Benchmarks by Michael Larabel for a future article.
v6.4 Processor: 2 x Intel Xeon Max 9480 @ 3.50GHz (112 Cores / 224 Threads), Motherboard: Supermicro X13DEM v1.10 (1.3 BIOS), Chipset: Intel Device 1bce, Memory: 512GB, Disk: 2 x 7682GB INTEL SSDPF2KX076TZ, Graphics: ASPEED, Monitor: VE228, Network: 2 x Broadcom BCM57508 NetXtreme-E 10Gb/25Gb/40Gb/50Gb/100Gb/200Gb
OS: Ubuntu 23.04, Kernel: 6.4.0-060400-generic (x86_64), Desktop: GNOME Shell 44.0, Display Server: X Server 1.21.1.7, Compiler: GCC 12.2.0, File-System: ext4, Screen Resolution: 1920x1080
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-12-Pa930Z/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-Pa930Z/gcc-12-12.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_cpufreq performance - CPU Microcode: 0x2c0001d1Java Notes: OpenJDK Runtime Environment (build 17.0.6+10-Ubuntu-1ubuntu2)Python Notes: Python 3.11.2Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
v6.5 Processor: 2 x Intel Xeon Max 9480 @ 3.50GHz (112 Cores / 224 Threads), Motherboard: Supermicro X13DEM v1.10 (1.3 BIOS), Chipset: Intel Device 1bce, Memory: 512GB, Disk: 2 x 7682GB INTEL SSDPF2KX076TZ, Graphics: ASPEED, Network: 2 x Broadcom BCM57508 NetXtreme-E 10Gb/25Gb/40Gb/50Gb/100Gb/200Gb
OS: Ubuntu 23.04, Kernel: 6.5.0-060500-generic (x86_64), Desktop: GNOME Shell 44.0, Display Server: X Server 1.21.1.7, Compiler: GCC 12.2.0, File-System: ext4, Screen Resolution: 1024x768
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-12-Pa930Z/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-Pa930Z/gcc-12-12.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_cpufreq performance - CPU Microcode: 0x2c0001d1Java Notes: OpenJDK Runtime Environment (build 17.0.6+10-Ubuntu-1ubuntu2)Python Notes: Python 3.11.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
v6.6 6 Oct OS: Ubuntu 23.04, Kernel: 6.6.0-060600rc4daily20231006-generic (x86_64), Desktop: GNOME Shell 44.0, Display Server: X Server 1.21.1.7, Compiler: GCC 12.2.0, File-System: ext4, Screen Resolution: 1920x1200
Blender OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.6 Blend File: Barbershop - Compute: CPU-Only v6.4 v6.5 v6.6 6 Oct 80 160 240 320 400 SE +/- 4.74, N = 9 SE +/- 14.36, N = 9 SE +/- 11.89, N = 9 315.02 336.42 385.35
PostgreSQL This is a benchmark of PostgreSQL using the integrated pgbench for facilitating the database benchmarks. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org ms, Fewer Is Better PostgreSQL 16 Scaling Factor: 1000 - Clients: 1000 - Mode: Read Write - Average Latency v6.4 v6.5 v6.6 6 Oct 11 22 33 44 55 SE +/- 0.55, N = 12 SE +/- 0.46, N = 12 SE +/- 0.55, N = 12 50.36 50.79 50.79 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.org TPS, More Is Better PostgreSQL 16 Scaling Factor: 1000 - Clients: 1000 - Mode: Read Write v6.4 v6.5 v6.6 6 Oct 4K 8K 12K 16K 20K SE +/- 218.15, N = 12 SE +/- 180.45, N = 12 SE +/- 216.13, N = 12 19884 19709 19714 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
TensorFlow This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.12 Device: CPU - Batch Size: 256 - Model: ResNet-50 v6.4 v6.5 v6.6 6 Oct 15 30 45 60 75 SE +/- 0.29, N = 3 SE +/- 0.76, N = 3 SE +/- 0.66, N = 9 65.22 65.91 65.97
High Performance Conjugate Gradient HPCG is the High Performance Conjugate Gradient and is a new scientific benchmark from Sandia National Lans focused for super-computer testing with modern real-world workloads compared to HPCC. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org GFLOP/s, More Is Better High Performance Conjugate Gradient 3.1 X Y Z: 144 144 144 - RT: 60 v6.4 v6.5 v6.6 6 Oct 20 40 60 80 100 SE +/- 0.76, N = 6 SE +/- 0.98, N = 9 SE +/- 0.42, N = 3 76.58 73.73 74.27 1. (CXX) g++ options: -O3 -ffast-math -ftree-vectorize -lmpi_cxx -lmpi
PostgreSQL This is a benchmark of PostgreSQL using the integrated pgbench for facilitating the database benchmarks. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org ms, Fewer Is Better PostgreSQL 16 Scaling Factor: 1000 - Clients: 1000 - Mode: Read Only - Average Latency v6.4 v6.5 v6.6 6 Oct 0.3285 0.657 0.9855 1.314 1.6425 SE +/- 0.018, N = 12 SE +/- 0.021, N = 12 SE +/- 0.018, N = 3 1.411 1.403 1.460 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.org TPS, More Is Better PostgreSQL 16 Scaling Factor: 1000 - Clients: 1000 - Mode: Read Only v6.4 v6.5 v6.6 6 Oct 150K 300K 450K 600K 750K SE +/- 8915.31, N = 12 SE +/- 10655.35, N = 12 SE +/- 8415.01, N = 3 710079 714484 685422 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.org ms, Fewer Is Better PostgreSQL 16 Scaling Factor: 1000 - Clients: 800 - Mode: Read Write - Average Latency v6.4 v6.5 v6.6 6 Oct 9 18 27 36 45 SE +/- 0.47, N = 12 SE +/- 0.42, N = 3 SE +/- 0.33, N = 12 39.25 37.45 39.04 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.org TPS, More Is Better PostgreSQL 16 Scaling Factor: 1000 - Clients: 800 - Mode: Read Write v6.4 v6.5 v6.6 6 Oct 5K 10K 15K 20K 25K SE +/- 240.76, N = 12 SE +/- 236.09, N = 3 SE +/- 172.49, N = 12 20416 21367 20510 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
Apache IoTDB Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Average Latency, Fewer Is Better Apache IoTDB 1.2 Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 100 v6.4 v6.5 v6.6 6 Oct 20 40 60 80 100 SE +/- 0.85, N = 15 SE +/- 0.97, N = 15 SE +/- 1.09, N = 13 77.00 76.54 75.47 MAX: 12622.02 MAX: 12675.23 MAX: 13909.62
OpenBenchmarking.org point/sec, More Is Better Apache IoTDB 1.2 Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 100 v6.4 v6.5 v6.6 6 Oct 13M 26M 39M 52M 65M SE +/- 581785.82, N = 15 SE +/- 676191.00, N = 15 SE +/- 697211.54, N = 13 59376956 59679878 60179852
Apache IoTDB Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Average Latency, Fewer Is Better Apache IoTDB 1.2 Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 400 v6.4 v6.5 v6.6 6 Oct 60 120 180 240 300 SE +/- 2.23, N = 12 SE +/- 2.65, N = 12 SE +/- 4.51, N = 3 277.40 273.92 277.50 MAX: 30379.43 MAX: 28783.48 MAX: 28355.83
OpenBenchmarking.org point/sec, More Is Better Apache IoTDB 1.2 Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 400 v6.4 v6.5 v6.6 6 Oct 14M 28M 42M 56M 70M SE +/- 524197.72, N = 12 SE +/- 628672.36, N = 12 SE +/- 718005.55, N = 3 65784478 66078835 65998468
OpenBenchmarking.org Average Latency, Fewer Is Better Apache IoTDB 1.2 Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 200 - Client Number: 400 v6.4 v6.5 v6.6 6 Oct 30 60 90 120 150 SE +/- 1.11, N = 15 SE +/- 1.66, N = 15 SE +/- 2.18, N = 15 140.40 139.82 134.57 MAX: 27765.12 MAX: 27644.97 MAX: 27260.07
OpenBenchmarking.org point/sec, More Is Better Apache IoTDB 1.2 Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 200 - Client Number: 400 v6.4 v6.5 v6.6 6 Oct 11M 22M 33M 44M 55M SE +/- 396843.42, N = 15 SE +/- 453048.86, N = 15 SE +/- 462612.67, N = 15 50423542 51104786 51764919
AOM AV1 This is a test of the AOMedia AV1 encoder (libaom) developed by AOMedia and Google as the AV1 Codec Library. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Frames Per Second, More Is Better AOM AV1 3.7 Encoder Mode: Speed 4 Two-Pass - Input: Bosphorus 4K v6.4 v6.5 v6.6 6 Oct 2 4 6 8 10 SE +/- 0.10, N = 15 SE +/- 0.24, N = 12 SE +/- 0.16, N = 15 7.07 7.72 7.13 1. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
