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.
Compare your own system(s) to this result file with the
Phoronix Test Suite by running the command:
phoronix-test-suite benchmark 2112063-TJ-CENTOSSTR24 CentOS Stream 8 - Phoronix Test Suite 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&sro&grr .
Processor Motherboard Chipset Memory Disk Graphics Monitor Network OS Kernel Desktop Display Server OpenGL Compiler File-System Screen Resolution Xeon Platinum 8380 2P EPYC 7763 2P CentOS Stream 8 CentOS Stream 8 2 x Intel Xeon Platinum 8380 @ 3.40GHz (80 Cores / 160 Threads) Intel M50CYP2SB2U (SE5C6200.86B.0022.D08.2103221623 BIOS) Intel Device 0998 504GB 7682GB INTEL SSDPF2KX076TZ llvmpipe VE228 2 x Intel X710 for 10GBASE-T + 2 x Intel E810-C for QSFP CentOS Stream 8 4.18.0-348.2.1.el8_5.x86_64 (x86_64) GNOME Shell 3.32.2 X Server 4.5 Mesa 21.3.0 (LLVM 12.0.1 256 bits) GCC 8.5.0 20210514 xfs 1920x1080 2 x AMD EPYC 7763 64-Core @ 2.45GHz (128 Cores / 256 Threads) AMD DAYTONA_X (RYM1001D BIOS) AMD Starship/Matisse 2 x Mellanox MT27710 OpenBenchmarking.org Kernel Details - Transparent Huge Pages: always Compiler 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: 0xa001119 Java Details - OpenJDK Runtime Environment (build 1.8.0_312-b07) Python Details - Python 3.6.8 Security 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
dacapobench: Tradebeans pgbench: 1000 - 250 - Read Write - Average Latency pgbench: 1000 - 250 - Read Write renaissance: Akka Unbalanced Cobwebbed Tree pgbench: 1000 - 500 - Read Only - Average Latency pgbench: 1000 - 500 - Read Only pgbench: 1000 - 500 - Read Write - Average Latency pgbench: 1000 - 500 - Read Write build-gcc: Time To Compile pgbench: 1000 - 250 - Read Only - Average Latency pgbench: 1000 - 250 - Read Only opencv: Features 2D ncnn: CPU - regnety_400m ncnn: CPU - squeezenet_ssd ncnn: CPU - yolov4-tiny ncnn: CPU - resnet50 ncnn: CPU - alexnet ncnn: CPU - resnet18 ncnn: CPU - vgg16 ncnn: CPU - googlenet ncnn: CPU - blazeface ncnn: CPU - efficientnet-b0 ncnn: CPU - mnasnet ncnn: CPU - shufflenet-v2 ncnn: CPU-v3-v3 - mobilenet-v3 ncnn: CPU-v2-v2 - mobilenet-v2 ncnn: CPU - mobilenet askap: tConvolve MT - Degridding askap: tConvolve MT - Gridding redis: MIX mnn: inception-v3 mnn: mobilenet-v1-1.0 mnn: MobileNetV2_224 mnn: SqueezeNetV1.0 mnn: resnet-v2-50 mnn: squeezenetv1.1 mnn: mobilenetV3 onnx: fcn-resnet101-11 - CPU redis: SET mocassin: Dust 2D tau100.0 onnx: shufflenet-v2-10 - CPU blender: Barbershop - CPU-Only renaissance: In-Memory Database Shootout plaidml: No - Inference - ResNet 50 - CPU gpaw: Carbon Nanotube redis: LPUSH and LPOP: lpush incompact3d: X3D-benchmarking input.i3d luxcorerender: Danish Mood - CPU luxcorerender: DLSC - CPU onnx: yolov4 - CPU compress-zstd: 19, Long Mode - Decompression Speed compress-zstd: 19, Long Mode - Compression Speed opencv: Object Detection lammps: 20k Atoms onnx: super-resolution-10 - CPU hpcg: onednn: Recurrent Neural Network Inference - u8s8f32 - CPU luxcorerender: LuxCore Benchmark - CPU tensorflow-lite: Inception V4 tensorflow-lite: Mobilenet Quant tensorflow-lite: Mobilenet Float tensorflow-lite: SqueezeNet appleseed: Material Tester build-llvm: Unix Makefiles tensorflow-lite: Inception ResNet V2 tensorflow-lite: NASNet Mobile stress-ng: CPU Cache stress-ng: Socket Activity renaissance: Apache Spark PageRank npb: EP.D redis: GET compress-zstd: 8, Long Mode - Decompression Speed compress-zstd: 8, Long Mode - Compression Speed build-llvm: Ninja askap: Hogbom Clean OpenMP pyperformance: raytrace rav1e: 5 stress-ng: Context Switching appleseed: Emily apache: 1000 onednn: Recurrent Neural Network Training - u8s8f32 - CPU pyperformance: python_startup nginx: 1000 apache: 500 nginx: 500 nginx: 200 apache: 200 x265: Bosphorus 4K build-nodejs: Time To Compile redis: GET build-linux-kernel: Time To Compile redis: LPUSH and LPOP: lpop sockperf: Latency Under Load pyperformance: 2to3 sockperf: Throughput dacapobench: H2 luxcorerender: Orange Juice - CPU onednn: Deconvolution Batch shapes_1d - u8s8f32 - CPU blender: Pabellon Barcelona - CPU-Only pyperformance: go npb: SP.C npb: BT.C incompact3d: input.i3d 129 Cells Per Direction npb: FT.C plaidml: No - Inference - VGG19 - CPU luxcorerender: Rainbow Colors and Prism - CPU blender: Classroom - CPU-Only npb: LU.C askap: tConvolve MPI - Gridding askap: tConvolve MPI - Degridding plaidml: No - Inference - VGG16 - CPU compress-7zip: Decompression Rating compress-7zip: Compression Rating ospray: San Miguel - Path Tracer compress-zstd: 8 - Decompression Speed compress-zstd: 8 - Compression Speed pyperformance: regex_compile redis: SET gromacs: MPI CPU - water_GMX50_bare lammps: Rhodopsin Protein npb: EP.C compress-zstd: 19 - Decompression Speed compress-zstd: 19 - Compression Speed onednn: IP Shapes 1D - u8s8f32 - CPU svt-av1: Preset 4 - Bosphorus 4K cryptsetup: Serpent-XTS 512b Encryption cryptsetup: Twofish-XTS 512b Decryption cryptsetup: Twofish-XTS 512b Encryption cryptsetup: Serpent-XTS 512b Decryption cryptsetup: AES-XTS 512b Decryption cryptsetup: AES-XTS 512b Encryption cryptsetup: Twofish-XTS 256b Decryption cryptsetup: Twofish-XTS 256b Encryption cryptsetup: Serpent-XTS 256b Decryption cryptsetup: Serpent-XTS 256b Encryption cryptsetup: AES-XTS 256b Decryption cryptsetup: AES-XTS 256b Encryption cryptsetup: PBKDF2-whirlpool cryptsetup: PBKDF2-sha512 pyperformance: pathlib pyperformance: django_template rbenchmark: appleseed: Disney Material stress-ng: Forking npb: MG.C pyperformance: pickle_pure_python phpbench: PHP Benchmark Suite stress-ng: NUMA askap: tConvolve OpenMP - Degridding askap: tConvolve OpenMP - Gridding namd: ATPase Simulation - 327,506 Atoms npb: CG.C stress-ng: Memory Copying stress-ng: Malloc stress-ng: MMAP stress-ng: MEMFD stress-ng: CPU Stress stress-ng: System V Message Passing stress-ng: Glibc Qsort Data Sorting stress-ng: SENDFILE stress-ng: Vector Math stress-ng: Glibc C String Functions stress-ng: Crypto stress-ng: Matrix Math stress-ng: Semaphores rav1e: 6 onednn: Convolution Batch Shapes Auto - u8s8f32 - CPU pyperformance: json_loads onednn: IP Shapes 3D - u8s8f32 - CPU pyperformance: chaos pybench: Total For Average Test Times ospray: San Miguel - SciVis pyperformance: nbody dacapobench: Eclipse pyperformance: float npb: IS.D pyperformance: crypto_pyaes oidn: RTLightmap.hdr.4096x4096 stockfish: Total Time npb: SP.B openssl: openssl: blender: BMW27 - CPU-Only incompact3d: input.i3d 193 Cells Per Direction mt-dgemm: Sustained Floating-Point Rate onednn: Matrix Multiply Batch Shapes Transformer - u8s8f32 - CPU rav1e: 10 stress-ng: RdRand svt-av1: Preset 8 - Bosphorus 4K sockperf: Latency Ping Pong oidn: RT.hdr_alb_nrm.3840x2160 oidn: RT.ldr_alb_nrm.3840x2160 embree: Pathtracer - Asian Dragon octave-benchmark: embree: Pathtracer ISPC - Asian Dragon opencv: DNN - Deep Neural Network svt-vp9: PSNR/SSIM Optimized - Bosphorus 1080p ospray: NASA Streamlines - Path Tracer embree: Pathtracer - Crown dacapobench: Jython embree: Pathtracer ISPC - Crown ospray: Magnetic Reconnection - SciVis svt-vp9: VMAF Optimized - Bosphorus 1080p svt-vp9: Visual Quality Optimized - Bosphorus 1080p ospray: NASA Streamlines - SciVis svt-hevc: 7 - Bosphorus 1080p onednn: Deconvolution Batch shapes_3d - u8s8f32 - CPU primesieve: 1e12 Prime Number Generation svt-hevc: 10 - Bosphorus 1080p ospray: Magnetic Reconnection - Path Tracer build-godot: Xeon Platinum 8380 2P EPYC 7763 2P CentOS Stream 8 CentOS Stream 8 3367 810.235 21541 14749.8 182.860 95665 1709.891 25426 654.209 48.137 382272 149913 38.69 24.41 26.61 27.09 10.11 14.83 31.73 23.19 6.88 14.50 11.11 11.96 11.72 11.82 20.73 12604.4 8372.03 464386.54 22.902 1.925 2.669 4.619 19.434 3.828 2.148 474 309851.24 190 10328 246.83 3801.3 6.01 359.410 1126290.26 319.572357 6.51 9.05 671 2674.3 18.3 72993 35.361 10804 40.6329 442.207 6.71 672585 34025.3 32947.8 47883.0 192.414095 180.010 580890 74916.4 17.63 3240.04 3035.3 2597.65 782586.56 3001.2 393.8 122.345 1086.451 621 2.078 11298849.84 122.715036 113997.00 664.114 10.5 198662.90 146927.60 196548.34 185945.29 142666.94 27.30 84.857 2357157.42 21.235 527747.44 17.371 355 327763 3843 14.31 0.363794 77.56 301 24243.78 49340.41 2.97357074 38431.53 52.24 12.89 61.53 39569.95 19830.2 17972.9 62.60 368659 477974 10.53 2845 2169.8 208 1829369.29 8.654 33.423 2611.81 2521.5 67.1 1.50147 4.808 602.6 363.3 368.1 562.2 3223.8 3198.4 363.0 367.9 560.8 602.4 3637.2 3651.8 602629 1535250 20.2 78.5 0.3949 58.134937 63152.48 44883.51 553 639429 10.00 18173.0 11254.8 0.27519 22885.75 8486.63 1446451541.16 3281.97 2826.87 112721.70 4971692.52 978.75 1034645.19 225795.62 7757894.79 12276.06 333297.17 12394981.29 2.797 1.19680 26.9 0.435150 127 1167 100 123 16229 120 2029.92 114 1.40 180099073 23554.85 1182907.4 17870.9 23.23 