AMD Milan-X Daytona RHEL 8.6 vs. RHEL 9.0 RHEL 8.6 vs. RHEL 9.0 defaults benchmark, EPYC Milan-X 7763 2P. Benchmarks by Michael Larabel looking at RHEL 9.0 performance. RHEL 9.0 test with tsc=nowatchdog added for a future article.
HTML result view exported from: https://openbenchmarking.org/result/2205295-NE-RHEL9395618&sro&grs .
AMD Milan-X Daytona RHEL 8.6 vs. RHEL 9.0 Processor Motherboard Chipset Memory Disk Graphics Monitor Network OS Kernel Desktop Display Server Compiler File-System Screen Resolution RHEL 8.6 RHEL 9.0 tsc=nowatchdog 2 x AMD EPYC 7773X 64-Core @ 2.20GHz (128 Cores / 256 Threads) AMD DAYTONA_X (TYM1008C BIOS) AMD Starship/Matisse 16 x 32 GB DDR4-3200MT/s 36ASF4G72PZ-3G2E2 800GB INTEL SSDPF21Q800GB ASPEED VE228 2 x Mellanox MT27710 Red Hat Enterprise Linux 8.6 4.18.0-372.9.1.el8.x86_64 (x86_64) GNOME Shell 3.32.2 X Server 1.20.11 GCC 8.5.0 20210514 xfs 1920x1080 Red Hat Enterprise Linux 9.0 5.14.0-70.13.1.el9_0.x86_64 (x86_64) GNOME Shell 40.9 GCC 11.2.1 20220127 OpenBenchmarking.org Kernel Details - Transparent Huge Pages: always Compiler Details - RHEL 8.6: --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 - RHEL 9.0 tsc=nowatchdog: --build=x86_64-redhat-linux --disable-libunwind-exceptions --enable-__cxa_atexit --enable-bootstrap --enable-cet --enable-checking=release --enable-gnu-indirect-function --enable-gnu-unique-object --enable-host-bind-now --enable-host-pie --enable-initfini-array --enable-languages=c,c++,fortran,lto --enable-link-serialization=1 --enable-multilib --enable-offload-targets=nvptx-none --enable-plugin --enable-shared --enable-threads=posix --mandir=/usr/share/man --with-arch_32=x86-64 --with-arch_64=x86-64-v2 --with-build-config=bootstrap-lto --with-gcc-major-version-only --with-linker-hash-style=gnu --with-tune=generic --without-cuda-driver --without-isl Processor Details - Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa001228 Java Details - RHEL 8.6: OpenJDK Runtime Environment (build 1.8.0_332-b09) - RHEL 9.0 tsc=nowatchdog: OpenJDK Runtime Environment 18.9 (build 11.0.15+10-LTS) Python Details - RHEL 8.6: Python 3.6.8 - RHEL 9.0 tsc=nowatchdog: Python 3.9.10 Security Details - RHEL 8.6: 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 Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling + srbds: Not affected + tsx_async_abort: Not affected - RHEL 9.0 tsc=nowatchdog: SELinux + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: 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 Retpolines IBPB: conditional IBRS_FW STIBP: always-on RSB filling + srbds: Not affected + tsx_async_abort: Not affected
AMD Milan-X Daytona RHEL 8.6 vs. RHEL 9.0 compress-zstd: 19, Long Mode - Compression Speed compress-zstd: 3, Long Mode - Compression Speed financebench: Bonds OpenMP stress-ng: System V Message Passing financebench: Repo OpenMP stress-ng: Context Switching compress-zstd: 19 - Compression Speed pyperformance: django_template stress-ng: MEMFD compress-zstd: 3 - Compression Speed mrbayes: Primate Phylogeny Analysis onnx: bertsquad-12 - CPU - Standard compress-zstd: 19, Long Mode - Decompression Speed luaradio: FM Deemphasis Filter stress-ng: Vector Math pyperformance: python_startup stress-ng: SENDFILE stress-ng: MMAP stress-ng: Matrix Math mnn: MobileNetV2_224 stress-ng: CPU Stress