Apache IoTDB Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Average Latency, Fewer Is Better Apache IoTDB 1.2 Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 200 - Client Number: 100 v6.4 v6.5 v6.6 6 Oct 8 16 24 32 40 SE +/- 0.38, N = 15 SE +/- 0.74, N = 15 SE +/- 0.42, N = 14 35.31 34.49 35.64 MAX: 23938.81 MAX: 23957.29 MAX: 23906.01
OpenBenchmarking.org point/sec, More Is Better Apache IoTDB 1.2 Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 200 - Client Number: 100 v6.4 v6.5 v6.6 6 Oct 11M 22M 33M 44M 55M SE +/- 482700.73, N = 15 SE +/- 518637.11, N = 15 SE +/- 501437.42, N = 14 51248299 51188398 50701925
PostgreSQL This is a benchmark of PostgreSQL using the integrated pgbench for facilitating the database benchmarks. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org ms, Fewer Is Better PostgreSQL 16 Scaling Factor: 100 - Clients: 1000 - Mode: Read Only - Average Latency v6.4 v6.5 v6.6 6 Oct 0.3533 0.7066 1.0599 1.4132 1.7665 SE +/- 0.042, N = 9 SE +/- 0.056, N = 12 SE +/- 0.044, N = 9 1.570 1.454 1.446 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.org TPS, More Is Better PostgreSQL 16 Scaling Factor: 100 - Clients: 1000 - Mode: Read Only v6.4 v6.5 v6.6 6 Oct 150K 300K 450K 600K 750K SE +/- 15922.26, N = 9 SE +/- 30469.92, N = 12 SE +/- 23263.68, N = 9 640235 700836 697103 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
Apache IoTDB Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Average Latency, Fewer Is Better Apache IoTDB 1.2 Device Count: 200 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 100 v6.4 v6.5 v6.6 6 Oct 30 60 90 120 150 SE +/- 1.53, N = 13 SE +/- 1.56, N = 15 SE +/- 2.12, N = 12 125.95 127.67 126.51 MAX: 23943.82 MAX: 23994.5 MAX: 23988.02
OpenBenchmarking.org point/sec, More Is Better Apache IoTDB 1.2 Device Count: 200 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 100 v6.4 v6.5 v6.6 6 Oct 11M 22M 33M 44M 55M SE +/- 506247.74, N = 13 SE +/- 410488.78, N = 15 SE +/- 547035.26, N = 12 52514008 52519898 52700021
PostgreSQL This is a benchmark of PostgreSQL using the integrated pgbench for facilitating the database benchmarks. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org ms, Fewer Is Better PostgreSQL 16 Scaling Factor: 100 - Clients: 1000 - Mode: Read Write - Average Latency v6.4 v6.5 v6.6 6 Oct 6 12 18 24 30 SE +/- 0.29, N = 4 SE +/- 0.21, N = 12 SE +/- 0.25, N = 12 23.30 23.64 23.86 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.org TPS, More Is Better PostgreSQL 16 Scaling Factor: 100 - Clients: 1000 - Mode: Read Write v6.4 v6.5 v6.6 6 Oct 9K 18K 27K 36K 45K SE +/- 535.00, N = 4 SE +/- 379.56, N = 12 SE +/- 434.91, N = 12 42947 42340 41961 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
Apache IoTDB Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Average Latency, Fewer Is Better Apache IoTDB 1.2 Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 200 - Client Number: 100 v6.4 v6.5 v6.6 6 Oct 10 20 30 40 50 SE +/- 0.46, N = 15 SE +/- 0.39, N = 15 SE +/- 0.42, N = 15 42.68 42.19 41.71 MAX: 15117.6 MAX: 16359.93 MAX: 18608.41
OpenBenchmarking.org point/sec, More Is Better Apache IoTDB 1.2 Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 200 - Client Number: 100 v6.4 v6.5 v6.6 6 Oct 9M 18M 27M 36M 45M SE +/- 387575.50, N = 15 SE +/- 308400.53, N = 15 SE +/- 384183.23, N = 15 41675265 42150368 42560026
PostgreSQL This is a benchmark of PostgreSQL using the integrated pgbench for facilitating the database benchmarks. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org ms, Fewer Is Better PostgreSQL 16 Scaling Factor: 1000 - Clients: 800 - Mode: Read Only - Average Latency v6.4 v6.5 v6.6 6 Oct 0.2567 0.5134 0.7701 1.0268 1.2835 SE +/- 0.016, N = 3 SE +/- 0.015, N = 3 SE +/- 0.015, N = 12 1.141 1.098 1.128 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.org TPS, More Is Better PostgreSQL 16 Scaling Factor: 1000 - Clients: 800 - Mode: Read Only v6.4 v6.5 v6.6 6 Oct 160K 320K 480K 640K 800K SE +/- 9464.82, N = 3 SE +/- 10180.90, N = 3 SE +/- 9607.07, N = 12 701213 729184 710389 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
VVenC VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Frames Per Second, More Is Better VVenC 1.9 Video Input: Bosphorus 4K - Video Preset: Faster v6.4 v6.5 v6.6 6 Oct 2 4 6 8 10 SE +/- 0.097, N = 15 SE +/- 0.137, N = 15 SE +/- 0.103, N = 15 8.450 8.500 8.326 1. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects
Apache IoTDB Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Average Latency, Fewer Is Better Apache IoTDB 1.2 Device Count: 100 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 100 v6.4 v6.5 v6.6 6 Oct 30 60 90 120 150 SE +/- 1.48, N = 15 SE +/- 1.82, N = 15 SE +/- 1.48, N = 15 155.22 156.06 156.10 MAX: 27068.23 MAX: 26765.03 MAX: 27008.49
OpenBenchmarking.org point/sec, More Is Better Apache IoTDB 1.2 Device Count: 100 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 100 v6.4 v6.5 v6.6 6 Oct 8M 16M 24M 32M 40M SE +/- 432863.04, N = 15 SE +/- 443563.55, N = 15 SE +/- 369838.39, N = 15 39122053 38539849 39142134
OpenBenchmarking.org Average Latency, Fewer Is Better Apache IoTDB 1.2 Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 400 v6.4 v6.5 v6.6 6 Oct 60 120 180 240 300 SE +/- 3.08, N = 10 SE +/- 3.68, N = 12 SE +/- 0.59, N = 3 265.99 266.74 279.15 MAX: 30115.56 MAX: 27647.27 MAX: 29007.59
OpenBenchmarking.org point/sec, More Is Better Apache IoTDB 1.2 Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 400 v6.4 v6.5 v6.6 6 Oct 13M 26M 39M 52M 65M SE +/- 450916.16, N = 10 SE +/- 654921.10, N = 12 SE +/- 455210.13, N = 3 58096600 59552847 56839448
OpenBenchmarking.org Average Latency, Fewer Is Better Apache IoTDB 1.2 Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 200 - Client Number: 400 v6.4 v6.5 v6.6 6 Oct 40 80 120 160 200 SE +/- 2.09, N = 3 SE +/- 1.65, N = 15 SE +/- 1.95, N = 15 163.36 152.95 153.49 MAX: 27171.99 MAX: 28282.63 MAX: 27353.32
OpenBenchmarking.org point/sec, More Is Better Apache IoTDB 1.2 Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 200 - Client Number: 400 v6.4 v6.5 v6.6 6 Oct 9M 18M 27M 36M 45M SE +/- 245644.87, N = 3 SE +/- 373729.23, N = 15 SE +/- 344346.26, N = 15 40520901 42201696 41638681
AOM AV1 This is a test of the AOMedia AV1 encoder (libaom) developed by AOMedia and Google as the AV1 Codec Library. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Frames Per Second, More Is Better AOM AV1 3.7 Encoder Mode: Speed 6 Two-Pass - Input: Bosphorus 4K v6.4 v6.5 v6.6 6 Oct 4 8 12 16 20 SE +/- 0.38, N = 15 SE +/- 0.42, N = 15 SE +/- 0.56, N = 15 14.30 14.09 12.38 1. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
OpenVINO OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2023.1 Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPU v6.4 v6.5 v6.6 6 Oct 0.0765 0.153 0.2295 0.306 0.3825 SE +/- 0.00, N = 15 SE +/- 0.00, N = 15 SE +/- 0.00, N = 15 0.34 0.34 0.33 MIN: 0.25 / MAX: 36.53 MIN: 0.24 / MAX: 33.93 MIN: 0.24 / MAX: 31.45 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org FPS, More Is Better OpenVINO 2023.1 Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPU v6.4 v6.5 v6.6 6 Oct 30K 60K 90K 120K 150K SE +/- 1041.05, N = 15 SE +/- 1061.43, N = 15 SE +/- 837.50, N = 15 112475.34 111886.46 117557.36 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