12.0736825 44.409275 0.226158 7.347 669474.48 58.515 5.229 2.93 2.96 83.5796 6.697 108.2310 9101 465.68 28.57 63.8227 3613 73.5347 125 476.33 359.89 142.86 299.17 0.189266 3.498 574.56 500 5456 1052.547 19378 36819.3 184.337 146223 2048.089 23334 706.430 121.668 74838 116.14 73.12 50.27 98.43 15.47 36.41 41.44 65.91 19.90 58.65 46.94 35.16 48.66 50.93 92.94 13224.0 8717.81 334361.81 31.184 4.460 5.120 8.089 35.929 9.506 6.876 193 209135.79 234 4485 180.32 6735.1 5.86 344.534 803762.14 345.578715 8.41 14.31 327 3340.9 12.9 35.492 4110 36.7376 1338.15 8.66 604490 50910.2 51660.2 59188.8 347.519804 167.799 655650 135223 39.74 3235.89 3519.2 3175.99 500099.84 3816.0 287.6 93.322 257.246 613 1.768 70077496.40 153.441231 106659.27 3033.04 12.6 93923.86 107388.17 94578.25 96705.09 120100.53 22.97 67.679 1763052.44 18.737 753633.00 16.892 375 345024 5264 21.94 0.825536 44.71 303 35030.13 74487.14 3.44985785 39242.31 45.08 17.17 36.15 63413.69 44990.1 45968.3 51.51 690972 792264 12.20 3494.6 2754.8 213 1463413.61 9.282 29.476 3244.22 3267.9 69.8 1.55423 5.066 724.8 376.5 376.8 712.1 1668.2 1666.5 377.4 376.0 711.2 725.6 1968.0 1971.7 644885 1863599 20.7 77.1 0.2405 47.489755 34192.06 48659.62 567 552170 8.25 6070.39 5689.70 0.22966 25292.48 11617.46 2941712927.33 4001.02 1438.30 245332.50 4524150.60 1346.44 1779916.24 439985.85 14162266.61 27068.88 440217.92 19337664.16 2.368 1.57892 27.6 0.908802 137 1150 125 136 17900 126 2701.61 112 1.26 297063599 30167.43 1851481.5 28503.6 14.48 12.1906455 45.449912 0.700087 5.527 63.948 4.236 2.53 2.54 72.1521 5.641 71.0323 424.87 38.46 110.1086 3995 100.4225 67.20 467.87 350.67 166.67 329.56 0.541402 2.449 591.35 500 OpenBenchmarking.org
DaCapo Benchmark Java Test: Tradebeans EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org msec, Fewer Is Better DaCapo Benchmark 9.12-MR1 Java Test: Tradebeans CentOS Stream 8 1200 2400 3600 4800 6000 SE +/- 191.21, N = 16 SE +/- 70.75, N = 20 5456 3367
PostgreSQL pgbench Scaling Factor: 1000 - Clients: 250 - Mode: Read Write - Average Latency EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better PostgreSQL pgbench 14.0 Scaling Factor: 1000 - Clients: 250 - Mode: Read Write - Average Latency CentOS Stream 8 200 400 600 800 1000 SE +/- 183.30, N = 12 SE +/- 140.61, N = 12 1052.55 810.24 1. (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 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org TPS, More Is Better PostgreSQL pgbench 14.0 Scaling Factor: 1000 - Clients: 250 - Mode: Read Write CentOS Stream 8 5K 10K 15K 20K 25K SE +/- 10026.46, N = 12 SE +/- 11129.26, N = 12 19378 21541 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lpthread -lrt -ldl -lm
Renaissance Test: Akka Unbalanced Cobwebbed Tree EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better Renaissance 0.12 Test: Akka Unbalanced Cobwebbed Tree CentOS Stream 8 8K 16K 24K 32K 40K SE +/- 133.92, N = 3 SE +/- 127.73, N = 9 36819.3 14749.8
PostgreSQL pgbench Scaling Factor: 1000 - Clients: 500 - Mode: Read Only - Average Latency EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better PostgreSQL pgbench 14.0 Scaling Factor: 1000 - Clients: 500 - Mode: Read Only - Average Latency CentOS Stream 8 40 80 120 160 200 SE +/- 25.01, N = 12 SE +/- 22.86, N = 9 184.34 182.86 1. (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 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org TPS, More Is Better PostgreSQL pgbench 14.0 Scaling Factor: 1000 - Clients: 500 - Mode: Read Only CentOS Stream 8 30K 60K 90K 120K 150K SE +/- 97067.17, N = 12 SE +/- 93232.04, N = 9 146223 95665 1. (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 - Average Latency EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better PostgreSQL pgbench 14.0 Scaling Factor: 1000 - Clients: 500 - Mode: Read Write - Average Latency CentOS Stream 8 400 800 1200 1600 2000 SE +/- 510.29, N = 9 SE +/- 363.48, N = 12 2048.09 1709.89 1. (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 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org TPS, More Is Better PostgreSQL pgbench 14.0 Scaling Factor: 1000 - Clients: 500 - Mode: Read Write CentOS Stream 8 5K 10K 15K 20K 25K SE +/- 11591.84, N = 9 SE +/- 10758.95, N = 12 23334 25426 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lpthread -lrt -ldl -lm
Timed GCC Compilation Time To Compile EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Seconds, Fewer Is Better Timed GCC Compilation 11.2.0 Time To Compile CentOS Stream 8 150 300 450 600 750 SE +/- 1.40, N = 3 SE +/- 0.56, N = 3 706.43 654.21
PostgreSQL pgbench Scaling Factor: 1000 - Clients: 250 - Mode: Read Only - Average Latency EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better PostgreSQL pgbench 14.0 Scaling Factor: 1000 - Clients: 250 - Mode: Read Only - Average Latency CentOS Stream 8 30 60 90 120 150 SE +/- 15.76, N = 9 SE +/- 12.04, N = 9 121.67 48.14 1. (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 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org TPS, More Is Better PostgreSQL pgbench 14.0 Scaling Factor: 1000 - Clients: 250 - Mode: Read Only CentOS Stream 8 80K 160K 240K 320K 400K SE +/- 72993.53, N = 9 SE +/- 191388.82, N = 9 74838 382272 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lpthread -lrt -ldl -lm
OpenCV Test: Features 2D OpenBenchmarking.org ms, Fewer Is Better OpenCV 4.5.4 Test: Features 2D Xeon Platinum 8380 2P: CentOS Stream 8 30K 60K 90K 120K 150K SE +/- 4393.54, N = 12 149913 1. (CXX) g++ options: -fPIC -fsigned-char -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -O3 -shared
NCNN Target: CPU - Model: regnety_400m EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU - Model: regnety_400m CentOS Stream 8 30 60 90 120 150 SE +/- 2.41, N = 9 SE +/- 0.49, N = 4 116.14 38.69 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
NCNN Target: CPU - Model: squeezenet_ssd EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU - Model: squeezenet_ssd CentOS Stream 8 16 32 48 64 80 SE +/- 7.24, N = 9 SE +/- 0.25, N = 4 73.12 24.41 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
NCNN Target: CPU - Model: yolov4-tiny EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU - Model: yolov4-tiny CentOS Stream 8 11 22 33 44 55 SE +/- 3.11, N = 9 SE +/- 0.17, N = 4 50.27 26.61 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
NCNN Target: CPU - Model: resnet50 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU - Model: resnet50 CentOS Stream 8 20 40 60 80 100 SE +/- 14.06, N = 9 SE +/- 0.43, N = 4 98.43 27.09 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
NCNN Target: CPU - Model: alexnet EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU - Model: alexnet CentOS Stream 8 4 8 12 16 20 SE +/- 0.80, N = 9 SE +/- 0.39, N = 4 15.47 10.11 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
NCNN Target: CPU - Model: resnet18 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU - Model: resnet18 CentOS Stream 8 8 16 24 32 40 SE +/- 1.55, N = 9 SE +/- 0.47, N = 4 36.41 14.83 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
NCNN Target: CPU - Model: vgg16 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU - Model: vgg16 CentOS Stream 8 9 18 27 36 45 SE +/- 1.42, N = 9 SE +/- 0.71, N = 4 41.44 31.73 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
NCNN Target: CPU - Model: googlenet EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU - Model: googlenet CentOS Stream 8 15 30 45 60 75 SE +/- 3.47, N = 9 SE +/- 0.82, N = 4 65.91 23.19 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
NCNN Target: CPU - Model: blazeface EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU - Model: blazeface CentOS Stream 8 5 10 15 20 25 SE +/- 0.54, N = 9 SE +/- 0.11, N = 4 19.90 6.88 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
NCNN Target: CPU - Model: efficientnet-b0 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU - Model: efficientnet-b0 CentOS Stream 8 13 26 39 52 65 SE +/- 1.83, N = 9 SE +/- 0.27, N = 4 58.65 14.50 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
NCNN Target: CPU - Model: mnasnet EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU - Model: mnasnet CentOS Stream 8 11 22 33 44 55 SE +/- 0.99, N = 9 SE +/- 0.12, N = 4 46.94 11.11 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
NCNN Target: CPU - Model: shufflenet-v2 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU - Model: shufflenet-v2 CentOS Stream 8 8 16 24 32 40 SE +/- 0.72, N = 9 SE +/- 0.30, N = 4 35.16 11.96 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
NCNN Target: CPU-v3-v3 - Model: mobilenet-v3 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU-v3-v3 - Model: mobilenet-v3 CentOS Stream 8 11 22 33 44 55 SE +/- 3.07, N = 9 SE +/- 0.07, N = 4 48.66 11.72 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