compress-zstd: 3, Long Mode - Compression Speed mnn: inception-v3 dacapobench: H2 jpegxl-decode: All build-mplayer: Time To Compile nginx: 500 pyperformance: nbody build2: Time To Compile nginx: 1000 phpbench: PHP Benchmark Suite nginx: 200 apache: 500 srsran: OFDM_Test pyperformance: crypto_pyaes luaradio: Five Back to Back FIR Filters build-apache: Time To Compile build-llvm: Unix Makefiles pyperformance: pickle_pure_python srsran: 5G PHY_DL_NR Test 52 PRB SISO 64-QAM pyperformance: regex_compile pyperformance: pathlib pyperformance: json_loads radiance: SMP Parallel apache: 200 coremark: CoreMark Size 666 - Iterations Per Second srsran: 4G PHY_DL_Test 100 PRB SISO 64-QAM dolfyn: Computational Fluid Dynamics askap: tConvolve MPI - Gridding srsran: 4G PHY_DL_Test 100 PRB SISO 256-QAM tjbench: Decompression Throughput etcpak: ETC1 java-jmh: Throughput build-llvm: Ninja compress-zstd: 3, Long Mode - Decompression Speed pennant: sedovbig simdjson: LargeRand build-php: Time To Compile srsran: 4G PHY_DL_Test 100 PRB MIMO 256-QAM x265: Bosphorus 4K svt-vp9: VMAF Optimized - Bosphorus 1080p onnx: ArcFace ResNet-100 - CPU - Standard srsran: 4G PHY_DL_Test 100 PRB MIMO 64-QAM stress-ng: Glibc Qsort Data Sorting compress-zstd: 19, Long Mode - Decompression Speed srsran: 4G PHY_DL_Test 100 PRB SISO 64-QAM build-nodejs: Time To Compile srsran: 4G PHY_DL_Test 100 PRB MIMO 64-QAM stress-ng: Memory Copying pyperformance: 2to3 srsran: 4G PHY_DL_Test 100 PRB MIMO 256-QAM jpegxl-decode: 1 compress-zstd: 19 - Decompression Speed compress-7zip: Decompression Rating incompact3d: input.i3d 193 Cells Per Direction srsran: 4G PHY_DL_Test 100 PRB SISO 256-QAM stress-ng: Semaphores graph500: 26 etcpak: DXT1 ngspice: C7552 build-gdb: Time To Compile jpegxl: PNG - 8 pyperformance: go apache: 1000 astcenc: Medium astcenc: Thorough pyperformance: chaos appleseed: Material Tester numpy: stress-ng: Glibc C String Functions renaissance: ALS Movie Lens dacapobench: Jython graph500: 26 etcpak: ETC1 + Dithering mysqlslap: 256 etcpak: ETC2 stress-ng: Forking askap: tConvolve OpenMP - Gridding appleseed: Emily graph500: 26 asmfish: 1024 Hash Memory, 26 Depth m-queens: Time To Solve askap: tConvolve MPI - Degridding svt-vp9: PSNR/SSIM Optimized - Bosphorus 1080p svt-vp9: Visual Quality Optimized - Bosphorus 1080p incompact3d: X3D-benchmarking input.i3d crafty: Elapsed Time renaissance: Rand Forest blender: BMW27 - CPU-Only oidn: RTLightmap.hdr.4096x4096 oidn: RT.ldr_alb_nrm.3840x2160 amg: ospray: gravity_spheres_volume/dim_512/ao/real_time srsran: 5G PHY_DL_NR Test 52 PRB SISO 64-QAM ospray: gravity_spheres_volume/dim_512/scivis/real_time securemark: SecureMark-TLS graph500: 26 stress-ng: Crypto hpcg: blender: Pabellon Barcelona - CPU-Only appleseed: Disney Material stress-ng: Malloc ospray: particle_volume/ao/real_time ospray: particle_volume/scivis/real_time liquid-dsp: 64 - 256 - 57 sqlite-speedtest: Timed Time - Size 1,000 ngspice: C2670 jpegxl: JPEG - 8 astcenc: Exhaustive oidn: RT.hdr_alb_nrm.3840x2160 build-linux-kernel: defconfig askap: tConvolve OpenMP - Degridding compress-zstd: 19, Long Mode - Compression Speed openssl: blender: Classroom - CPU-Only blender: Fishy Cat - CPU-Only liquid-dsp: 128 - 256 - 57 liquid-dsp: 160 - 256 - 57 pybench: Total For Average Test Times