Apache IoTDB Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Average Latency, Fewer Is Better Apache IoTDB 1.2 Device Count: 200 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 100 v6.4 v6.5 v6.6 6 Oct 20 40 60 80 100 SE +/- 0.96, N = 15 SE +/- 1.35, N = 15 SE +/- 1.57, N = 3 97.16 99.71 102.21 MAX: 24088.47 MAX: 24866.21 MAX: 23899.32
OpenBenchmarking.org point/sec, More Is Better Apache IoTDB 1.2 Device Count: 200 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 100 v6.4 v6.5 v6.6 6 Oct 9M 18M 27M 36M 45M SE +/- 358083.32, N = 15 SE +/- 472609.51, N = 15 SE +/- 465340.61, N = 3 43036898 42147671 41245153
TensorFlow This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.12 Device: CPU - Batch Size: 64 - Model: ResNet-50 v6.4 v6.5 v6.6 6 Oct 12 24 36 48 60 SE +/- 0.72, N = 3 SE +/- 0.49, N = 12 SE +/- 0.57, N = 3 53.04 51.42 51.40
Apache IoTDB Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Average Latency, Fewer Is Better Apache IoTDB 1.2 Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 100 v6.4 v6.5 v6.6 6 Oct 20 40 60 80 100 SE +/- 0.13, N = 3 SE +/- 1.12, N = 3 SE +/- 1.01, N = 3 104.93 108.51 102.90 MAX: 23900.89 MAX: 23916.24 MAX: 23899.89
OpenBenchmarking.org point/sec, More Is Better Apache IoTDB 1.2 Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 100 v6.4 v6.5 v6.6 6 Oct 16M 32M 48M 64M 80M SE +/- 265706.48, N = 3 SE +/- 551372.39, N = 3 SE +/- 658836.89, N = 3 72605513 70592527 73920159
OpenBenchmarking.org Average Latency, Fewer Is Better Apache IoTDB 1.2 Device Count: 200 - Batch Size Per Write: 100 - Sensor Count: 200 - Client Number: 100 v6.4 v6.5 v6.6 6 Oct 16 32 48 64 80 SE +/- 0.55, N = 12 SE +/- 0.56, N = 15 SE +/- 0.49, N = 12 70.77 69.65 69.84 MAX: 23941.04 MAX: 23930.93 MAX: 23963.52
OpenBenchmarking.org point/sec, More Is Better Apache IoTDB 1.2 Device Count: 200 - Batch Size Per Write: 100 - Sensor Count: 200 - Client Number: 100 v6.4 v6.5 v6.6 6 Oct 5M 10M 15M 20M 25M SE +/- 171233.24, N = 12 SE +/- 188478.96, N = 15 SE +/- 173884.88, N = 12 24582941 24825272 24839001
AOM AV1 This is a test of the AOMedia AV1 encoder (libaom) developed by AOMedia and Google as the AV1 Codec Library. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Frames Per Second, More Is Better AOM AV1 3.7 Encoder Mode: Speed 0 Two-Pass - Input: Bosphorus 4K v6.4 v6.5 v6.6 6 Oct 0.0878 0.1756 0.2634 0.3512 0.439 SE +/- 0.00, N = 15 SE +/- 0.00, N = 15 SE +/- 0.01, N = 15 0.39 0.39 0.37 1. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
Apache IoTDB Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org point/sec, More Is Better Apache IoTDB 1.2 Device Count: 100 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 100 v6.4 v6.5 v6.6 6 Oct 7M 14M 21M 28M 35M SE +/- 332901.55, N = 15 SE +/- 232925.17, N = 15 SE +/- 327309.26, N = 5 30189985 30632905 29931254
OpenBenchmarking.org Ops per sec, More Is Better Apache Hadoop 3.3.6 Operation: Open - Threads: 500 - Files: 100000 v6.4 v6.5 v6.6 6 Oct 60K 120K 180K 240K 300K SE +/- 9141.48, N = 12 SE +/- 20767.40, N = 15 SE +/- 8486.55, N = 15 265652 254640 267299
VVenC VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Frames Per Second, More Is Better VVenC 1.9 Video Input: Bosphorus 4K - Video Preset: Fast v6.4 v6.5 v6.6 6 Oct 1.1644 2.3288 3.4932 4.6576 5.822 SE +/- 0.058, N = 3 SE +/- 0.066, N = 3 SE +/- 0.064, N = 12 5.075 5.175 5.076 1. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects
Redis 7.0.12 + memtier_benchmark Memtier_benchmark is a NoSQL Redis/Memcache traffic generation plus benchmarking tool developed by Redis Labs. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Ops/sec, More Is Better Redis 7.0.12 + memtier_benchmark 2.0 Protocol: Redis - Clients: 50 - Set To Get Ratio: 1:5 v6.4 v6.5 v6.6 6 Oct 500K 1000K 1500K 2000K 2500K SE +/- 57616.61, N = 12 SE +/- 31237.79, N = 15 SE +/- 21494.50, N = 3 2331936.55 2128295.60 2188365.95 1. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre
Apache IoTDB Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Average Latency, Fewer Is Better Apache IoTDB 1.2 Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 400 v6.4 v6.5 v6.6 6 Oct 80 160 240 320 400 SE +/- 4.26, N = 3 SE +/- 5.94, N = 4 SE +/- 4.55, N = 3 370.82 368.15 374.60 MAX: 30958.08 MAX: 30924.16 MAX: 29403.16
OpenBenchmarking.org point/sec, More Is Better Apache IoTDB 1.2 Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 400 v6.4 v6.5 v6.6 6 Oct 14M 28M 42M 56M 70M SE +/- 525580.37, N = 3 SE +/- 800541.34, N = 4 SE +/- 791635.87, N = 3 66098942 67394323 66372800
OpenBenchmarking.org Average Latency, Fewer Is Better Apache IoTDB 1.2 Device Count: 100 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 100 v6.5 v6.6 6 Oct 30 60 90 120 150 SE +/- 1.28, N = 15 SE +/- 1.88, N = 5 127.92 132.61 MAX: 26759.03 MAX: 25773.81
Apache IoTDB Apache IotDB is a time series database and this benchmark is facilitated using the IoT Benchmaark [https://github.com/thulab/iot-benchmark/]. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Average Latency, Fewer Is Better Apache IoTDB 1.2 Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 100 v6.4 v6.5 v6.6 6 Oct 30 60 90 120 150 SE +/- 0.10, N = 3 SE +/- 1.29, N = 3 SE +/- 1.13, N = 3 111.75 111.14 111.92 MAX: 10157.01 MAX: 11352.98 MAX: 10281.65
OpenBenchmarking.org point/sec, More Is Better Apache IoTDB 1.2 Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 800 - Client Number: 100 v6.4 v6.5 v6.6 6 Oct 14M 28M 42M 56M 70M SE +/- 46596.93, N = 3 SE +/- 691489.65, N = 3 SE +/- 494444.55, N = 3 67045377 67023750 66764218
OpenBenchmarking.org Average Latency, Fewer Is Better Apache IoTDB 1.2 Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 100 v6.4 v6.5 v6.6 6 Oct 16 32 48 64 80 SE +/- 0.22, N = 3 SE +/- 1.10, N = 3 SE +/- 0.62, N = 3 69.88 70.09 69.65 MAX: 23895.2 MAX: 23936.83 MAX: 23900.31
OpenBenchmarking.org point/sec, More Is Better Apache IoTDB 1.2 Device Count: 800 - Batch Size Per Write: 100 - Sensor Count: 500 - Client Number: 100 v6.4 v6.5 v6.6 6 Oct 14M 28M 42M 56M 70M SE +/- 690519.22, N = 3 SE +/- 929043.22, N = 3 SE +/- 407522.59, N = 3 66287906 66558906 67263606
Rodinia Rodinia is a suite focused upon accelerating compute-intensive applications with accelerators. CUDA, OpenMP, and OpenCL parallel models are supported by the included applications. This profile utilizes select OpenCL, NVIDIA CUDA and OpenMP test binaries at the moment. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better Rodinia 3.1 Test: OpenMP HotSpot3D v6.4 v6.5 v6.6 6 Oct 20 40 60 80 100 SE +/- 0.21, N = 3 SE +/- 0.17, N = 3 SE +/- 1.06, N = 15 80.64 80.78 83.83 1. (CXX) g++ options: -O2 -lOpenCL
OpenVINO OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2023.1 Model: Machine Translation EN To DE FP16 - Device: CPU v6.4 v6.5 v6.6 6 Oct 13 26 39 52 65 SE +/- 0.43, N = 3 SE +/- 0.65, N = 15 SE +/- 0.54, N = 6 55.31 57.84 56.60 MIN: 38.54 / MAX: 312.66 MIN: 34.52 / MAX: 409.65 MIN: 39.06 / MAX: 243.12 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org FPS, More Is Better OpenVINO 2023.1 Model: Machine Translation EN To DE FP16 - Device: CPU v6.4 v6.5 v6.6 6 Oct 140 280 420 560 700 SE +/- 5.12, N = 3 SE +/- 7.36, N = 15 SE +/- 6.10, N = 6 667.64 639.08 652.77 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