NCNN Target: CPU-v2-v2 - Model: mobilenet-v2 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU-v2-v2 - Model: mobilenet-v2 CentOS Stream 8 11 22 33 44 55 SE +/- 2.60, N = 9 SE +/- 0.14, N = 4 50.93 11.82 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
NCNN Target: CPU - Model: mobilenet EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better NCNN 20210720 Target: CPU - Model: mobilenet CentOS Stream 8 20 40 60 80 100 SE +/- 5.90, N = 9 SE +/- 0.23, N = 4 92.94 20.73 1. (CXX) g++ options: -O3 -rdynamic -lgomp -lpthread -pthread
ASKAP Test: tConvolve MT - Degridding EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Million Grid Points Per Second, More Is Better ASKAP 1.0 Test: tConvolve MT - Degridding CentOS Stream 8 3K 6K 9K 12K 15K SE +/- 126.65, N = 3 SE +/- 373.79, N = 9 13224.0 12604.4 1. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp
ASKAP Test: tConvolve MT - Gridding EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Million Grid Points Per Second, More Is Better ASKAP 1.0 Test: tConvolve MT - Gridding CentOS Stream 8 2K 4K 6K 8K 10K SE +/- 41.20, N = 3 SE +/- 172.71, N = 9 8717.81 8372.03 1. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp
Redis Memtier / Redis Benchmark Test: MIX EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Operations Per Second, More Is Better Redis Memtier / Redis Benchmark Test: MIX CentOS Stream 8 100K 200K 300K 400K 500K SE +/- 8460.70, N = 12 SE +/- 8588.53, N = 12 334361.81 464386.54 1. (CXX) g++ options: -O2 -levent -levent_openssl -lcrypto -lssl -lpthread -lz -lpcre 2. Redis server v=5.0.3 sha=00000000:0 malloc=jemalloc-5.1.0 bits=64 build=28849dbea6f07cc8
Mobile Neural Network Model: inception-v3 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better Mobile Neural Network 1.2 Model: inception-v3 CentOS Stream 8 7 14 21 28 35 SE +/- 0.20, N = 12 SE +/- 0.07, N = 15 31.18 22.90 1. (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: mobilenet-v1-1.0 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better Mobile Neural Network 1.2 Model: mobilenet-v1-1.0 CentOS Stream 8 1.0035 2.007 3.0105 4.014 5.0175 SE +/- 0.132, N = 12 SE +/- 0.008, N = 15 4.460 1.925 1. (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 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better Mobile Neural Network 1.2 Model: MobileNetV2_224 CentOS Stream 8 1.152 2.304 3.456 4.608 5.76 SE +/- 0.151, N = 12 SE +/- 0.035, N = 15 5.120 2.669 1. (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 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better Mobile Neural Network 1.2 Model: SqueezeNetV1.0 CentOS Stream 8 2 4 6 8 10 SE +/- 0.119, N = 12 SE +/- 0.141, N = 15 8.089 4.619 1. (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 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better Mobile Neural Network 1.2 Model: resnet-v2-50 CentOS Stream 8 8 16 24 32 40 SE +/- 4.72, N = 12 SE +/- 0.05, N = 15 35.93 19.43 1. (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 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better Mobile Neural Network 1.2 Model: squeezenetv1.1 CentOS Stream 8 3 6 9 12 15 SE +/- 1.075, N = 12 SE +/- 0.052, N = 15 9.506 3.828 1. (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 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better Mobile Neural Network 1.2 Model: mobilenetV3 CentOS Stream 8 2 4 6 8 10 SE +/- 0.985, N = 12 SE +/- 0.025, N = 15 6.876 2.148 1. (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
ONNX Runtime Model: fcn-resnet101-11 - Device: CPU EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.10 Model: fcn-resnet101-11 - Device: CPU CentOS Stream 8 100 200 300 400 500 SE +/- 2.15, N = 12 SE +/- 15.75, N = 12 193 474 1. (CXX) g++ options: -ffunction-sections -fdata-sections -march=native -mtune=native -O3 -flto -fno-fat-lto-objects -ldl -lrt -pthread -lpthread
Redis Memtier / Redis Benchmark Test: SET EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Operations Per Second, More Is Better Redis Memtier / Redis Benchmark Test: SET CentOS Stream 8 70K 140K 210K 280K 350K SE +/- 7402.50, N = 12 SE +/- 10740.70, N = 9 209135.79 309851.24 1. (CXX) g++ options: -O2 -levent -levent_openssl -lcrypto -lssl -lpthread -lz -lpcre 2. Redis server v=5.0.3 sha=00000000:0 malloc=jemalloc-5.1.0 bits=64 build=28849dbea6f07cc8
Monte Carlo Simulations of Ionised Nebulae Input: Dust 2D tau100.0 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Seconds, Fewer Is Better Monte Carlo Simulations of Ionised Nebulae 2019-03-24 Input: Dust 2D tau100.0 CentOS Stream 8 50 100 150 200 250 SE +/- 2.65, N = 9 234 190 1. (F9X) gfortran options: -cpp -Jsource/ -ffree-line-length-0 -lm -std=legacy -O3 -O2 -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi
ONNX Runtime Model: shufflenet-v2-10 - Device: CPU EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.10 Model: shufflenet-v2-10 - Device: CPU CentOS Stream 8 2K 4K 6K 8K 10K SE +/- 303.89, N = 9 SE +/- 172.85, N = 12 4485 10328 1. (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 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.0 Blend File: Barbershop - Compute: CPU-Only CentOS Stream 8 50 100 150 200 250 SE +/- 25.57, N = 9 SE +/- 0.29, N = 3 180.32 246.83
Renaissance Test: In-Memory Database Shootout EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better Renaissance 0.12 Test: In-Memory Database Shootout CentOS Stream 8 1400 2800 4200 5600 7000 SE +/- 50.27, N = 10 SE +/- 38.74, N = 15 6735.1 3801.3
PlaidML FP16: No - Mode: Inference - Network: ResNet 50 - Device: CPU EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org FPS, More Is Better PlaidML FP16: No - Mode: Inference - Network: ResNet 50 - Device: CPU CentOS Stream 8 2 4 6 8 10 SE +/- 0.04, N = 3 SE +/- 0.02, N = 3 5.86 6.01
GPAW Input: Carbon Nanotube EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Seconds, Fewer Is Better GPAW 20.1 Input: Carbon Nanotube CentOS Stream 8 80 160 240 320 400 SE +/- 0.69, N = 3 SE +/- 0.67, N = 3 344.53 359.41 1. (CC) gcc options: -pthread -shared -lxc -lblas -lpython3.6m -lmpi -O2 -pipe -fexceptions -m64 -mtune=generic -fcf-protection -fPIC -fwrapv -UNDEBUG -std=c99
Redis Memtier / Redis Benchmark Test: LPUSH and LPOP: lpush EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Requests Per Second, More Is Better Redis Memtier / Redis Benchmark Test: LPUSH and LPOP: lpush CentOS Stream 8 200K 400K 600K 800K 1000K SE +/- 15174.05, N = 15 SE +/- 25162.13, N = 15 803762.14 1126290.26 1. (CXX) g++ options: -O2 -levent -levent_openssl -lcrypto -lssl -lpthread -lz -lpcre 2. Redis server v=5.0.3 sha=00000000:0 malloc=jemalloc-5.1.0 bits=64 build=28849dbea6f07cc8
Xcompact3d Incompact3d Input: X3D-benchmarking input.i3d EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Seconds, Fewer Is Better Xcompact3d Incompact3d 2021-03-11 Input: X3D-benchmarking input.i3d CentOS Stream 8 80 160 240 320 400 SE +/- 0.09, N = 3 SE +/- 0.11, N = 3 345.58 319.57 1. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi
LuxCoreRender Scene: Danish Mood - Acceleration: CPU EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org M samples/sec, More Is Better LuxCoreRender 2.5 Scene: Danish Mood - Acceleration: CPU CentOS Stream 8 2 4 6 8 10 SE +/- 0.13, N = 15 SE +/- 0.05, N = 15 8.41 6.51
LuxCoreRender Scene: DLSC - Acceleration: CPU EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org M samples/sec, More Is Better LuxCoreRender 2.5 Scene: DLSC - Acceleration: CPU CentOS Stream 8 4 8 12 16 20 SE +/- 0.31, N = 15 SE +/- 0.06, N = 15 14.31 9.05
ONNX Runtime Model: yolov4 - Device: CPU EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.10 Model: yolov4 - Device: CPU CentOS Stream 8 140 280 420 560 700 SE +/- 6.47, N = 12 SE +/- 7.12, N = 3 327 671 1. (CXX) g++ options: -ffunction-sections -fdata-sections -march=native -mtune=native -O3 -flto -fno-fat-lto-objects -ldl -lrt -pthread -lpthread
Zstd Compression Compression Level: 19, Long Mode - Decompression Speed EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org MB/s, More Is Better Zstd Compression Compression Level: 19, Long Mode - Decompression Speed CentOS Stream 8 700 1400 2100 2800 3500 SE +/- 7.85, N = 12 SE +/- 9.70, N = 7 3340.9 2674.3 1. *** zstd command line interface 64-bits v1.4.4, by Yann Collet ***
Zstd Compression Compression Level: 19, Long Mode - Compression Speed EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org MB/s, More Is Better Zstd Compression Compression Level: 19, Long Mode - Compression Speed CentOS Stream 8 5 10 15 20 25 SE +/- 0.22, N = 12 SE +/- 0.17, N = 7 12.9 18.3 1. *** zstd command line interface 64-bits v1.4.4, by Yann Collet ***