openssl: namd: ATPase Simulation - 327,506 Atoms onnx: GPT-2 - CPU - Standard ospray: gravity_spheres_volume/dim_512/pathtracer/real_time compress-zstd: 19 - Compression Speed stress-ng: CPU Cache luajit: Composite stress-ng: Atomic blender: Barbershop - CPU-Only onnx: yolov4 - CPU - Standard pyperformance: float simdjson: Kostya onnx: super-resolution-10 - CPU - Standard mnn: mobilenet-v1-1.0 mnn: SqueezeNetV1.0 mnn: resnet-v2-50 mnn: squeezenetv1.1 mnn: mobilenetV3 stress-ng: Futex mysqlslap: 128 askap: Hogbom Clean OpenMP compress-7zip: Compression Rating himeno: Poisson Pressure Solver ospray: particle_volume/pathtracer/real_time luaradio: Complex Phase compress-zstd: 3, Long Mode - Decompression Speed compress-zstd: 19 - Decompression Speed compress-zstd: 3 - Compression Speed renaissance: Apache Spark ALS renaissance: Scala Dotty dacapobench: Tradebeans incompact3d: input.i3d 129 Cells Per Direction pennant: leblancbig RHEL 8.6 RHEL 9.0 tsc=nowatchdog 23.0 317.4 114294.052083 4945066.41 61022.161458 13347685.81 73.2 82.2 1536.35 9464.9 183.191 687 2559.5 173.5 565856.71 12.3 1665050.95 3907.79 465746.50 4.581 216996.46 583.7 31.875 4230 301.07 7.395 97314.91 143 49.683 97374.02 525005 97400.21 111137.30 108966667 116 461.1 19.875 162.676 587 49.9 225 21.4 27.7 207.245 121914.60 4180906.330324 117.5 20.680 66640.2 122.8 189.450707 243.882 166378871533.27 91.367 3663.8 5.757579 0.83 34.193 116.9 23.86 387.22 1081 327.4 1356.46 3155.0 329.7 96.034 110.7 12203.20 393 361.7 50.78 3099.0 721671 10.1577638 360.8 18815432.41 360311000 965.990 112.595 51.773 0.89 316 101529.21 4.0684 5.3483 138 340.503722 431.91 13787576.86 9758.4 4316 952659000 225.758 387 162.669 33665.32 10750.9 166.807074 492999000 295354804 5.689 64589.7 399.83 316.95 306.666916 8615936 762.8 14.98 0.95 1.93 2051105333 16.5845 110.9 15.9199 218474 987760000 162556.21 38.2882 45.30 47.228196 601718674.55 128.057 123.180 3054800000 62.676 147.732 26.56 5.9835 1.92 19.717 14868.3 60.5 1763867.7 37.24 19.11 5653166667 5710066667 1235 27000.2 0.22581 6187 25.6151 109.2 16.95 1295.56 123081.86 142.62 265 133 2.58 4826 4.709 8.485 28.916 6.053 6.003 192662.22 479 319.776 737301 3547.511761 164.562 297.9 3182.7 2675.0 6998.9 2695.0 866.5 8248 3.16484340 4.230145 60.9 665.3 57020.289062 9517502.50 38371.824219 20538498.04 111.1 57.4 2153.91 6865.8 141.380 536 3240.5 138.7 705122.85 10.1 2017286.05 4727.15 387212.68 3.821 258331.25 686.5 27.265 3656 347.44 8.506 84779.30 164 56.723 85299.53 597949 85903.76 98091.14 123000000 130 415.4 22.061 180.464 532 54.9 205 19.5 25.4 190.257 112402.32 4531044.492776 127.0 19.191 71567.1 130.9 177.853890 259.309 156642355939.16 97.015 3885.9 5.429139 0.88 36.244 123.9 22.54 409.83 1143 345.8 1285.16 3329.7 347.5 101.108 116.4 12831.44 374 378.5 53.13 3240.2 754378 10.6130292 375.0 18135356.63 373684000 1001.425 108.685 53.576 0.92 306 104832.79 4.1918 5.5049 142 330.970478 444.26 14151608.76 10002.0 4420 973443000 230.588 379 159.455 34328.25 10946.1 163.837675 501887000 290367032 5.595 63620.8 405.69 320.93 303.135742 8518882 754.3 14.82 0.94 1.95 2032026667 16.7367 111.9 16.0618 220380 996357000 163933.11 37.9698 45.67 47.609604 596924219.93 129.072 124.141 3075900000 63.058 146.873 26.71 5.9508 1.93 19.819 14792 60.2 1771726.6 37.08 19.19 5673033333 5729800000 1239 27060.3 0.22630 6200 25.5623 109.4 16.94 1295.01 123128.59 142.61 265 133 2.58 5878 3.900 8.714 24.404 6.744 4.924 1910034.62 416 361.819 706479 3563.322525 170.109 230.4 3974.8 3251.0 6985.5 1857.2 897.0 9859 2.50625911 3.772760 OpenBenchmarking.org
