS Stream 8246810SE +/- 0.025, N = 15SE +/- 0.985, N = 122.1486.8761. (CXX) g++ options: -std=c++11 -O3 -fvisibility=hidden -fomit-frame-pointer -fstrict-aliasing -ffunction-sections -fdata-sections -ffast-math -fno-rtti -fno-exceptions -rdynamic -pthread -ldl

Redis Memtier / Redis Benchmark

Test: SET

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgOperations Per Second, More Is BetterRedis Memtier / Redis BenchmarkTest: SETCentOS Stream 870K140K210K280K350KSE +/- 10740.70, N = 9SE +/- 7402.50, N = 12309851.24209135.791. (CXX) g++ options: -O2 -levent -levent_openssl -lcrypto -lssl -lpthread -lz -lpcre2. Redis server v=5.0.3 sha=00000000:0 malloc=jemalloc-5.1.0 bits=64 build=28849dbea6f07cc8

Redis Memtier / Redis Benchmark

Test: MIX

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgOperations Per Second, More Is BetterRedis Memtier / Redis BenchmarkTest: MIXCentOS Stream 8100K200K300K400K500KSE +/- 8588.53, N = 12SE +/- 8460.70, N = 12464386.54334361.811. (CXX) g++ options: -O2 -levent -levent_openssl -lcrypto -lssl -lpthread -lz -lpcre2. Redis server v=5.0.3 sha=00000000:0 malloc=jemalloc-5.1.0 bits=64 build=28849dbea6f07cc8

Redis Memtier / Redis Benchmark

Test: LPUSH and LPOP: lpush

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgRequests Per Second, More Is BetterRedis Memtier / Redis BenchmarkTest: LPUSH and LPOP: lpushCentOS Stream 8200K400K600K800K1000KSE +/- 25162.13, N = 15SE +/- 15174.05, N = 151126290.26803762.141. (CXX) g++ options: -O2 -levent -levent_openssl -lcrypto -lssl -lpthread -lz -lpcre2. Redis server v=5.0.3 sha=00000000:0 malloc=jemalloc-5.1.0 bits=64 build=28849dbea6f07cc8

Stress-NG

Test: Context Switching

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: Context SwitchingCentOS Stream 815M30M45M60M75MSE +/- 809501.98, N = 15SE +/- 834986.74, N = 411298849.8470077496.401. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc

Stress-NG

Test: Socket Activity

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: Socket ActivityCentOS Stream 87001400210028003500SE +/- 953.56, N = 15SE +/- 887.18, N = 153240.043235.891. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc

Stress-NG

Test: CPU Cache

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.13.02Test: CPU CacheCentOS Stream 8918273645SE +/- 0.42, N = 15SE +/- 2.09, N = 1517.6339.741. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc

Redis

Test: SET

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgRequests Per Second, More Is BetterRedis 6.0.9Test: SETCentOS Stream 8400K800K1200K1600K2000KSE +/- 16442.53, N = 3SE +/- 68212.04, N = 121829369.291463413.611. (CXX) g++ options: -MM -MT -g3 -fvisibility=hidden -O3

Redis

Test: GET

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgRequests Per Second, More Is BetterRedis 6.0.9Test: GETCentOS Stream 8500K1000K1500K2000K2500KSE +/- 33069.54, N = 15SE +/- 57548.10, N = 152357157.421763052.441. (CXX) g++ options: -MM -MT -g3 -fvisibility=hidden -O3

PostgreSQL pgbench

Scaling Factor: 1000 - Clients: 500 - Mode: Read Write - Average Latency

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterPostgreSQL pgbench 14.0Scaling Factor: 1000 - Clients: 500 - Mode: Read Write - Average LatencyCentOS Stream 8400800120016002000SE +/- 363.48, N = 12SE +/- 510.29, N = 91709.892048.091. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lpthread -lrt -ldl -lm

PostgreSQL pgbench

Scaling Factor: 1000 - Clients: 500 - Mode: Read Write

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 14.0Scaling Factor: 1000 - Clients: 500 - Mode: Read WriteCentOS Stream 85K10K15K20K25KSE +/- 10758.95, N = 12SE +/- 11591.84, N = 925426233341. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lpthread -lrt -ldl -lm