Rodinia Rodinia is a suite focused upon accelerating compute-intensive applications with accelerators. CUDA, OpenMP, and OpenCL parallel models are supported by the included applications. This profile utilizes select OpenCL, NVIDIA CUDA and OpenMP test binaries at the moment. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better Rodinia 3.1 Test: OpenMP Streamcluster v6.4 v6.5 v6.6 6 Oct 11 22 33 44 55 SE +/- 2.00, N = 15 SE +/- 0.58, N = 3 SE +/- 2.47, N = 15 42.61 45.67 47.37 1. (CXX) g++ options: -O2 -lOpenCL
OpenRadioss OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better OpenRadioss 2023.09.15 Model: Bird Strike on Windshield v6.4 v6.5 v6.6 6 Oct 40 80 120 160 200 SE +/- 0.82, N = 3 SE +/- 0.28, N = 3 SE +/- 0.50, N = 3 159.21 159.91 161.95
TensorFlow This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.12 Device: CPU - Batch Size: 32 - Model: ResNet-50 v6.4 v6.5 v6.6 6 Oct 11 22 33 44 55 SE +/- 0.54, N = 3 SE +/- 0.58, N = 12 SE +/- 0.66, N = 3 45.20 46.41 46.63
OpenRadioss OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better OpenRadioss 2023.09.15 Model: Chrysler Neon 1M v6.4 v6.5 v6.6 6 Oct 30 60 90 120 150 SE +/- 0.01, N = 3 SE +/- 1.07, N = 3 SE +/- 0.37, N = 3 128.50 130.43 129.79
Redis 7.0.12 + memtier_benchmark Memtier_benchmark is a NoSQL Redis/Memcache traffic generation plus benchmarking tool developed by Redis Labs. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Ops/sec, More Is Better Redis 7.0.12 + memtier_benchmark 2.0 Protocol: Redis - Clients: 100 - Set To Get Ratio: 1:10 v6.4 v6.5 v6.6 6 Oct 500K 1000K 1500K 2000K 2500K SE +/- 23334.59, N = 3 SE +/- 37051.26, N = 12 SE +/- 26927.97, N = 3 2262895.86 2068343.10 2231964.35 1. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre
Stress-NG Stress-NG is a Linux stress tool developed by Colin Ian King. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: Pipe v6.4 v6.5 v6.6 6 Oct 8M 16M 24M 32M 40M SE +/- 487787.75, N = 15 SE +/- 1055793.62, N = 15 SE +/- 581009.73, N = 15 31203885.98 38337996.37 27024816.26 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: Glibc C String Functions v6.4 v6.5 v6.6 6 Oct 20M 40M 60M 80M 100M SE +/- 906580.48, N = 15 SE +/- 1000931.10, N = 15 SE +/- 772766.53, N = 15 81547074.59 77642781.17 83443787.58 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: Mutex v6.4 v6.5 v6.6 6 Oct 6M 12M 18M 24M 30M SE +/- 381162.66, N = 15 SE +/- 376350.69, N = 15 SE +/- 709062.89, N = 15 23140484.14 25266270.16 27388113.63 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
TensorFlow This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.12 Device: CPU - Batch Size: 16 - Model: ResNet-50 v6.4 v6.5 v6.6 6 Oct 9 18 27 36 45 SE +/- 0.35, N = 3 SE +/- 0.31, N = 9 SE +/- 0.38, N = 15 36.85 38.22 38.01
PostgreSQL This is a benchmark of PostgreSQL using the integrated pgbench for facilitating the database benchmarks. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org ms, Fewer Is Better PostgreSQL 16 Scaling Factor: 100 - Clients: 800 - Mode: Read Write - Average Latency v6.4 v6.5 v6.6 6 Oct 4 8 12 16 20 SE +/- 0.22, N = 3 SE +/- 0.15, N = 3 SE +/- 0.21, N = 4 17.56 18.28 17.98 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.org TPS, More Is Better PostgreSQL 16 Scaling Factor: 100 - Clients: 800 - Mode: Read Write v6.4 v6.5 v6.6 6 Oct 10K 20K 30K 40K 50K SE +/- 557.77, N = 3 SE +/- 370.85, N = 3 SE +/- 508.74, N = 4 45581 43765 44504 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
SVT-AV1 OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 1.7 Encoder Mode: Preset 4 - Input: Bosphorus 4K v6.4 v6.5 v6.6 6 Oct 0.9297 1.8594 2.7891 3.7188 4.6485 SE +/- 0.031, N = 15 SE +/- 0.040, N = 6 SE +/- 0.030, N = 11 4.132 4.016 4.120 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenRadioss OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better OpenRadioss 2023.09.15 Model: INIVOL and Fluid Structure Interaction Drop Container v6.4 v6.5 v6.6 6 Oct 30 60 90 120 150 SE +/- 0.42, N = 3 SE +/- 0.29, N = 3 SE +/- 0.74, N = 3 138.97 141.85 138.78
OpenVINO OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2023.1 Model: Weld Porosity Detection FP16-INT8 - Device: CPU v6.4 v6.5 v6.6 6 Oct 0.9518 1.9036 2.8554 3.8072 4.759 SE +/- 0.03, N = 15 SE +/- 0.03, N = 3 SE +/- 0.03, N = 3 4.23 3.70 3.67 MIN: 2.61 / MAX: 111.71 MIN: 2.57 / MAX: 108.71 MIN: 2.51 / MAX: 63.65 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org FPS, More Is Better OpenVINO 2023.1 Model: Weld Porosity Detection FP16-INT8 - Device: CPU v6.4 v6.5 v6.6 6 Oct 6K 12K 18K 24K 30K SE +/- 231.74, N = 15 SE +/- 236.24, N = 3 SE +/- 186.80, N = 3 25464.62 29736.31 29779.38 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2023.1 Model: Age Gender Recognition Retail 0013 FP16 - Device: CPU v6.4 v6.5 v6.6 6 Oct 0.1283 0.2566 0.3849 0.5132 0.6415 SE +/- 0.01, N = 15 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 0.57 0.56 0.53 MIN: 0.29 / MAX: 74.65 MIN: 0.36 / MAX: 51.75 MIN: 0.37 / MAX: 56.75 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org FPS, More Is Better OpenVINO 2023.1 Model: Age Gender Recognition Retail 0013 FP16 - Device: CPU v6.4 v6.5 v6.6 6 Oct 20K 40K 60K 80K 100K SE +/- 1782.53, N = 15 SE +/- 544.05, N = 3 SE +/- 537.27, N = 3 74954.03 94168.28 93693.64 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
Stress-NG Stress-NG is a Linux stress tool developed by Colin Ian King. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: Matrix Math v6.4 v6.5 v6.6 6 Oct 80K 160K 240K 320K 400K SE +/- 15424.87, N = 12 SE +/- 5325.22, N = 12 SE +/- 6612.89, N = 15 396715.18 349646.89 370932.98 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
High Performance Conjugate Gradient HPCG is the High Performance Conjugate Gradient and is a new scientific benchmark from Sandia National Lans focused for super-computer testing with modern real-world workloads compared to HPCC. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org GFLOP/s, More Is Better High Performance Conjugate Gradient 3.1 X Y Z: 104 104 104 - RT: 60 v6.4 v6.5 v6.6 6 Oct 20 40 60 80 100 SE +/- 0.65, N = 3 SE +/- 0.94, N = 3 SE +/- 0.92, N = 3 104.68 102.19 101.74 1. (CXX) g++ options: -O3 -ffast-math -ftree-vectorize -lmpi_cxx -lmpi
Stress-NG Stress-NG is a Linux stress tool developed by Colin Ian King. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: IO_uring v6.4 v6.5 v6.6 6 Oct 700K 1400K 2100K 2800K 3500K SE +/- 33896.79, N = 15 SE +/- 24671.90, N = 15 SE +/- 44896.75, N = 3 2619140.96 2557131.35 3424829.35 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
OpenRadioss OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better OpenRadioss 2023.09.15 Model: Rubber O-Ring Seal Installation v6.4 v6.5 v6.6 6 Oct 30 60 90 120 150 SE +/- 0.71, N = 3 SE +/- 0.71, N = 3 SE +/- 0.19, N = 3 116.35 119.47 116.46
OpenFOAM OpenFOAM is the leading free, open-source software for computational fluid dynamics (CFD). This test profile currently uses the drivaerFastback test case for analyzing automotive aerodynamics or alternatively the older motorBike input. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better OpenFOAM 10 Input: drivaerFastback, Medium Mesh Size - Execution Time v6.4 v6.5 v6.6 6 Oct 50 100 150 200 250 195.89 197.95 206.43 1. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfiniteVolume -lmeshTools -lparallel -llagrangian -lregionModels -lgenericPatchFields -lOpenFOAM -ldl -lm
OpenBenchmarking.org Seconds, Fewer Is Better OpenFOAM 10 Input: drivaerFastback, Medium Mesh Size - Mesh Time v6.4 v6.5 v6.6 6 Oct 40 80 120 160 200 175.93 174.04 175.39 1. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfiniteVolume -lmeshTools -lparallel -llagrangian -lregionModels -lgenericPatchFields -lOpenFOAM -ldl -lm