Zstd Compression Compression Level: 19, Long Mode - Compression Speed OpenBenchmarking.org MB/s, More Is Better Zstd Compression Compression Level: 19, Long Mode - Compression Speed RHEL 8.6 RHEL 9.0 tsc=nowatchdog 14 28 42 56 70 SE +/- 0.12, N = 3 SE +/- 0.22, N = 3 23.0 60.9 1. RHEL 8.6: *** zstd command line interface 64-bits v1.4.4, by Yann Collet *** 2. RHEL 9.0 tsc=nowatchdog: *** zstd command line interface 64-bits v1.5.1, by Yann Collet ***
Zstd Compression Compression Level: 3, Long Mode - Compression Speed OpenBenchmarking.org MB/s, More Is Better Zstd Compression Compression Level: 3, Long Mode - Compression Speed RHEL 8.6 RHEL 9.0 tsc=nowatchdog 140 280 420 560 700 SE +/- 3.47, N = 5 SE +/- 8.91, N = 15 317.4 665.3 1. RHEL 8.6: *** zstd command line interface 64-bits v1.4.4, by Yann Collet *** 2. RHEL 9.0 tsc=nowatchdog: *** zstd command line interface 64-bits v1.5.1, by Yann Collet ***
FinanceBench Benchmark: Bonds OpenMP OpenBenchmarking.org ms, Fewer Is Better FinanceBench 2016-07-25 Benchmark: Bonds OpenMP RHEL 8.6 RHEL 9.0 tsc=nowatchdog 20K 40K 60K 80K 100K SE +/- 33.81, N = 3 SE +/- 34.47, N = 3 114294.05 57020.29 1. (CXX) g++ options: -O3 -march=native -fopenmp
Stress-NG Test: System V Message Passing OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.14 Test: System V Message Passing RHEL 8.6 RHEL 9.0 tsc=nowatchdog 2M 4M 6M 8M 10M SE +/- 5035.39, N = 3 SE +/- 35135.71, N = 3 4945066.41 9517502.50 1. (CC) gcc options: -O2 -std=gnu99 -lm -lc -lcrypt -ldl -ljpeg -lrt -lsctp -lz -pthread
FinanceBench Benchmark: Repo OpenMP OpenBenchmarking.org ms, Fewer Is Better FinanceBench 2016-07-25 Benchmark: Repo OpenMP RHEL 8.6 RHEL 9.0 tsc=nowatchdog 13K 26K 39K 52K 65K SE +/- 155.78, N = 3 SE +/- 56.79, N = 3 61022.16 38371.82 1. (CXX) g++ options: -O3 -march=native -fopenmp
Stress-NG Test: Context Switching OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.14 Test: Context Switching RHEL 8.6 RHEL 9.0 tsc=nowatchdog 4M 8M 12M 16M 20M SE +/- 65642.17, N = 3 SE +/- 166932.19, N = 3 13347685.81 20538498.04 1. (CC) gcc options: -O2 -std=gnu99 -lm -lc -lcrypt -ldl -ljpeg -lrt -lsctp -lz -pthread
Zstd Compression Compression Level: 19 - Compression Speed OpenBenchmarking.org MB/s, More Is Better Zstd Compression Compression Level: 19 - Compression Speed RHEL 8.6 RHEL 9.0 tsc=nowatchdog 20 40 60 80 100 SE +/- 0.19, N = 3 SE +/- 0.43, N = 3 73.2 111.1 1. RHEL 8.6: *** zstd command line interface 64-bits v1.4.4, by Yann Collet *** 2. RHEL 9.0 tsc=nowatchdog: *** zstd command line interface 64-bits v1.5.1, by Yann Collet ***
PyPerformance Benchmark: django_template OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.0.0 Benchmark: django_template RHEL 8.6 RHEL 9.0 tsc=nowatchdog 20 40 60 80 100 SE +/- 0.10, N = 3 SE +/- 0.07, N = 3 82.2 57.4
Stress-NG Test: MEMFD OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.14 Test: MEMFD RHEL 8.6 RHEL 9.0 tsc=nowatchdog 500 1000 1500 2000 2500 SE +/- 1.32, N = 3 SE +/- 4.60, N = 3 1536.35 2153.91 1. (CC) gcc options: -O2 -std=gnu99 -lm -lc -lcrypt -ldl -ljpeg -lrt -lsctp -lz -pthread