OpenCV Test: Object Detection OpenBenchmarking.org ms, Fewer Is Better OpenCV 4.5.4 Test: Object Detection Xeon Platinum 8380 2P: CentOS Stream 8 16K 32K 48K 64K 80K SE +/- 1456.11, N = 12 72993 1. (CXX) g++ options: -fPIC -fsigned-char -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -O3 -shared
LAMMPS Molecular Dynamics Simulator Model: 20k Atoms EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ns/day, More Is Better LAMMPS Molecular Dynamics Simulator 29Oct2020 Model: 20k Atoms CentOS Stream 8 8 16 24 32 40 SE +/- 0.14, N = 3 SE +/- 0.02, N = 3 35.49 35.36 1. (CXX) g++ options: -O3 -pthread -lm
ONNX Runtime Model: super-resolution-10 - Device: CPU EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.10 Model: super-resolution-10 - Device: CPU CentOS Stream 8 2K 4K 6K 8K 10K SE +/- 295.99, N = 9 SE +/- 8.65, N = 3 4110 10804 1. (CXX) g++ options: -ffunction-sections -fdata-sections -march=native -mtune=native -O3 -flto -fno-fat-lto-objects -ldl -lrt -pthread -lpthread
High Performance Conjugate Gradient EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org GFLOP/s, More Is Better High Performance Conjugate Gradient 3.1 CentOS Stream 8 9 18 27 36 45 SE +/- 0.03, N = 3 SE +/- 0.02, N = 3 36.74 40.63 1. (CXX) g++ options: -O3 -ffast-math -ftree-vectorize -pthread -lmpi_cxx -lmpi
oneDNN Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPU EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.1.2 Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPU CentOS Stream 8 300 600 900 1200 1500 SE +/- 41.11, N = 12 SE +/- 1.93, N = 3 1338.15 442.21 1. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
LuxCoreRender Scene: LuxCore Benchmark - Acceleration: CPU EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org M samples/sec, More Is Better LuxCoreRender 2.5 Scene: LuxCore Benchmark - Acceleration: CPU CentOS Stream 8 2 4 6 8 10 SE +/- 0.02, N = 3 SE +/- 0.13, N = 15 8.66 6.71
TensorFlow Lite Model: Inception V4 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Microseconds, Fewer Is Better TensorFlow Lite 2020-08-23 Model: Inception V4 CentOS Stream 8 140K 280K 420K 560K 700K SE +/- 9303.83, N = 15 SE +/- 4325.52, N = 3 604490 672585
TensorFlow Lite Model: Mobilenet Quant EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Microseconds, Fewer Is Better TensorFlow Lite 2020-08-23 Model: Mobilenet Quant CentOS Stream 8 11K 22K 33K 44K 55K SE +/- 1002.46, N = 15 SE +/- 84.87, N = 3 50910.2 34025.3
TensorFlow Lite Model: Mobilenet Float EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Microseconds, Fewer Is Better TensorFlow Lite 2020-08-23 Model: Mobilenet Float CentOS Stream 8 11K 22K 33K 44K 55K SE +/- 1146.99, N = 15 SE +/- 52.61, N = 3 51660.2 32947.8
TensorFlow Lite Model: SqueezeNet EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Microseconds, Fewer Is Better TensorFlow Lite 2020-08-23 Model: SqueezeNet CentOS Stream 8 13K 26K 39K 52K 65K SE +/- 1414.69, N = 15 SE +/- 105.74, N = 3 59188.8 47883.0
Appleseed Scene: Material Tester EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Seconds, Fewer Is Better Appleseed 2.0 Beta Scene: Material Tester CentOS Stream 8 80 160 240 320 400 347.52 192.41
Timed LLVM Compilation Build System: Unix Makefiles EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Seconds, Fewer Is Better Timed LLVM Compilation 13.0 Build System: Unix Makefiles CentOS Stream 8 40 80 120 160 200 SE +/- 0.51, N = 3 SE +/- 1.48, N = 3 167.80 180.01
TensorFlow Lite Model: Inception ResNet V2 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Microseconds, Fewer Is Better TensorFlow Lite 2020-08-23 Model: Inception ResNet V2 CentOS Stream 8 140K 280K 420K 560K 700K SE +/- 21565.13, N = 12 SE +/- 2874.78, N = 3 655650 580890
TensorFlow Lite Model: NASNet Mobile EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Microseconds, Fewer Is Better TensorFlow Lite 2020-08-23 Model: NASNet Mobile CentOS Stream 8 30K 60K 90K 120K 150K SE +/- 1721.19, N = 12 SE +/- 155.05, N = 3 135223.0 74916.4
Stress-NG Test: CPU Cache EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.13.02 Test: CPU Cache CentOS Stream 8 9 18 27 36 45 SE +/- 2.09, N = 15 SE +/- 0.42, N = 15 39.74 17.63 1. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc
Stress-NG Test: Socket Activity EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.13.02 Test: Socket Activity CentOS Stream 8 700 1400 2100 2800 3500 SE +/- 887.18, N = 15 SE +/- 953.56, N = 15 3235.89 3240.04 1. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc
Renaissance Test: Apache Spark PageRank EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better Renaissance 0.12 Test: Apache Spark PageRank CentOS Stream 8 800 1600 2400 3200 4000 SE +/- 42.84, N = 4 SE +/- 28.17, N = 3 3519.2 3035.3
NAS Parallel Benchmarks Test / Class: EP.D EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: EP.D CentOS Stream 8 700 1400 2100 2800 3500 SE +/- 84.69, N = 12 SE +/- 23.33, N = 3 3175.99 2597.65 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi
Redis Memtier / Redis Benchmark Test: GET EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Operations Per Second, More Is Better Redis Memtier / Redis Benchmark Test: GET CentOS Stream 8 200K 400K 600K 800K 1000K SE +/- 1838.84, N = 3 SE +/- 1984.40, N = 3 500099.84 782586.56 1. (CXX) g++ options: -O2 -levent -levent_openssl -lcrypto -lssl -lpthread -lz -lpcre 2. Redis server v=5.0.3 sha=00000000:0 malloc=jemalloc-5.1.0 bits=64 build=28849dbea6f07cc8
Zstd Compression Compression Level: 8, Long Mode - Decompression Speed EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org MB/s, More Is Better Zstd Compression Compression Level: 8, Long Mode - Decompression Speed CentOS Stream 8 800 1600 2400 3200 4000 SE +/- 3.85, N = 15 SE +/- 6.95, N = 3 3816.0 3001.2 1. *** zstd command line interface 64-bits v1.4.4, by Yann Collet ***
Zstd Compression Compression Level: 8, Long Mode - Compression Speed EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org MB/s, More Is Better Zstd Compression Compression Level: 8, Long Mode - Compression Speed CentOS Stream 8 90 180 270 360 450 SE +/- 3.28, N = 15 SE +/- 0.98, N = 3 287.6 393.8 1. *** zstd command line interface 64-bits v1.4.4, by Yann Collet ***
Timed LLVM Compilation Build System: Ninja EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Seconds, Fewer Is Better Timed LLVM Compilation 13.0 Build System: Ninja CentOS Stream 8 30 60 90 120 150 SE +/- 0.29, N = 3 SE +/- 0.28, N = 3 93.32 122.35
ASKAP Test: Hogbom Clean OpenMP EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Iterations Per Second, More Is Better ASKAP 1.0 Test: Hogbom Clean OpenMP CentOS Stream 8 200 400 600 800 1000 SE +/- 9.50, N = 15 SE +/- 39.14, N = 15 257.25 1086.45 1. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp
PyPerformance Benchmark: raytrace EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.0.0 Benchmark: raytrace CentOS Stream 8 130 260 390 520 650 SE +/- 0.67, N = 3 SE +/- 1.76, N = 3 613 621
rav1e Speed: 5 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Frames Per Second, More Is Better rav1e 0.5 Speed: 5 CentOS Stream 8 0.4676 0.9352 1.4028 1.8704 2.338 SE +/- 0.013, N = 15 SE +/- 0.008, N = 3 1.768 2.078
Stress-NG Test: Context Switching EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.13.02 Test: Context Switching CentOS Stream 8 15M 30M 45M 60M 75M SE +/- 834986.74, N = 4 SE +/- 809501.98, N = 15 70077496.40 11298849.84 1. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc
Appleseed Scene: Emily EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Seconds, Fewer Is Better Appleseed 2.0 Beta Scene: Emily CentOS Stream 8 30 60 90 120 150 153.44 122.72
Apache HTTP Server Concurrent Requests: 1000 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Requests Per Second, More Is Better Apache HTTP Server 2.4.48 Concurrent Requests: 1000 CentOS Stream 8 20K 40K 60K 80K 100K SE +/- 710.10, N = 3 SE +/- 382.73, N = 3 106659.27 113997.00 1. (CC) gcc options: -shared -fPIC -O2 -pthread
oneDNN Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPU EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.1.2 Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPU CentOS Stream 8 700 1400 2100 2800 3500 SE +/- 32.78, N = 4 SE +/- 2.02, N = 3 3033.04 664.11 1. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
PyPerformance Benchmark: python_startup EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.0.0 Benchmark: python_startup CentOS Stream 8 3 6 9 12 15 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 12.6 10.5
nginx Concurrent Requests: 1000 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Requests Per Second, More Is Better nginx 1.21.1 Concurrent Requests: 1000 CentOS Stream 8 40K 80K 120K 160K 200K SE +/- 437.18, N = 3 SE +/- 415.61, N = 3 93923.86 198662.90 1. (CC) gcc options: -ldl -lpthread -lcrypt -lz -O3 -march=native