Stress-NG Stress-NG is a Linux stress tool developed by Colin Ian King. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: Function Call v6.4 v6.5 v6.6 6 Oct 14K 28K 42K 56K 70K SE +/- 811.29, N = 15 SE +/- 545.69, N = 15 SE +/- 112.81, N = 3 63249.99 61230.89 65566.39 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
Blender OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.6 Blend File: Fishy Cat - Compute: CPU-Only v6.4 v6.5 v6.6 6 Oct 9 18 27 36 45 SE +/- 0.15, N = 3 SE +/- 0.41, N = 15 SE +/- 0.10, N = 3 40.12 41.62 40.42
OpenRadioss OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better OpenRadioss 2023.09.15 Model: Bumper Beam v6.4 v6.5 v6.6 6 Oct 20 40 60 80 100 SE +/- 0.64, N = 3 SE +/- 0.23, N = 3 SE +/- 0.37, N = 3 95.17 96.05 95.08
Stress-NG Stress-NG is a Linux stress tool developed by Colin Ian King. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: Floating Point v6.4 v6.5 v6.6 6 Oct 7K 14K 21K 28K 35K SE +/- 348.48, N = 15 SE +/- 305.43, N = 3 SE +/- 238.47, N = 10 31253.93 29342.67 31541.66 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: Memory Copying v6.4 v6.5 v6.6 6 Oct 6K 12K 18K 24K 30K SE +/- 629.40, N = 15 SE +/- 228.57, N = 3 SE +/- 224.34, N = 7 25816.50 25493.26 25800.06 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
Blender OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.6 Blend File: Pabellon Barcelona - Compute: CPU-Only v6.4 v6.5 v6.6 6 Oct 20 40 60 80 100 SE +/- 0.36, N = 3 SE +/- 0.56, N = 3 SE +/- 0.83, N = 4 74.88 73.87 75.58
OpenVINO OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2023.1 Model: Vehicle Detection FP16 - Device: CPU v6.4 v6.5 v6.6 6 Oct 4 8 12 16 20 SE +/- 0.13, N = 6 SE +/- 0.04, N = 3 SE +/- 0.01, N = 3 14.35 12.74 12.67 MIN: 9.07 / MAX: 112.91 MIN: 8.82 / MAX: 56.2 MIN: 8.91 / MAX: 252.77 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org FPS, More Is Better OpenVINO 2023.1 Model: Vehicle Detection FP16 - Device: CPU v6.4 v6.5 v6.6 6 Oct 600 1200 1800 2400 3000 SE +/- 24.17, N = 6 SE +/- 8.62, N = 3 SE +/- 0.78, N = 3 2572.63 2896.61 2911.44 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
Rodinia Rodinia is a suite focused upon accelerating compute-intensive applications with accelerators. CUDA, OpenMP, and OpenCL parallel models are supported by the included applications. This profile utilizes select OpenCL, NVIDIA CUDA and OpenMP test binaries at the moment. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better Rodinia 3.1 Test: OpenMP Leukocyte v6.4 v6.5 v6.6 6 Oct 8 16 24 32 40 SE +/- 0.52, N = 15 SE +/- 0.08, N = 3 SE +/- 0.40, N = 3 34.17 31.49 32.18 1. (CXX) g++ options: -O2 -lOpenCL
Stress-NG Stress-NG is a Linux stress tool developed by Colin Ian King. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: Pthread v6.4 v6.5 v6.6 6 Oct 8K 16K 24K 32K 40K SE +/- 43.26, N = 3 SE +/- 820.44, N = 15 SE +/- 106.91, N = 3 36288.24 38759.74 35491.62 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
AOM AV1 This is a test of the AOMedia AV1 encoder (libaom) developed by AOMedia and Google as the AV1 Codec Library. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Frames Per Second, More Is Better AOM AV1 3.7 Encoder Mode: Speed 9 Realtime - Input: Bosphorus 4K v6.4 v6.5 v6.6 6 Oct 11 22 33 44 55 SE +/- 1.51, N = 15 SE +/- 0.88, N = 15 SE +/- 1.27, N = 15 47.54 45.14 46.08 1. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
Stress-NG Stress-NG is a Linux stress tool developed by Colin Ian King. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: Fused Multiply-Add v6.4 v6.5 v6.6 6 Oct 60M 120M 180M 240M 300M SE +/- 2984105.55, N = 15 SE +/- 3180404.53, N = 3 SE +/- 480576.96, N = 3 253366692.27 243730838.96 261241623.38 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: Hash v6.4 v6.5 v6.6 6 Oct 4M 8M 12M 16M 20M SE +/- 343354.96, N = 15 SE +/- 109703.46, N = 3 SE +/- 232420.75, N = 3 17389901.32 16889506.12 18109792.85 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: NUMA v6.4 v6.5 v6.6 6 Oct 120 240 360 480 600 SE +/- 7.82, N = 15 SE +/- 2.88, N = 3 SE +/- 1.03, N = 3 514.09 575.37 474.39 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
OpenBenchmarking.org MIPS, More Is Better 7-Zip Compression 22.01 Test: Compression Rating v6.4 v6.5 v6.6 6 Oct 60K 120K 180K 240K 300K SE +/- 2800.68, N = 3 SE +/- 1889.96, N = 3 SE +/- 2866.87, N = 3 283556 279135 273731 1. (CXX) g++ options: -lpthread -ldl -O2 -fPIC
Stress-NG Stress-NG is a Linux stress tool developed by Colin Ian King. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: Matrix 3D Math v6.4 v6.5 v6.6 6 Oct 7K 14K 21K 28K 35K SE +/- 329.39, N = 3 SE +/- 2392.89, N = 14 SE +/- 48.00, N = 3 28901.02 32314.93 30007.33 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
AOM AV1 This is a test of the AOMedia AV1 encoder (libaom) developed by AOMedia and Google as the AV1 Codec Library. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Frames Per Second, More Is Better AOM AV1 3.7 Encoder Mode: Speed 8 Realtime - Input: Bosphorus 4K v6.4 v6.5 v6.6 6 Oct 10 20 30 40 50 SE +/- 1.53, N = 15 SE +/- 1.26, N = 12 SE +/- 1.29, N = 12 41.67 42.38 41.95 1. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
OpenVINO OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2023.1 Model: Face Detection FP16-INT8 - Device: CPU v6.4 v6.5 v6.6 6 Oct 70 140 210 280 350 SE +/- 0.31, N = 3 SE +/- 0.56, N = 3 SE +/- 0.67, N = 3 337.22 341.06 341.19 MIN: 252.67 / MAX: 470.11 MIN: 238.2 / MAX: 437.74 MIN: 242.27 / MAX: 401.59 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org FPS, More Is Better OpenVINO 2023.1 Model: Face Detection FP16-INT8 - Device: CPU v6.4 v6.5 v6.6 6 Oct 70 140 210 280 350 SE +/- 0.28, N = 3 SE +/- 0.51, N = 3 SE +/- 0.63, N = 3 331.20 327.68 327.61 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
AOM AV1 This is a test of the AOMedia AV1 encoder (libaom) developed by AOMedia and Google as the AV1 Codec Library. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Frames Per Second, More Is Better AOM AV1 3.7 Encoder Mode: Speed 6 Realtime - Input: Bosphorus 4K v6.4 v6.5 v6.6 6 Oct 10 20 30 40 50 SE +/- 0.79, N = 15 SE +/- 1.36, N = 12 SE +/- 2.39, N = 12 41.51 42.42 45.27 1. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
nekRS nekRS is an open-source Navier Stokes solver based on the spectral element method. NekRS supports both CPU and GPU/accelerator support though this test profile is currently configured for CPU execution. NekRS is part of Nek5000 of the Mathematics and Computer Science MCS at Argonne National Laboratory. This nekRS benchmark is primarily relevant to large core count HPC servers and otherwise may be very time consuming on smaller systems. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org flops/rank, More Is Better nekRS 23.0 Input: Kershaw v6.4 v6.5 v6.6 6 Oct 1300M 2600M 3900M 5200M 6500M SE +/- 38898803.89, N = 3 SE +/- 56766399.45, N = 6 SE +/- 77700626.41, N = 3 5865983333 5809220000 5817096667 1. (CXX) g++ options: -fopenmp -O2 -march=native -mtune=native -ftree-vectorize -rdynamic -lmpi_cxx -lmpi