PostgreSQL pgbench

Scaling Factor: 1000 - Clients: 250 - Mode: Read Write - Average Latency

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterPostgreSQL pgbench 14.0Scaling Factor: 1000 - Clients: 250 - Mode: Read Write - Average LatencyCentOS Stream 82004006008001000SE +/- 140.61, N = 12SE +/- 183.30, N = 12810.241052.551. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lpthread -lrt -ldl -lm

PostgreSQL pgbench

Scaling Factor: 1000 - Clients: 250 - Mode: Read Write

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 14.0Scaling Factor: 1000 - Clients: 250 - Mode: Read WriteCentOS Stream 85K10K15K20K25KSE +/- 11129.26, N = 12SE +/- 10026.46, N = 1221541193781. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lpthread -lrt -ldl -lm

PostgreSQL pgbench

Scaling Factor: 1000 - Clients: 500 - Mode: Read Only - Average Latency

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterPostgreSQL pgbench 14.0Scaling Factor: 1000 - Clients: 500 - Mode: Read Only - Average LatencyCentOS Stream 84080120160200SE +/- 22.86, N = 9SE +/- 25.01, N = 12182.86184.341. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lpthread -lrt -ldl -lm

PostgreSQL pgbench

Scaling Factor: 1000 - Clients: 500 - Mode: Read Only

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 14.0Scaling Factor: 1000 - Clients: 500 - Mode: Read OnlyCentOS Stream 830K60K90K120K150KSE +/- 93232.04, N = 9SE +/- 97067.17, N = 12956651462231. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lpthread -lrt -ldl -lm

PostgreSQL pgbench

Scaling Factor: 1000 - Clients: 250 - Mode: Read Only - Average Latency

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetterPostgreSQL pgbench 14.0Scaling Factor: 1000 - Clients: 250 - Mode: Read Only - Average LatencyCentOS Stream 8306090120150SE +/- 12.04, N = 9SE +/- 15.76, N = 948.14121.671. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lpthread -lrt -ldl -lm

PostgreSQL pgbench

Scaling Factor: 1000 - Clients: 250 - Mode: Read Only

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 14.0Scaling Factor: 1000 - Clients: 250 - Mode: Read OnlyCentOS Stream 880K160K240K320K400KSE +/- 191388.82, N = 9SE +/- 72993.53, N = 9382272748381. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lpthread -lrt -ldl -lm

TensorFlow Lite

Model: Inception ResNet V2

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: Inception ResNet V2CentOS Stream 8140K280K420K560K700KSE +/- 2874.78, N = 3SE +/- 21565.13, N = 12580890655650

TensorFlow Lite

Model: Mobilenet Quant

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: Mobilenet QuantCentOS Stream 811K22K33K44K55KSE +/- 84.87, N = 3SE +/- 1002.46, N = 1534025.350910.2

TensorFlow Lite

Model: Mobilenet Float

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: Mobilenet FloatCentOS Stream 811K22K33K44K55KSE +/- 52.61, N = 3SE +/- 1146.99, N = 1532947.851660.2

TensorFlow Lite

Model: SqueezeNet

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMicroseconds, Fewer Is BetterTensorFlow Lite 2020-08-23Model: SqueezeNetCentOS Stream 813K26K39K52K65KSE +/- 105.74, N = 3SE +/- 1414.69, N = 1547883.059188.8

ASKAP

Test: Hogbom Clean OpenMP

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgIterations Per Second, More Is BetterASKAP 1.0Test: Hogbom Clean OpenMPCentOS Stream 82004006008001000SE +/- 39.14, N = 15SE +/- 9.50, N = 151086.45257.251. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp

ASKAP

Test: tConvolve OpenMP - Degridding

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMillion Grid Points Per Second, More Is BetterASKAP 1.0Test: tConvolve OpenMP - DegriddingCentOS Stream 84K8K12K16K20KSE +/- 422.63, N = 3SE +/- 163.43, N = 1518173.006070.391. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp

ASKAP

Test: tConvolve OpenMP - Gridding

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMillion Grid Points Per Second, More Is BetterASKAP 1.0Test: tConvolve OpenMP - GriddingCentOS Stream 82K4K6K8K10KSE +/- 160.77, N = 3SE +/- 266.31, N = 1511254.805689.701. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp

ASKAP

Test: tConvolve MT - Degridding

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMillion Grid Points Per Second, More Is BetterASKAP 1.0Test: tConvolve MT - DegriddingCentOS Stream 83K6K9K12K15KSE +/- 373.79, N = 9SE +/- 126.65, N = 312604.413224.01. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp

ASKAP

Test: tConvolve MT - Gridding

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgMillion Grid Points Per Second, More Is BetterASKAP 1.0Test: tConvolve MT - GriddingCentOS Stream 82K4K6K8K10KSE +/- 172.71, N = 9SE +/- 41.20, N = 38372.038717.811. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp

oneDNN

Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPU

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPUCentOS Stream 830060090012001500SE +/- 1.93, N = 3SE +/- 41.11, N = 12442.211338.151. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

oneDNN

Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPU

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPUCentOS Stream 80.35530.71061.06591.42121.7765SE +/- 0.01980, N = 12SE +/- 0.05247, N = 151.196801.578921. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

oneDNN

Harness: IP Shapes 1D - Data Type: u8s8f32 - Engine: CPU

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgms, Fewer Is BetteroneDNN 2.1.2Harness: IP Shapes 1D - Data Type: u8s8f32 - Engine: CPUCentOS Stream 80.34970.69941.04911.39881.7485SE +/- 0.06642, N = 12SE +/- 0.00429, N = 31.501471.554231. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl

LuxCoreRender

Scene: Rainbow Colors and Prism - Acceleration: CPU

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: Rainbow Colors and Prism - Acceleration: CPUCentOS Stream 848121620SE +/- 0.85, N = 12SE +/- 0.57, N = 1212.8917.17

LuxCoreRender

Scene: LuxCore Benchmark - Acceleration: CPU

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: LuxCore Benchmark - Acceleration: CPUCentOS Stream 8246810SE +/- 0.13, N = 15SE +/- 0.02, N = 36.718.66

LuxCoreRender

Scene: DLSC - Acceleration: CPU

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgM samples/sec, More Is BetterLuxCoreRender 2.5Scene: DLSC - Acceleration: CPUCentOS Stream 848121620SE +/- 0.06, N = 15SE +/- 0.31, N = 159.0514.31

DaCapo Benchmark

Java Test: Tradebeans

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 9.12-MR1Java Test: TradebeansCentOS Stream 812002400360048006000SE +/- 70.75, N = 20SE +/- 191.21, N = 1633675456

DaCapo Benchmark

Java Test: H2

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 9.12-MR1Java Test: H2CentOS Stream 811002200330044005500SE +/- 49.75, N = 20SE +/- 76.12, N = 2038435264

LAMMPS Molecular Dynamics Simulator

Model: Rhodopsin Protein

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgns/day, More Is BetterLAMMPS Molecular Dynamics Simulator 29Oct2020Model: Rhodopsin ProteinCentOS Stream 8816243240SE +/- 0.75, N = 14SE +/- 0.31, N = 1533.4229.481. (CXX) g++ options: -O3 -pthread -lm

Xcompact3d Incompact3d

Input: input.i3d 129 Cells Per Direction

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgSeconds, Fewer Is BetterXcompact3d Incompact3d 2021-03-11Input: input.i3d 129 Cells Per DirectionCentOS Stream 80.77621.55242.32863.10483.881SE +/- 0.04066101, N = 15SE +/- 0.05694678, N = 152.973570743.449857851. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi

NAS Parallel Benchmarks

Test / Class: MG.C

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: MG.CCentOS Stream 810K20K30K40K50KSE +/- 39.43, N = 3SE +/- 883.57, N = 1244883.5148659.621. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi

NAS Parallel Benchmarks

Test / Class: FT.C

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: FT.CCentOS Stream 88K16K24K32K40KSE +/- 25.02, N = 3SE +/- 1314.27, N = 1238431.5339242.311. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi

NAS Parallel Benchmarks

Test / Class: EP.D

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: EP.DCentOS Stream 87001400210028003500SE +/- 23.33, N = 3SE +/- 84.69, N = 122597.653175.991. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi

NAS Parallel Benchmarks

Test / Class: EP.C

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.4Test / Class: EP.CCentOS Stream 87001400210028003500SE +/- 2.20, N = 3SE +/- 78.53, N = 152611.813244.221. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi

Sockperf

Test: Latency Under Load

Xeon Platinum 8380 2PEPYC 7763 2POpenBenchmarking.orgusec, Fewer Is BetterSockperf 3.7Test: Latency Under LoadCentOS Stream 848121620SE +/- 1.82, N = 25SE +/- 1.02, N = 2517.3716.891. (CXX) g++ options: --param -O3 -rdynamic -ldl -lpthread

Phoronix Test Suite v10.8.4