Apache HTTP Server Concurrent Requests: 500 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Requests Per Second, More Is Better Apache HTTP Server 2.4.48 Concurrent Requests: 500 CentOS Stream 8 30K 60K 90K 120K 150K SE +/- 495.48, N = 3 SE +/- 884.81, N = 3 107388.17 146927.60 1. (CC) gcc options: -shared -fPIC -O2 -pthread
nginx Concurrent Requests: 500 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Requests Per Second, More Is Better nginx 1.21.1 Concurrent Requests: 500 CentOS Stream 8 40K 80K 120K 160K 200K SE +/- 533.88, N = 3 SE +/- 392.84, N = 3 94578.25 196548.34 1. (CC) gcc options: -ldl -lpthread -lcrypt -lz -O3 -march=native
nginx Concurrent Requests: 200 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Requests Per Second, More Is Better nginx 1.21.1 Concurrent Requests: 200 CentOS Stream 8 40K 80K 120K 160K 200K SE +/- 211.37, N = 3 SE +/- 1210.93, N = 3 96705.09 185945.29 1. (CC) gcc options: -ldl -lpthread -lcrypt -lz -O3 -march=native
Apache HTTP Server Concurrent Requests: 200 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Requests Per Second, More Is Better Apache HTTP Server 2.4.48 Concurrent Requests: 200 CentOS Stream 8 30K 60K 90K 120K 150K SE +/- 752.69, N = 3 SE +/- 925.34, N = 3 120100.53 142666.94 1. (CC) gcc options: -shared -fPIC -O2 -pthread
x265 Video Input: Bosphorus 4K EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Frames Per Second, More Is Better x265 3.4 Video Input: Bosphorus 4K CentOS Stream 8 6 12 18 24 30 SE +/- 0.26, N = 5 SE +/- 0.40, N = 15 22.97 27.30 1. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl
Timed Node.js Compilation Time To Compile EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Seconds, Fewer Is Better Timed Node.js Compilation 15.11 Time To Compile CentOS Stream 8 20 40 60 80 100 SE +/- 0.63, N = 3 SE +/- 0.30, N = 3 67.68 84.86
Redis Test: GET EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Requests Per Second, More Is Better Redis 6.0.9 Test: GET CentOS Stream 8 500K 1000K 1500K 2000K 2500K SE +/- 57548.10, N = 15 SE +/- 33069.54, N = 15 1763052.44 2357157.42 1. (CXX) g++ options: -MM -MT -g3 -fvisibility=hidden -O3
Timed Linux Kernel Compilation Time To Compile EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Seconds, Fewer Is Better Timed Linux Kernel Compilation 5.14 Time To Compile CentOS Stream 8 5 10 15 20 25 SE +/- 0.14, N = 14 SE +/- 0.18, N = 8 18.74 21.24
Redis Memtier / Redis Benchmark Test: LPUSH and LPOP: lpop EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Requests Per Second, More Is Better Redis Memtier / Redis Benchmark Test: LPUSH and LPOP: lpop CentOS Stream 8 160K 320K 480K 640K 800K SE +/- 3605.89, N = 3 SE +/- 1237.45, N = 3 753633.00 527747.44 1. (CXX) g++ options: -O2 -levent -levent_openssl -lcrypto -lssl -lpthread -lz -lpcre 2. Redis server v=5.0.3 sha=00000000:0 malloc=jemalloc-5.1.0 bits=64 build=28849dbea6f07cc8
Sockperf Test: Latency Under Load EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org usec, Fewer Is Better Sockperf 3.7 Test: Latency Under Load CentOS Stream 8 4 8 12 16 20 SE +/- 1.02, N = 25 SE +/- 1.82, N = 25 16.89 17.37 1. (CXX) g++ options: --param -O3 -rdynamic -ldl -lpthread
PyPerformance Benchmark: 2to3 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.0.0 Benchmark: 2to3 CentOS Stream 8 80 160 240 320 400 SE +/- 0.58, N = 3 375 355
Sockperf Test: Throughput EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Messages Per Second, More Is Better Sockperf 3.7 Test: Throughput CentOS Stream 8 70K 140K 210K 280K 350K SE +/- 2067.12, N = 25 SE +/- 4047.87, N = 23 345024 327763 1. (CXX) g++ options: --param -O3 -rdynamic -ldl -lpthread
DaCapo Benchmark Java Test: H2 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org msec, Fewer Is Better DaCapo Benchmark 9.12-MR1 Java Test: H2 CentOS Stream 8 1100 2200 3300 4400 5500 SE +/- 76.12, N = 20 SE +/- 49.75, N = 20 5264 3843
LuxCoreRender Scene: Orange Juice - Acceleration: CPU EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org M samples/sec, More Is Better LuxCoreRender 2.5 Scene: Orange Juice - Acceleration: CPU CentOS Stream 8 5 10 15 20 25 SE +/- 0.11, N = 3 SE +/- 0.11, N = 3 21.94 14.31
oneDNN Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPU EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.1.2 Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPU CentOS Stream 8 0.1857 0.3714 0.5571 0.7428 0.9285 SE +/- 0.008999, N = 15 SE +/- 0.000501, N = 3 0.825536 0.363794 1. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
Blender Blend File: Pabellon Barcelona - Compute: CPU-Only EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.0 Blend File: Pabellon Barcelona - Compute: CPU-Only CentOS Stream 8 20 40 60 80 100 SE +/- 0.08, N = 3 SE +/- 0.13, N = 3 44.71 77.56
PyPerformance Benchmark: go EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.0.0 Benchmark: go CentOS Stream 8 70 140 210 280 350 SE +/- 0.58, N = 3 303 301
NAS Parallel Benchmarks Test / Class: SP.C EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: SP.C CentOS Stream 8 8K 16K 24K 32K 40K SE +/- 77.88, N = 3 SE +/- 14.36, N = 3 35030.13 24243.78 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi
NAS Parallel Benchmarks Test / Class: BT.C EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: BT.C CentOS Stream 8 16K 32K 48K 64K 80K SE +/- 90.74, N = 3 SE +/- 24.48, N = 3 74487.14 49340.41 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi
Xcompact3d Incompact3d Input: input.i3d 129 Cells Per Direction EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Seconds, Fewer Is Better Xcompact3d Incompact3d 2021-03-11 Input: input.i3d 129 Cells Per Direction CentOS Stream 8 0.7762 1.5524 2.3286 3.1048 3.881 SE +/- 0.05694678, N = 15 SE +/- 0.04066101, N = 15 3.44985785 2.97357074 1. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi
NAS Parallel Benchmarks Test / Class: FT.C EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: FT.C CentOS Stream 8 8K 16K 24K 32K 40K SE +/- 1314.27, N = 12 SE +/- 25.02, N = 3 39242.31 38431.53 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi
PlaidML FP16: No - Mode: Inference - Network: VGG19 - Device: CPU EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org FPS, More Is Better PlaidML FP16: No - Mode: Inference - Network: VGG19 - Device: CPU CentOS Stream 8 12 24 36 48 60 SE +/- 0.37, N = 3 SE +/- 0.28, N = 3 45.08 52.24
LuxCoreRender Scene: Rainbow Colors and Prism - Acceleration: CPU EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org M samples/sec, More Is Better LuxCoreRender 2.5 Scene: Rainbow Colors and Prism - Acceleration: CPU CentOS Stream 8 4 8 12 16 20 SE +/- 0.57, N = 12 SE +/- 0.85, N = 12 17.17 12.89
Blender Blend File: Classroom - Compute: CPU-Only EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.0 Blend File: Classroom - Compute: CPU-Only CentOS Stream 8 14 28 42 56 70 SE +/- 0.06, N = 3 SE +/- 0.13, N = 3 36.15 61.53
NAS Parallel Benchmarks Test / Class: LU.C EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: LU.C CentOS Stream 8 14K 28K 42K 56K 70K SE +/- 52.36, N = 3 SE +/- 84.61, N = 3 63413.69 39569.95 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi
ASKAP Test: tConvolve MPI - Gridding EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Mpix/sec, More Is Better ASKAP 1.0 Test: tConvolve MPI - Gridding CentOS Stream 8 10K 20K 30K 40K 50K SE +/- 423.01, N = 3 SE +/- 131.68, N = 3 44990.1 19830.2 1. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp
ASKAP Test: tConvolve MPI - Degridding EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Mpix/sec, More Is Better ASKAP 1.0 Test: tConvolve MPI - Degridding CentOS Stream 8 10K 20K 30K 40K 50K SE +/- 167.17, N = 3 SE +/- 71.10, N = 3 45968.3 17972.9 1. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp
PlaidML FP16: No - Mode: Inference - Network: VGG16 - Device: CPU EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org FPS, More Is Better PlaidML FP16: No - Mode: Inference - Network: VGG16 - Device: CPU CentOS Stream 8 14 28 42 56 70 SE +/- 0.21, N = 3 SE +/- 0.27, N = 3 51.51 62.60
7-Zip Compression Test: Decompression Rating EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org MIPS, More Is Better 7-Zip Compression 21.06 Test: Decompression Rating CentOS Stream 8 150K 300K 450K 600K 750K SE +/- 7232.75, N = 4 SE +/- 441.97, N = 3 690972 368659 1. (CXX) g++ options: -lpthread -ldl -O2 -fPIC
7-Zip Compression Test: Compression Rating EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org MIPS, More Is Better 7-Zip Compression 21.06 Test: Compression Rating CentOS Stream 8 200K 400K 600K 800K 1000K SE +/- 8841.92, N = 4 SE +/- 2452.32, N = 3 792264 477974 1. (CXX) g++ options: -lpthread -ldl -O2 -fPIC