AOM AV1 This is a test of the AOMedia AV1 encoder (libaom) developed by AOMedia and Google as the AV1 Codec Library. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Frames Per Second, More Is Better AOM AV1 3.7 Encoder Mode: Speed 10 Realtime - Input: Bosphorus 4K v6.4 v6.5 v6.6 6 Oct 12 24 36 48 60 SE +/- 1.50, N = 12 SE +/- 2.34, N = 15 SE +/- 1.97, N = 15 44.23 51.60 47.18 1. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
OpenBenchmarking.org Frames Per Second, More Is Better AOM AV1 3.7 Encoder Mode: Speed 11 Realtime - Input: Bosphorus 4K v6.4 v6.5 v6.6 6 Oct 11 22 33 44 55 SE +/- 1.22, N = 15 SE +/- 1.08, N = 15 SE +/- 2.00, N = 12 49.09 45.45 48.21 1. (CXX) g++ options: -O3 -std=c++11 -U_FORTIFY_SOURCE -lm
Stress-NG Stress-NG is a Linux stress tool developed by Colin Ian King. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: CPU Stress v6.4 v6.5 v6.6 6 Oct 40K 80K 120K 160K 200K SE +/- 341.26, N = 3 SE +/- 1870.15, N = 13 SE +/- 692.17, N = 3 200494.80 198750.35 203979.83 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
nekRS nekRS is an open-source Navier Stokes solver based on the spectral element method. NekRS supports both CPU and GPU/accelerator support though this test profile is currently configured for CPU execution. NekRS is part of Nek5000 of the Mathematics and Computer Science MCS at Argonne National Laboratory. This nekRS benchmark is primarily relevant to large core count HPC servers and otherwise may be very time consuming on smaller systems. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org flops/rank, More Is Better nekRS 23.0 Input: TurboPipe Periodic v6.4 v6.5 v6.6 6 Oct 900M 1800M 2700M 3600M 4500M SE +/- 40424400.36, N = 3 SE +/- 30449494.25, N = 3 SE +/- 4200842.51, N = 3 4098953333 4101490000 4272803333 1. (CXX) g++ options: -fopenmp -O2 -march=native -mtune=native -ftree-vectorize -rdynamic -lmpi_cxx -lmpi
OpenVINO OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2023.1 Model: Person Vehicle Bike Detection FP16 - Device: CPU v6.4 v6.5 v6.6 6 Oct 4 8 12 16 20 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 17.70 17.93 17.83 MIN: 12.28 / MAX: 42.46 MIN: 12.49 / MAX: 40.07 MIN: 12.43 / MAX: 36.82 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org FPS, More Is Better OpenVINO 2023.1 Model: Person Vehicle Bike Detection FP16 - Device: CPU v6.4 v6.5 v6.6 6 Oct 1400 2800 4200 5600 7000 SE +/- 8.07, N = 3 SE +/- 8.34, N = 3 SE +/- 6.56, N = 3 6316.44 6234.83 6265.86 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2023.1 Model: Face Detection FP16 - Device: CPU v6.4 v6.5 v6.6 6 Oct 60 120 180 240 300 SE +/- 1.25, N = 3 SE +/- 0.39, N = 3 SE +/- 0.68, N = 3 293.60 232.33 231.45 MIN: 175.19 / MAX: 761.95 MIN: 171.26 / MAX: 448.32 MIN: 167.08 / MAX: 429 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org FPS, More Is Better OpenVINO 2023.1 Model: Face Detection FP16 - Device: CPU v6.4 v6.5 v6.6 6 Oct 40 80 120 160 200 SE +/- 0.59, N = 3 SE +/- 0.26, N = 3 SE +/- 0.46, N = 3 125.69 159.00 159.56 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2023.1 Model: Road Segmentation ADAS FP16-INT8 - Device: CPU v6.4 v6.5 v6.6 6 Oct 20 40 60 80 100 SE +/- 0.13, N = 3 SE +/- 0.16, N = 3 SE +/- 0.11, N = 3 74.40 74.65 74.13 MIN: 53.51 / MAX: 281.3 MIN: 56.35 / MAX: 221.5 MIN: 53.68 / MAX: 267.9 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org FPS, More Is Better OpenVINO 2023.1 Model: Road Segmentation ADAS FP16-INT8 - Device: CPU v6.4 v6.5 v6.6 6 Oct 300 600 900 1200 1500 SE +/- 2.60, N = 3 SE +/- 3.25, N = 3 SE +/- 2.05, N = 3 1503.54 1498.56 1508.77 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2023.1 Model: Person Detection FP32 - Device: CPU v6.4 v6.5 v6.6 6 Oct 20 40 60 80 100 SE +/- 0.95, N = 3 SE +/- 0.02, N = 3 SE +/- 0.10, N = 3 77.70 67.01 65.42 MIN: 49.99 / MAX: 537.56 MIN: 47.99 / MAX: 141.48 MIN: 51.27 / MAX: 181.2 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org FPS, More Is Better OpenVINO 2023.1 Model: Person Detection FP32 - Device: CPU v6.4 v6.5 v6.6 6 Oct 120 240 360 480 600 SE +/- 5.84, N = 3 SE +/- 0.11, N = 3 SE +/- 0.92, N = 3 475.66 551.66 565.02 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2023.1 Model: Person Detection FP16 - Device: CPU v6.4 v6.5 v6.6 6 Oct 20 40 60 80 100 SE +/- 0.97, N = 3 SE +/- 0.06, N = 3 SE +/- 0.08, N = 3 77.97 66.83 65.43 MIN: 47.82 / MAX: 365.81 MIN: 47.69 / MAX: 144.02 MIN: 52.22 / MAX: 151.59 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org FPS, More Is Better OpenVINO 2023.1 Model: Person Detection FP16 - Device: CPU v6.4 v6.5 v6.6 6 Oct 120 240 360 480 600 SE +/- 5.93, N = 3 SE +/- 0.50, N = 3 SE +/- 0.67, N = 3 474.05 553.15 564.91 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
Stress-NG Stress-NG is a Linux stress tool developed by Colin Ian King. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: Wide Vector Math v6.4 v6.5 v6.6 6 Oct 1.3M 2.6M 3.9M 5.2M 6.5M SE +/- 72348.80, N = 3 SE +/- 142745.75, N = 12 SE +/- 3935.20, N = 3 5961543.12 5808342.74 6205140.99 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
OpenVINO OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2023.1 Model: Vehicle Detection FP16-INT8 - Device: CPU v6.4 v6.5 v6.6 6 Oct 5 10 15 20 25 SE +/- 0.09, N = 3 SE +/- 0.02, N = 3 SE +/- 0.03, N = 3 22.28 22.25 22.10 MIN: 14.79 / MAX: 71.65 MIN: 15.09 / MAX: 75.35 MIN: 16.11 / MAX: 81 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org FPS, More Is Better OpenVINO 2023.1 Model: Vehicle Detection FP16-INT8 - Device: CPU v6.4 v6.5 v6.6 6 Oct 1100 2200 3300 4400 5500 SE +/- 20.34, N = 3 SE +/- 3.85, N = 3 SE +/- 8.34, N = 3 5018.03 5018.07 5053.25 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2023.1 Model: Face Detection Retail FP16 - Device: CPU v6.4 v6.5 v6.6 6 Oct 3 6 9 12 15 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 9.85 10.15 9.98 MIN: 7.51 / MAX: 55.05 MIN: 7.22 / MAX: 47.29 MIN: 7.7 / MAX: 41.05 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org FPS, More Is Better OpenVINO 2023.1 Model: Face Detection Retail FP16 - Device: CPU v6.4 v6.5 v6.6 6 Oct 2K 4K 6K 8K 10K SE +/- 16.48, N = 3 SE +/- 10.14, N = 3 SE +/- 21.17, N = 3 11342.54 11013.29 11190.46 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2023.1 Model: Road Segmentation ADAS FP16 - Device: CPU v6.4 v6.5 v6.6 6 Oct 8 16 24 32 40 SE +/- 0.39, N = 3 SE +/- 0.03, N = 3 SE +/- 0.05, N = 3 33.21 31.98 31.88 MIN: 23.34 / MAX: 169.71 MIN: 23.5 / MAX: 154.49 MIN: 23.76 / MAX: 159.97 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org FPS, More Is Better OpenVINO 2023.1 Model: Road Segmentation ADAS FP16 - Device: CPU v6.4 v6.5 v6.6 6 Oct 200 400 600 800 1000 SE +/- 13.29, N = 3 SE +/- 1.10, N = 3 SE +/- 1.66, N = 3 1112.46 1155.64 1159.20 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2023.1 Model: Face Detection Retail FP16-INT8 - Device: CPU v6.4 v6.5 v6.6 6 Oct 2 4 6 8 10 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 6.89 6.89 6.88 MIN: 5.52 / MAX: 36.52 MIN: 5.7 / MAX: 43.92 MIN: 5.82 / MAX: 40.08 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org FPS, More Is Better OpenVINO 2023.1 Model: Face Detection Retail FP16-INT8 - Device: CPU v6.4 v6.5 v6.6 6 Oct 3K 6K 9K 12K 15K SE +/- 27.81, N = 3 SE +/- 32.50, N = 3 SE +/- 14.33, N = 3 16182.55 16152.10 16160.29 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2023.1 Model: Weld Porosity Detection FP16 - Device: CPU v6.4 v6.5 v6.6 6 Oct 2 4 6 8 10 SE +/- 0.09, N = 3 SE +/- 0.03, N = 3 SE +/- 0.04, N = 3 6.60 6.21 6.16 MIN: 4.28 / MAX: 87.64 MIN: 4.25 / MAX: 71.23 MIN: 4.39 / MAX: 72.48 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org FPS, More Is Better OpenVINO 2023.1 Model: Weld Porosity Detection FP16 - Device: CPU v6.4 v6.5 v6.6 6 Oct 4K 8K 12K 16K 20K SE +/- 240.02, N = 3 SE +/- 92.46, N = 3 SE +/- 116.21, N = 3 16690.36 17909.28 18105.63 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2023.1 Model: Handwritten English Recognition FP16-INT8 - Device: CPU v6.4 v6.5 v6.6 6 Oct 10 20 30 40 50 SE +/- 0.11, N = 3 SE +/- 0.05, N = 3 SE +/- 0.06, N = 3 45.63 45.76 45.53 MIN: 38.64 / MAX: 91.4 MIN: 38.33 / MAX: 110.96 MIN: 40.67 / MAX: 112.3 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org FPS, More Is Better OpenVINO 2023.1 Model: Handwritten English Recognition FP16-INT8 - Device: CPU v6.4 v6.5 v6.6 6 Oct 500 1000 1500 2000 2500 SE +/- 5.85, N = 3 SE +/- 2.61, N = 3 SE +/- 3.57, N = 3 2452.14 2445.54 2457.57 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org ms, Fewer Is Better OpenVINO 2023.1 Model: Handwritten English Recognition FP16 - Device: CPU v6.4 v6.5 v6.6 6 Oct 7 14 21 28 35 SE +/- 0.09, N = 3 SE +/- 0.08, N = 3 SE +/- 0.07, N = 3 31.10 31.32 31.19 MIN: 24.91 / MAX: 99.63 MIN: 25.09 / MAX: 91.66 MIN: 24.8 / MAX: 100.26 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
OpenBenchmarking.org FPS, More Is Better OpenVINO 2023.1 Model: Handwritten English Recognition FP16 - Device: CPU v6.4 v6.5 v6.6 6 Oct 800 1600 2400 3200 4000 SE +/- 9.62, N = 3 SE +/- 8.61, N = 3 SE +/- 7.83, N = 3 3597.65 3572.89 3587.67 1. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl
VVenC VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Frames Per Second, More Is Better VVenC 1.9 Video Input: Bosphorus 1080p - Video Preset: Faster v6.4 v6.5 v6.6 6 Oct 7 14 21 28 35 SE +/- 0.34, N = 4 SE +/- 0.08, N = 3 SE +/- 0.48, N = 12 28.27 28.72 27.33 1. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects
Blender OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.6 Blend File: Classroom - Compute: CPU-Only v6.4 v6.5 v6.6 6 Oct 12 24 36 48 60 SE +/- 0.28, N = 3 SE +/- 0.26, N = 3 SE +/- 0.24, N = 3 53.14 54.01 53.11
Stress-NG Stress-NG is a Linux stress tool developed by Colin Ian King. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: CPU Cache v6.4 v6.5 v6.6 6 Oct 200K 400K 600K 800K 1000K SE +/- 7749.59, N = 3 SE +/- 4820.43, N = 3 SE +/- 1835.02, N = 3 938764.06 827481.85 709241.85 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
OpenRadioss OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/ and https://github.com/OpenRadioss/ModelExchange/tree/main/Examples. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better OpenRadioss 2023.09.15 Model: Cell Phone Drop Test v6.4 v6.5 v6.6 6 Oct 8 16 24 32 40 SE +/- 0.01, N = 3 SE +/- 0.17, N = 3 SE +/- 0.25, N = 3 32.37 33.81 33.34
VVenC VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Frames Per Second, More Is Better VVenC 1.9 Video Input: Bosphorus 1080p - Video Preset: Fast v6.4 v6.5 v6.6 6 Oct 4 8 12 16 20 SE +/- 0.19, N = 3 SE +/- 0.19, N = 3 SE +/- 0.04, N = 3 16.34 16.28 16.40 1. (CXX) g++ options: -O3 -flto=auto -fno-fat-lto-objects
SVT-AV1 OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 1.7 Encoder Mode: Preset 12 - Input: Bosphorus 4K v6.4 v6.5 v6.6 6 Oct 30 60 90 120 150 SE +/- 8.17, N = 15 SE +/- 4.93, N = 15 SE +/- 6.81, N = 15 133.93 114.32 122.12 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
Stress-NG Stress-NG is a Linux stress tool developed by Colin Ian King. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: Vector Math v6.4 v6.5 v6.6 6 Oct 80K 160K 240K 320K 400K SE +/- 3943.55, N = 5 SE +/- 3836.13, N = 3 SE +/- 522.79, N = 3 381412.82 359634.35 384963.64 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: Semaphores v6.4 v6.5 v6.6 6 Oct 40M 80M 120M 160M 200M SE +/- 1308655.40, N = 3 SE +/- 1418717.66, N = 3 SE +/- 1941141.12, N = 5 177471711.51 154525821.93 180684828.43 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
Rodinia Rodinia is a suite focused upon accelerating compute-intensive applications with accelerators. CUDA, OpenMP, and OpenCL parallel models are supported by the included applications. This profile utilizes select OpenCL, NVIDIA CUDA and OpenMP test binaries at the moment. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better Rodinia 3.1 Test: OpenMP LavaMD v6.4 v6.5 v6.6 6 Oct 8 16 24 32 40 SE +/- 0.23, N = 3 SE +/- 0.14, N = 3 SE +/- 0.14, N = 3 35.24 35.28 35.06 1. (CXX) g++ options: -O2 -lOpenCL
Stress-NG Stress-NG is a Linux stress tool developed by Colin Ian King. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: AVX-512 VNNI v6.4 v6.5 v6.6 6 Oct 3M 6M 9M 12M 15M SE +/- 48287.23, N = 3 SE +/- 137441.83, N = 3 SE +/- 45820.46, N = 3 11752668.18 11016794.41 11946993.70 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: AVL Tree v6.4 v6.5 v6.6 6 Oct 200 400 600 800 1000 SE +/- 1.81, N = 3 SE +/- 2.77, N = 3 SE +/- 1.34, N = 3 790.23 794.30 816.16 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: Glibc Qsort Data Sorting v6.4 v6.5 v6.6 6 Oct 400 800 1200 1600 2000 SE +/- 0.72, N = 3 SE +/- 3.46, N = 3 SE +/- 5.70, N = 3 1875.83 1916.20 1931.83 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: Vector Floating Point v6.4 v6.5 v6.6 6 Oct 50K 100K 150K 200K 250K SE +/- 620.60, N = 3 SE +/- 1308.77, N = 3 SE +/- 1538.11, N = 3 244931.41 240694.16 248017.78 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: Vector Shuffle v6.4 v6.5 v6.6 6 Oct 120K 240K 360K 480K 600K SE +/- 1079.72, N = 3 SE +/- 652.21, N = 3 SE +/- 838.85, N = 3 551599.32 549410.10 562355.24 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: x86_64 RdRand v6.4 v6.5 v6.6 6 Oct 140K 280K 420K 560K 700K SE +/- 674.33, N = 3 SE +/- 196.88, N = 3 SE +/- 280.73, N = 3 637179.52 648063.91 642917.30 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: Zlib v6.4 v6.5 v6.6 6 Oct 2K 4K 6K 8K 10K SE +/- 22.76, N = 3 SE +/- 29.13, N = 3 SE +/- 11.00, N = 3 8220.83 8330.22 8374.58 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: MMAP v6.4 v6.5 v6.6 6 Oct 2K 4K 6K 8K 10K SE +/- 47.78, N = 3 SE +/- 60.98, N = 3 SE +/- 76.34, N = 3 7996.78 8195.87 8382.54 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: Malloc v6.4 v6.5 v6.6 6 Oct 50M 100M 150M 200M 250M SE +/- 129847.43, N = 3 SE +/- 655876.42, N = 3 SE +/- 619909.09, N = 3 217211109.03 219735444.05 222268148.76 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: System V Message Passing v6.4 v6.5 v6.6 6 Oct 4M 8M 12M 16M 20M SE +/- 54732.60, N = 3 SE +/- 57359.04, N = 3 SE +/- 6699.83, N = 3 18811154.40 19106643.99 17488847.83 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lc -lcrypt -ldl -lEGL -lGLESv2 -ljpeg -lmpfr -lpthread -lrt -lsctp -lz
SVT-AV1 OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 1.7 Encoder Mode: Preset 13 - Input: Bosphorus 4K v6.4 v6.5 v6.6 6 Oct 30 60 90 120 150 SE +/- 9.48, N = 12 SE +/- 5.99, N = 15 SE +/- 1.29, N = 3 145.92 112.26 116.74 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenFOAM OpenFOAM is the leading free, open-source software for computational fluid dynamics (CFD). This test profile currently uses the drivaerFastback test case for analyzing automotive aerodynamics or alternatively the older motorBike input. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better OpenFOAM 10 Input: drivaerFastback, Small Mesh Size - Execution Time v6.4 v6.5 v6.6 6 Oct 9 18 27 36 45 37.84 36.77 37.02 1. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfiniteVolume -lmeshTools -lparallel -llagrangian -lregionModels -lgenericPatchFields -lOpenFOAM -ldl -lm