OSPray Demo: San Miguel - Renderer: Path Tracer EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org FPS, More Is Better OSPray 1.8.5 Demo: San Miguel - Renderer: Path Tracer CentOS Stream 8 3 6 9 12 15 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 12.20 10.53
Zstd Compression Compression Level: 8 - Decompression Speed EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org MB/s, More Is Better Zstd Compression Compression Level: 8 - Decompression Speed CentOS Stream 8 700 1400 2100 2800 3500 SE +/- 10.76, N = 4 3494.6 2845.0 1. *** zstd command line interface 64-bits v1.4.4, by Yann Collet ***
Zstd Compression Compression Level: 8 - Compression Speed EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org MB/s, More Is Better Zstd Compression Compression Level: 8 - Compression Speed CentOS Stream 8 600 1200 1800 2400 3000 SE +/- 32.84, N = 4 SE +/- 11.63, N = 3 2754.8 2169.8 1. *** zstd command line interface 64-bits v1.4.4, by Yann Collet ***
PyPerformance Benchmark: regex_compile EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.0.0 Benchmark: regex_compile CentOS Stream 8 50 100 150 200 250 SE +/- 0.33, N = 3 213 208
Redis Test: SET EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Requests Per Second, More Is Better Redis 6.0.9 Test: SET CentOS Stream 8 400K 800K 1200K 1600K 2000K SE +/- 68212.04, N = 12 SE +/- 16442.53, N = 3 1463413.61 1829369.29 1. (CXX) g++ options: -MM -MT -g3 -fvisibility=hidden -O3
GROMACS Implementation: MPI CPU - Input: water_GMX50_bare EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Ns Per Day, More Is Better GROMACS 2021.2 Implementation: MPI CPU - Input: water_GMX50_bare CentOS Stream 8 3 6 9 12 15 SE +/- 0.033, N = 3 SE +/- 0.020, N = 3 9.282 8.654 1. (CXX) g++ options: -O3 -pthread
LAMMPS Molecular Dynamics Simulator Model: Rhodopsin Protein EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ns/day, More Is Better LAMMPS Molecular Dynamics Simulator 29Oct2020 Model: Rhodopsin Protein CentOS Stream 8 8 16 24 32 40 SE +/- 0.31, N = 15 SE +/- 0.75, N = 14 29.48 33.42 1. (CXX) g++ options: -O3 -pthread -lm
NAS Parallel Benchmarks Test / Class: EP.C EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: EP.C CentOS Stream 8 700 1400 2100 2800 3500 SE +/- 78.53, N = 15 SE +/- 2.20, N = 3 3244.22 2611.81 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi
Zstd Compression Compression Level: 19 - Decompression Speed EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org MB/s, More Is Better Zstd Compression Compression Level: 19 - Decompression Speed CentOS Stream 8 700 1400 2100 2800 3500 SE +/- 8.99, N = 3 SE +/- 15.70, N = 3 3267.9 2521.5 1. *** zstd command line interface 64-bits v1.4.4, by Yann Collet ***
Zstd Compression Compression Level: 19 - Compression Speed EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org MB/s, More Is Better Zstd Compression Compression Level: 19 - Compression Speed CentOS Stream 8 16 32 48 64 80 SE +/- 0.74, N = 3 SE +/- 0.77, N = 3 69.8 67.1 1. *** zstd command line interface 64-bits v1.4.4, by Yann Collet ***
oneDNN Harness: IP Shapes 1D - Data Type: u8s8f32 - Engine: CPU EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.1.2 Harness: IP Shapes 1D - Data Type: u8s8f32 - Engine: CPU CentOS Stream 8 0.3497 0.6994 1.0491 1.3988 1.7485 SE +/- 0.00429, N = 3 SE +/- 0.06642, N = 12 1.55423 1.50147 1. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
SVT-AV1 Encoder Mode: Preset 4 - Input: Bosphorus 4K EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 0.8.7 Encoder Mode: Preset 4 - Input: Bosphorus 4K CentOS Stream 8 1.1399 2.2798 3.4197 4.5596 5.6995 SE +/- 0.012, N = 3 SE +/- 0.032, N = 3 5.066 4.808 1. (CXX) g++ options: -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq -pie
Cryptsetup Serpent-XTS 512b Encryption EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org MiB/s, More Is Better Cryptsetup Serpent-XTS 512b Encryption CentOS Stream 8 160 320 480 640 800 SE +/- 2.17, N = 3 SE +/- 0.85, N = 2 724.8 602.6
Cryptsetup Twofish-XTS 512b Decryption EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org MiB/s, More Is Better Cryptsetup Twofish-XTS 512b Decryption CentOS Stream 8 80 160 240 320 400 SE +/- 0.59, N = 3 SE +/- 0.22, N = 3 376.5 363.3
Cryptsetup Twofish-XTS 512b Encryption EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org MiB/s, More Is Better Cryptsetup Twofish-XTS 512b Encryption CentOS Stream 8 80 160 240 320 400 SE +/- 0.15, N = 3 SE +/- 0.22, N = 3 376.8 368.1
Cryptsetup Serpent-XTS 512b Decryption EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org MiB/s, More Is Better Cryptsetup Serpent-XTS 512b Decryption CentOS Stream 8 150 300 450 600 750 SE +/- 0.12, N = 3 SE +/- 0.55, N = 3 712.1 562.2
Cryptsetup AES-XTS 512b Decryption EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org MiB/s, More Is Better Cryptsetup AES-XTS 512b Decryption CentOS Stream 8 700 1400 2100 2800 3500 SE +/- 5.57, N = 3 SE +/- 8.37, N = 3 1668.2 3223.8
Cryptsetup AES-XTS 512b Encryption EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org MiB/s, More Is Better Cryptsetup AES-XTS 512b Encryption CentOS Stream 8 700 1400 2100 2800 3500 SE +/- 5.02, N = 3 SE +/- 15.69, N = 3 1666.5 3198.4
Cryptsetup Twofish-XTS 256b Decryption EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org MiB/s, More Is Better Cryptsetup Twofish-XTS 256b Decryption CentOS Stream 8 80 160 240 320 400 SE +/- 0.43, N = 3 SE +/- 0.23, N = 3 377.4 363.0
Cryptsetup Twofish-XTS 256b Encryption EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org MiB/s, More Is Better Cryptsetup Twofish-XTS 256b Encryption CentOS Stream 8 80 160 240 320 400 SE +/- 1.87, N = 3 SE +/- 0.23, N = 3 376.0 367.9
Cryptsetup Serpent-XTS 256b Decryption EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org MiB/s, More Is Better Cryptsetup Serpent-XTS 256b Decryption CentOS Stream 8 150 300 450 600 750 SE +/- 1.05, N = 3 SE +/- 0.12, N = 3 711.2 560.8
Cryptsetup Serpent-XTS 256b Encryption EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org MiB/s, More Is Better Cryptsetup Serpent-XTS 256b Encryption CentOS Stream 8 160 320 480 640 800 SE +/- 1.07, N = 3 SE +/- 0.33, N = 3 725.6 602.4
Cryptsetup AES-XTS 256b Decryption EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org MiB/s, More Is Better Cryptsetup AES-XTS 256b Decryption CentOS Stream 8 800 1600 2400 3200 4000 SE +/- 6.88, N = 3 SE +/- 22.87, N = 3 1968.0 3637.2
Cryptsetup AES-XTS 256b Encryption EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org MiB/s, More Is Better Cryptsetup AES-XTS 256b Encryption CentOS Stream 8 800 1600 2400 3200 4000 SE +/- 4.85, N = 3 SE +/- 13.26, N = 3 1971.7 3651.8
Cryptsetup PBKDF2-whirlpool EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Iterations Per Second, More Is Better Cryptsetup PBKDF2-whirlpool CentOS Stream 8 140K 280K 420K 560K 700K SE +/- 1213.62, N = 3 644885 602629
Cryptsetup PBKDF2-sha512 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Iterations Per Second, More Is Better Cryptsetup PBKDF2-sha512 CentOS Stream 8 400K 800K 1200K 1600K 2000K SE +/- 3974.38, N = 3 1863599 1535250
PyPerformance Benchmark: pathlib EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.0.0 Benchmark: pathlib CentOS Stream 8 5 10 15 20 25 SE +/- 0.03, N = 3 SE +/- 0.00, N = 3 20.7 20.2
PyPerformance Benchmark: django_template EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.0.0 Benchmark: django_template CentOS Stream 8 20 40 60 80 100 SE +/- 0.12, N = 3 SE +/- 0.12, N = 3 77.1 78.5
R Benchmark EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Seconds, Fewer Is Better R Benchmark CentOS Stream 8 0.0889 0.1778 0.2667 0.3556 0.4445 SE +/- 0.0010, N = 3 SE +/- 0.0015, N = 3 0.2405 0.3949 1. R scripting front-end version 4.1.2 (2021-11-01)
Appleseed Scene: Disney Material EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Seconds, Fewer Is Better Appleseed 2.0 Beta Scene: Disney Material CentOS Stream 8 13 26 39 52 65 47.49 58.13
Stress-NG Test: Forking EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.13.02 Test: Forking CentOS Stream 8 14K 28K 42K 56K 70K SE +/- 411.00, N = 4 SE +/- 29.17, N = 3 34192.06 63152.48 1. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc
NAS Parallel Benchmarks Test / Class: MG.C EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: MG.C CentOS Stream 8 10K 20K 30K 40K 50K SE +/- 883.57, N = 12 SE +/- 39.43, N = 3 48659.62 44883.51 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi
PyPerformance Benchmark: pickle_pure_python EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.0.0 Benchmark: pickle_pure_python CentOS Stream 8 120 240 360 480 600 SE +/- 0.67, N = 3 567 553
PHPBench PHP Benchmark Suite EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Score, More Is Better PHPBench 0.8.1 PHP Benchmark Suite CentOS Stream 8 140K 280K 420K 560K 700K SE +/- 1363.68, N = 3 SE +/- 2276.07, N = 3 552170 639429