OpenBenchmarking.org Seconds, Fewer Is Better OpenFOAM 10 Input: drivaerFastback, Small Mesh Size - Mesh Time v6.4 v6.5 v6.6 6 Oct 9 18 27 36 45 39.36 38.49 37.35 1. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfiniteVolume -lmeshTools -lparallel -llagrangian -lregionModels -lgenericPatchFields -lOpenFOAM -ldl -lm
Blender OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.6 Blend File: BMW27 - Compute: CPU-Only v6.4 v6.5 v6.6 6 Oct 5 10 15 20 25 SE +/- 0.17, N = 3 SE +/- 0.18, N = 3 SE +/- 0.16, N = 3 21.48 21.96 21.12
NAS Parallel Benchmarks NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: FT.C v6.4 v6.5 v6.6 6 Oct 20K 40K 60K 80K 100K SE +/- 1023.16, N = 15 SE +/- 2189.90, N = 15 SE +/- 2154.10, N = 12 103248.13 112362.39 114793.58 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.4
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: BT.C v6.4 v6.5 v6.6 6 Oct 70K 140K 210K 280K 350K SE +/- 2411.99, N = 11 SE +/- 3511.64, N = 3 SE +/- 958.09, N = 3 322119.99 306002.20 312977.30 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.4
SVT-AV1 OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 1.7 Encoder Mode: Preset 4 - Input: Bosphorus 1080p v6.4 v6.5 v6.6 6 Oct 3 6 9 12 15 SE +/- 0.04, N = 3 SE +/- 0.06, N = 3 SE +/- 0.10, N = 3 11.59 11.10 11.56 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
NAS Parallel Benchmarks NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: SP.B v6.4 v6.5 v6.6 6 Oct 40K 80K 120K 160K 200K SE +/- 1564.28, N = 13 SE +/- 1289.87, N = 15 SE +/- 1609.35, N = 15 176009.15 178582.58 180233.00 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.4
SVT-AV1 OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 1.7 Encoder Mode: Preset 12 - Input: Bosphorus 1080p v6.4 v6.5 v6.6 6 Oct 90 180 270 360 450 SE +/- 29.20, N = 15 SE +/- 28.60, N = 15 SE +/- 23.45, N = 15 426.84 406.64 431.13 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
NAS Parallel Benchmarks NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: SP.C v6.4 v6.5 v6.6 6 Oct 40K 80K 120K 160K 200K SE +/- 1944.01, N = 5 SE +/- 2009.88, N = 3 SE +/- 2038.85, N = 3 193221.95 190905.82 191378.82 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.4
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: CG.C v6.4 v6.5 v6.6 6 Oct 11K 22K 33K 44K 55K SE +/- 525.17, N = 6 SE +/- 486.03, N = 15 SE +/- 305.31, N = 3 51894.85 52032.10 53274.79 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.4
SVT-AV1 OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 1.7 Encoder Mode: Preset 13 - Input: Bosphorus 1080p v6.4 v6.5 v6.6 6 Oct 110 220 330 440 550 SE +/- 28.12, N = 15 SE +/- 27.06, N = 15 SE +/- 29.52, N = 15 518.77 459.29 468.57 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
Rodinia Rodinia is a suite focused upon accelerating compute-intensive applications with accelerators. CUDA, OpenMP, and OpenCL parallel models are supported by the included applications. This profile utilizes select OpenCL, NVIDIA CUDA and OpenMP test binaries at the moment. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better Rodinia 3.1 Test: OpenMP CFD Solver v6.4 v6.5 v6.6 6 Oct 2 4 6 8 10 SE +/- 0.058, N = 3 SE +/- 0.093, N = 3 SE +/- 0.076, N = 3 8.713 8.886 8.164 1. (CXX) g++ options: -O2 -lOpenCL
NAS Parallel Benchmarks NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: LU.C v6.4 v6.5 v6.6 6 Oct 60K 120K 180K 240K 300K SE +/- 2141.41, N = 3 SE +/- 3576.66, N = 3 SE +/- 946.75, N = 3 269542.21 268492.02 273013.53 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.4
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: EP.C v6.4 v6.5 v6.6 6 Oct 2K 4K 6K 8K 10K SE +/- 222.04, N = 12 SE +/- 266.91, N = 15 SE +/- 228.39, N = 12 10844.90 10479.79 10840.56 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.4
NAS Parallel Benchmarks NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. This test profile offers selecting the different NPB tests/problems and varying problem sizes. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Total Mop/s, More Is Better NAS Parallel Benchmarks 3.4 Test / Class: MG.C v6.4 v6.5 v6.6 6 Oct 50K 100K 150K 200K 250K SE +/- 2713.74, N = 15 SE +/- 2603.01, N = 3 SE +/- 2417.02, N = 15 239408.11 234268.14 239660.40 1. (F9X) gfortran options: -O3 -march=native -lmpi_usempif08 -lmpi_mpifh -lmpi -lopen-rte -lopen-pal -lhwloc -levent_core -levent_pthreads -lm -lz 2. Open MPI 4.1.4
v6.4 Processor: 2 x Intel Xeon Max 9480 @ 3.50GHz (112 Cores / 224 Threads), Motherboard: Supermicro X13DEM v1.10 (1.3 BIOS), Chipset: Intel Device 1bce, Memory: 512GB, Disk: 2 x 7682GB INTEL SSDPF2KX076TZ, Graphics: ASPEED, Monitor: VE228, Network: 2 x Broadcom BCM57508 NetXtreme-E 10Gb/25Gb/40Gb/50Gb/100Gb/200Gb
OS: Ubuntu 23.04, Kernel: 6.4.0-060400-generic (x86_64), Desktop: GNOME Shell 44.0, Display Server: X Server 1.21.1.7, Compiler: GCC 12.2.0, File-System: ext4, Screen Resolution: 1920x1080
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-12-Pa930Z/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-Pa930Z/gcc-12-12.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_cpufreq performance - CPU Microcode: 0x2c0001d1Java Notes: OpenJDK Runtime Environment (build 17.0.6+10-Ubuntu-1ubuntu2)Python Notes: Python 3.11.2Security Notes: itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 4 October 2023 15:43 by user phoronix.
v6.5 Processor: 2 x Intel Xeon Max 9480 @ 3.50GHz (112 Cores / 224 Threads), Motherboard: Supermicro X13DEM v1.10 (1.3 BIOS), Chipset: Intel Device 1bce, Memory: 512GB, Disk: 2 x 7682GB INTEL SSDPF2KX076TZ, Graphics: ASPEED, Network: 2 x Broadcom BCM57508 NetXtreme-E 10Gb/25Gb/40Gb/50Gb/100Gb/200Gb
OS: Ubuntu 23.04, Kernel: 6.5.0-060500-generic (x86_64), Desktop: GNOME Shell 44.0, Display Server: X Server 1.21.1.7, Compiler: GCC 12.2.0, File-System: ext4, Screen Resolution: 1024x768
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-12-Pa930Z/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-Pa930Z/gcc-12-12.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_cpufreq performance - CPU Microcode: 0x2c0001d1Java Notes: OpenJDK Runtime Environment (build 17.0.6+10-Ubuntu-1ubuntu2)Python Notes: Python 3.11.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 5 October 2023 21:21 by user phoronix.
v6.6 6 Oct Processor: 2 x Intel Xeon Max 9480 @ 3.50GHz (112 Cores / 224 Threads), Motherboard: Supermicro X13DEM v1.10 (1.3 BIOS), Chipset: Intel Device 1bce, Memory: 512GB, Disk: 2 x 7682GB INTEL SSDPF2KX076TZ, Graphics: ASPEED, Network: 2 x Broadcom BCM57508 NetXtreme-E 10Gb/25Gb/40Gb/50Gb/100Gb/200Gb
OS: Ubuntu 23.04, Kernel: 6.6.0-060600rc4daily20231006-generic (x86_64), Desktop: GNOME Shell 44.0, Display Server: X Server 1.21.1.7, Compiler: GCC 12.2.0, File-System: ext4, Screen Resolution: 1920x1200
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-12-Pa930Z/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-Pa930Z/gcc-12-12.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_cpufreq performance - CPU Microcode: 0x2c0001d1Java Notes: OpenJDK Runtime Environment (build 17.0.6+10-Ubuntu-1ubuntu2)Python Notes: Python 3.11.2Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 6 October 2023 19:45 by user phoronix.