Stress-NG Test: NUMA EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.13.02 Test: NUMA CentOS Stream 8 3 6 9 12 15 SE +/- 0.02, N = 3 SE +/- 0.13, N = 3 8.25 10.00 1. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc
ASKAP Test: tConvolve OpenMP - Degridding EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Million Grid Points Per Second, More Is Better ASKAP 1.0 Test: tConvolve OpenMP - Degridding CentOS Stream 8 4K 8K 12K 16K 20K SE +/- 163.43, N = 15 SE +/- 422.63, N = 3 6070.39 18173.00 1. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp
ASKAP Test: tConvolve OpenMP - Gridding EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Million Grid Points Per Second, More Is Better ASKAP 1.0 Test: tConvolve OpenMP - Gridding CentOS Stream 8 2K 4K 6K 8K 10K SE +/- 266.31, N = 15 SE +/- 160.77, N = 3 5689.70 11254.80 1. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp
NAMD ATPase Simulation - 327,506 Atoms EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org days/ns, Fewer Is Better NAMD 2.14 ATPase Simulation - 327,506 Atoms CentOS Stream 8 0.0619 0.1238 0.1857 0.2476 0.3095 SE +/- 0.00044, N = 3 SE +/- 0.00054, N = 3 0.22966 0.27519
NAS Parallel Benchmarks Test / Class: CG.C EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: CG.C CentOS Stream 8 5K 10K 15K 20K 25K SE +/- 182.81, N = 3 SE +/- 166.54, N = 15 25292.48 22885.75 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi
Stress-NG Test: Memory Copying EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.13.02 Test: Memory Copying CentOS Stream 8 2K 4K 6K 8K 10K SE +/- 24.77, N = 3 SE +/- 12.87, N = 3 11617.46 8486.63 1. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc
Stress-NG Test: Malloc EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.13.02 Test: Malloc CentOS Stream 8 600M 1200M 1800M 2400M 3000M SE +/- 2008786.61, N = 3 SE +/- 3355046.59, N = 3 2941712927.33 1446451541.16 1. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc
Stress-NG Test: MMAP EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.13.02 Test: MMAP CentOS Stream 8 900 1800 2700 3600 4500 SE +/- 13.65, N = 3 SE +/- 3.28, N = 3 4001.02 3281.97 1. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc
Stress-NG Test: MEMFD EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.13.02 Test: MEMFD CentOS Stream 8 600 1200 1800 2400 3000 SE +/- 2.31, N = 3 SE +/- 2.47, N = 3 1438.30 2826.87 1. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc
Stress-NG Test: CPU Stress EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.13.02 Test: CPU Stress CentOS Stream 8 50K 100K 150K 200K 250K SE +/- 376.22, N = 3 SE +/- 149.97, N = 3 245332.50 112721.70 1. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc
Stress-NG Test: System V Message Passing EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.13.02 Test: System V Message Passing CentOS Stream 8 1.1M 2.2M 3.3M 4.4M 5.5M SE +/- 1927.81, N = 3 SE +/- 5457.50, N = 3 4524150.60 4971692.52 1. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc
Stress-NG Test: Glibc Qsort Data Sorting EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.13.02 Test: Glibc Qsort Data Sorting CentOS Stream 8 300 600 900 1200 1500 SE +/- 0.90, N = 3 SE +/- 0.88, N = 3 1346.44 978.75 1. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc
Stress-NG Test: SENDFILE EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.13.02 Test: SENDFILE CentOS Stream 8 400K 800K 1200K 1600K 2000K SE +/- 12384.02, N = 3 SE +/- 5479.93, N = 3 1779916.24 1034645.19 1. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc
Stress-NG Test: Vector Math EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.13.02 Test: Vector Math CentOS Stream 8 90K 180K 270K 360K 450K SE +/- 154.22, N = 3 SE +/- 54.64, N = 3 439985.85 225795.62 1. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc
Stress-NG Test: Glibc C String Functions EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.13.02 Test: Glibc C String Functions CentOS Stream 8 3M 6M 9M 12M 15M SE +/- 40185.46, N = 3 SE +/- 52776.39, N = 3 14162266.61 7757894.79 1. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc
Stress-NG Test: Crypto EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.13.02 Test: Crypto CentOS Stream 8 6K 12K 18K 24K 30K SE +/- 42.88, N = 3 SE +/- 1.08, N = 3 27068.88 12276.06 1. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc
Stress-NG Test: Matrix Math EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.13.02 Test: Matrix Math CentOS Stream 8 90K 180K 270K 360K 450K SE +/- 287.18, N = 3 SE +/- 1114.79, N = 3 440217.92 333297.17 1. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc
Stress-NG Test: Semaphores EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.13.02 Test: Semaphores CentOS Stream 8 4M 8M 12M 16M 20M SE +/- 116082.15, N = 3 SE +/- 10851.57, N = 3 19337664.16 12394981.29 1. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc
rav1e Speed: 6 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Frames Per Second, More Is Better rav1e 0.5 Speed: 6 CentOS Stream 8 0.6293 1.2586 1.8879 2.5172 3.1465 SE +/- 0.028, N = 4 SE +/- 0.013, N = 3 2.368 2.797
oneDNN Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPU EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.1.2 Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPU CentOS Stream 8 0.3553 0.7106 1.0659 1.4212 1.7765 SE +/- 0.05247, N = 15 SE +/- 0.01980, N = 12 1.57892 1.19680 1. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
PyPerformance Benchmark: json_loads EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.0.0 Benchmark: json_loads CentOS Stream 8 6 12 18 24 30 SE +/- 0.09, N = 3 SE +/- 0.03, N = 3 27.6 26.9
oneDNN Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPU EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.1.2 Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPU CentOS Stream 8 0.2045 0.409 0.6135 0.818 1.0225 SE +/- 0.007250, N = 15 SE +/- 0.004759, N = 3 0.908802 0.435150 1. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
PyPerformance Benchmark: chaos EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.0.0 Benchmark: chaos CentOS Stream 8 30 60 90 120 150 137 127
PyBench Total For Average Test Times EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Milliseconds, Fewer Is Better PyBench 2018-02-16 Total For Average Test Times CentOS Stream 8 300 600 900 1200 1500 SE +/- 1.86, N = 3 SE +/- 2.19, N = 3 1150 1167
OSPray Demo: San Miguel - Renderer: SciVis EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org FPS, More Is Better OSPray 1.8.5 Demo: San Miguel - Renderer: SciVis CentOS Stream 8 30 60 90 120 150 125 100
PyPerformance Benchmark: nbody EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.0.0 Benchmark: nbody CentOS Stream 8 30 60 90 120 150 136 123
DaCapo Benchmark Java Test: Eclipse EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org msec, Fewer Is Better DaCapo Benchmark 9.12-MR1 Java Test: Eclipse CentOS Stream 8 4K 8K 12K 16K 20K SE +/- 109.08, N = 4 SE +/- 134.99, N = 4 17900 16229
PyPerformance Benchmark: float EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.0.0 Benchmark: float CentOS Stream 8 30 60 90 120 150 SE +/- 0.33, N = 3 126 120
NAS Parallel Benchmarks Test / Class: IS.D EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: IS.D CentOS Stream 8 600 1200 1800 2400 3000 SE +/- 4.62, N = 3 SE +/- 5.09, N = 3 2701.61 2029.92 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi
PyPerformance Benchmark: crypto_pyaes EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.0.0 Benchmark: crypto_pyaes CentOS Stream 8 30 60 90 120 150 112 114
Intel Open Image Denoise Run: RTLightmap.hdr.4096x4096 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Images / Sec, More Is Better Intel Open Image Denoise 1.4.0 Run: RTLightmap.hdr.4096x4096 CentOS Stream 8 0.315 0.63 0.945 1.26 1.575 SE +/- 0.00, N = 3 SE +/- 0.02, N = 3 1.26 1.40
Stockfish Total Time EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Nodes Per Second, More Is Better Stockfish 13 Total Time CentOS Stream 8 60M 120M 180M 240M 300M SE +/- 2759095.27, N = 3 SE +/- 1573780.07, N = 3 297063599 180099073 1. (CXX) g++ options: -lgcov -m64 -lpthread -fno-exceptions -std=c++17 -fprofile-use -fno-peel-loops -fno-tracer -pedantic -O3 -msse -msse3 -mpopcnt -mavx2 -mavx512f -mavx512bw -mavx512vnni -mavx512dq -mavx512vl -msse4.1 -mssse3 -msse2 -mbmi2 -flto -flto=jobserver
NAS Parallel Benchmarks Test / Class: SP.B EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: SP.B CentOS Stream 8 6K 12K 18K 24K 30K SE +/- 174.51, N = 3 SE +/- 11.94, N = 3 30167.43 23554.85 1. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi
OpenSSL EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org verify/s, More Is Better OpenSSL CentOS Stream 8 400K 800K 1200K 1600K 2000K SE +/- 1553.72, N = 3 SE +/- 728.21, N = 3 1851481.5 1182907.4 1. OpenSSL 1.1.1k FIPS 25 Mar 2021
OpenSSL EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org sign/s, More Is Better OpenSSL CentOS Stream 8 6K 12K 18K 24K 30K SE +/- 15.51, N = 3 SE +/- 48.50, N = 3 28503.6 17870.9 1. OpenSSL 1.1.1k FIPS 25 Mar 2021
Blender Blend File: BMW27 - Compute: CPU-Only EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.0 Blend File: BMW27 - Compute: CPU-Only CentOS Stream 8 6 12 18 24 30 SE +/- 0.04, N = 3 SE +/- 0.07, N = 3 14.48 23.23
Xcompact3d Incompact3d Input: input.i3d 193 Cells Per Direction EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Seconds, Fewer Is Better Xcompact3d Incompact3d 2021-03-11 Input: input.i3d 193 Cells Per Direction CentOS Stream 8 3 6 9 12 15 SE +/- 0.16, N = 3 SE +/- 0.08, N = 3 12.19 12.07 1. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi
ACES DGEMM Sustained Floating-Point Rate EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org GFLOP/s, More Is Better ACES DGEMM 1.0 Sustained Floating-Point Rate CentOS Stream 8 10 20 30 40 50 SE +/- 0.55, N = 4 SE +/- 0.49, N = 15 45.45 44.41 1. (CC) gcc options: -O3 -march=native -fopenmp
oneDNN Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPU EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.1.2 Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPU CentOS Stream 8 0.1575 0.315 0.4725 0.63 0.7875 SE +/- 0.007446, N = 5 SE +/- 0.001114, N = 3 0.700087 0.226158 1. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
rav1e Speed: 10 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Frames Per Second, More Is Better rav1e 0.5 Speed: 10 CentOS Stream 8 2 4 6 8 10 SE +/- 0.076, N = 3 SE +/- 0.099, N = 3 5.527 7.347
Stress-NG Test: RdRand OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.13.02 Test: RdRand CentOS Stream 8 140K 280K 420K 560K 700K SE +/- 53.75, N = 3 669474.48 1. (CC) gcc options: -O2 -std=gnu99 -lm -lcrypt -lrt -lz -ldl -pthread -lc
SVT-AV1 Encoder Mode: Preset 8 - Input: Bosphorus 4K EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 0.8.7 Encoder Mode: Preset 8 - Input: Bosphorus 4K CentOS Stream 8 14 28 42 56 70 SE +/- 0.38, N = 3 SE +/- 0.81, N = 3 63.95 58.52 1. (CXX) g++ options: -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq -pie
Sockperf Test: Latency Ping Pong EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org usec, Fewer Is Better Sockperf 3.7 Test: Latency Ping Pong CentOS Stream 8 1.1765 2.353 3.5295 4.706 5.8825 SE +/- 0.019, N = 5 SE +/- 0.009, N = 5 4.236 5.229 1. (CXX) g++ options: --param -O3 -rdynamic -ldl -lpthread
Intel Open Image Denoise Run: RT.hdr_alb_nrm.3840x2160 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Images / Sec, More Is Better Intel Open Image Denoise 1.4.0 Run: RT.hdr_alb_nrm.3840x2160 CentOS Stream 8 0.6593 1.3186 1.9779 2.6372 3.2965 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 2.53 2.93
Intel Open Image Denoise Run: RT.ldr_alb_nrm.3840x2160 EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Images / Sec, More Is Better Intel Open Image Denoise 1.4.0 Run: RT.ldr_alb_nrm.3840x2160 CentOS Stream 8 0.666 1.332 1.998 2.664 3.33 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 2.54 2.96
Embree Binary: Pathtracer - Model: Asian Dragon EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Frames Per Second, More Is Better Embree 3.13 Binary: Pathtracer - Model: Asian Dragon CentOS Stream 8 20 40 60 80 100 SE +/- 0.87, N = 4 SE +/- 0.24, N = 3 72.15 83.58
GNU Octave Benchmark EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Seconds, Fewer Is Better GNU Octave Benchmark 5.2.0 CentOS Stream 8 2 4 6 8 10 SE +/- 0.013, N = 5 SE +/- 0.006, N = 5 5.641 6.697
Embree Binary: Pathtracer ISPC - Model: Asian Dragon EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Frames Per Second, More Is Better Embree 3.13 Binary: Pathtracer ISPC - Model: Asian Dragon CentOS Stream 8 20 40 60 80 100 SE +/- 0.80, N = 4 SE +/- 0.25, N = 3 71.03 108.23
OpenCV Test: DNN - Deep Neural Network OpenBenchmarking.org ms, Fewer Is Better OpenCV 4.5.4 Test: DNN - Deep Neural Network Xeon Platinum 8380 2P: CentOS Stream 8 2K 4K 6K 8K 10K SE +/- 122.38, N = 3 9101 1. (CXX) g++ options: -fPIC -fsigned-char -pthread -fomit-frame-pointer -ffunction-sections -fdata-sections -msse -msse2 -msse3 -fvisibility=hidden -O3 -shared
SVT-VP9 Tuning: PSNR/SSIM Optimized - Input: Bosphorus 1080p EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Frames Per Second, More Is Better SVT-VP9 0.3 Tuning: PSNR/SSIM Optimized - Input: Bosphorus 1080p CentOS Stream 8 100 200 300 400 500 SE +/- 4.45, N = 3 SE +/- 4.13, N = 15 424.87 465.68 1. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm
OSPray Demo: NASA Streamlines - Renderer: Path Tracer EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org FPS, More Is Better OSPray 1.8.5 Demo: NASA Streamlines - Renderer: Path Tracer CentOS Stream 8 9 18 27 36 45 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 38.46 28.57
Embree Binary: Pathtracer - Model: Crown EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Frames Per Second, More Is Better Embree 3.13 Binary: Pathtracer - Model: Crown CentOS Stream 8 20 40 60 80 100 SE +/- 0.41, N = 3 SE +/- 0.10, N = 3 110.11 63.82
DaCapo Benchmark Java Test: Jython EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org msec, Fewer Is Better DaCapo Benchmark 9.12-MR1 Java Test: Jython CentOS Stream 8 900 1800 2700 3600 4500 SE +/- 31.02, N = 4 SE +/- 15.96, N = 4 3995 3613
Embree Binary: Pathtracer ISPC - Model: Crown EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Frames Per Second, More Is Better Embree 3.13 Binary: Pathtracer ISPC - Model: Crown CentOS Stream 8 20 40 60 80 100 SE +/- 0.44, N = 3 SE +/- 0.26, N = 3 100.42 73.53
OSPray Demo: Magnetic Reconnection - Renderer: SciVis EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org FPS, More Is Better OSPray 1.8.5 Demo: Magnetic Reconnection - Renderer: SciVis CentOS Stream 8 30 60 90 120 150 SE +/- 0.53, N = 9 67.20 125.00
SVT-VP9 Tuning: VMAF Optimized - Input: Bosphorus 1080p EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Frames Per Second, More Is Better SVT-VP9 0.3 Tuning: VMAF Optimized - Input: Bosphorus 1080p CentOS Stream 8 100 200 300 400 500 SE +/- 5.40, N = 4 SE +/- 5.38, N = 4 467.87 476.33 1. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm
SVT-VP9 Tuning: Visual Quality Optimized - Input: Bosphorus 1080p EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Frames Per Second, More Is Better SVT-VP9 0.3 Tuning: Visual Quality Optimized - Input: Bosphorus 1080p CentOS Stream 8 80 160 240 320 400 SE +/- 2.18, N = 3 SE +/- 4.61, N = 3 350.67 359.89 1. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm
OSPray Demo: NASA Streamlines - Renderer: SciVis EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org FPS, More Is Better OSPray 1.8.5 Demo: NASA Streamlines - Renderer: SciVis CentOS Stream 8 40 80 120 160 200 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 166.67 142.86
SVT-HEVC Tuning: 7 - Input: Bosphorus 1080p EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Frames Per Second, More Is Better SVT-HEVC 1.5.0 Tuning: 7 - Input: Bosphorus 1080p CentOS Stream 8 70 140 210 280 350 SE +/- 0.88, N = 3 SE +/- 1.81, N = 3 329.56 299.17 1. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -lpthread -lrt
oneDNN Harness: Deconvolution Batch shapes_3d - Data Type: u8s8f32 - Engine: CPU EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org ms, Fewer Is Better oneDNN 2.1.2 Harness: Deconvolution Batch shapes_3d - Data Type: u8s8f32 - Engine: CPU CentOS Stream 8 0.1218 0.2436 0.3654 0.4872 0.609 SE +/- 0.002376, N = 3 SE +/- 0.000855, N = 3 0.541402 0.189266 1. (CXX) g++ options: -O3 -march=native -std=c++11 -fopenmp -msse4.1 -fPIC -pie -lpthread -ldl
Primesieve 1e12 Prime Number Generation EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Seconds, Fewer Is Better Primesieve 7.7 1e12 Prime Number Generation CentOS Stream 8 0.7871 1.5742 2.3613 3.1484 3.9355 SE +/- 0.014, N = 3 SE +/- 0.015, N = 3 2.449 3.498 1. (CXX) g++ options: -O3 -lpthread
SVT-HEVC Tuning: 10 - Input: Bosphorus 1080p EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org Frames Per Second, More Is Better SVT-HEVC 1.5.0 Tuning: 10 - Input: Bosphorus 1080p CentOS Stream 8 130 260 390 520 650 SE +/- 2.53, N = 3 SE +/- 2.87, N = 3 591.35 574.56 1. (CC) gcc options: -fPIE -fPIC -O3 -O2 -pie -rdynamic -lpthread -lrt
OSPray Demo: Magnetic Reconnection - Renderer: Path Tracer EPYC 7763 2P Xeon Platinum 8380 2P OpenBenchmarking.org FPS, More Is Better OSPray 1.8.5 Demo: Magnetic Reconnection - Renderer: Path Tracer CentOS Stream 8 110 220 330 440 550 500 500
Phoronix Test Suite v10.8.5