Zstd Compression Compression Level: 3 - Compression Speed OpenBenchmarking.org MB/s, More Is Better Zstd Compression Compression Level: 3 - Compression Speed RHEL 8.6 RHEL 9.0 tsc=nowatchdog 2K 4K 6K 8K 10K SE +/- 47.98, N = 3 SE +/- 25.33, N = 3 9464.9 6865.8 1. RHEL 8.6: *** zstd command line interface 64-bits v1.4.4, by Yann Collet *** 2. RHEL 9.0 tsc=nowatchdog: *** zstd command line interface 64-bits v1.5.1, by Yann Collet ***
Timed MrBayes Analysis Primate Phylogeny Analysis OpenBenchmarking.org Seconds, Fewer Is Better Timed MrBayes Analysis 3.2.7 Primate Phylogeny Analysis RHEL 8.6 RHEL 9.0 tsc=nowatchdog 40 80 120 160 200 SE +/- 3.16, N = 9 SE +/- 1.43, N = 3 183.19 141.38 1. (CC) gcc options: -mmmx -msse -msse2 -msse3 -mssse3 -msse4.1 -msse4.2 -msse4a -msha -maes -mavx -mfma -mavx2 -mrdrnd -mbmi -mbmi2 -madx -mabm -O3 -std=c99 -pedantic -lm
ONNX Runtime Model: bertsquad-12 - Device: CPU - Executor: Standard OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.11 Model: bertsquad-12 - Device: CPU - Executor: Standard RHEL 8.6 RHEL 9.0 tsc=nowatchdog 150 300 450 600 750 SE +/- 4.62, N = 3 SE +/- 6.47, N = 4 687 536 -pthread -lpthread 1. (CXX) g++ options: -ffunction-sections -fdata-sections -march=native -mtune=native -O3 -flto -fno-fat-lto-objects -ldl -lrt
Zstd Compression Compression Level: 19, Long Mode - Decompression Speed OpenBenchmarking.org MB/s, More Is Better Zstd Compression Compression Level: 19, Long Mode - Decompression Speed RHEL 8.6 RHEL 9.0 tsc=nowatchdog 700 1400 2100 2800 3500 SE +/- 7.65, N = 3 SE +/- 5.15, N = 3 2559.5 3240.5 1. RHEL 8.6: *** zstd command line interface 64-bits v1.4.4, by Yann Collet *** 2. RHEL 9.0 tsc=nowatchdog: *** zstd command line interface 64-bits v1.5.1, by Yann Collet ***
LuaRadio Test: FM Deemphasis Filter OpenBenchmarking.org MiB/s, More Is Better LuaRadio 0.9.1 Test: FM Deemphasis Filter RHEL 8.6 RHEL 9.0 tsc=nowatchdog 40 80 120 160 200 SE +/- 1.42, N = 3 SE +/- 0.75, N = 3 173.5 138.7
Stress-NG Test: Vector Math OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.14 Test: Vector Math RHEL 8.6 RHEL 9.0 tsc=nowatchdog 150K 300K 450K 600K 750K SE +/- 157.91, N = 3 SE +/- 396.91, N = 3 565856.71 705122.85 1. (CC) gcc options: -O2 -std=gnu99 -lm -lc -lcrypt -ldl -ljpeg -lrt -lsctp -lz -pthread
PyPerformance Benchmark: python_startup OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.0.0 Benchmark: python_startup RHEL 8.6 RHEL 9.0 tsc=nowatchdog 3 6 9 12 15 SE +/- 0.00, N = 3 SE +/- 0.03, N = 3 12.3 10.1
Stress-NG Test: SENDFILE OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.14 Test: SENDFILE RHEL 8.6 RHEL 9.0 tsc=nowatchdog 400K 800K 1200K 1600K 2000K SE +/- 12268.55, N = 3 SE +/- 4051.51, N = 3 1665050.95 2017286.05 1. (CC) gcc options: -O2 -std=gnu99 -lm -lc -lcrypt -ldl -ljpeg -lrt -lsctp -lz -pthread
Stress-NG Test: MMAP OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.14 Test: MMAP RHEL 8.6 RHEL 9.0 tsc=nowatchdog 1000 2000 3000 4000 5000 SE +/- 29.68, N = 3 SE +/- 18.66, N = 3 3907.79 4727.15 1. (CC) gcc options: -O2 -std=gnu99 -lm -lc -lcrypt -ldl -ljpeg -lrt -lsctp -lz -pthread
Stress-NG Test: Matrix Math OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.14 Test: Matrix Math RHEL 8.6 RHEL 9.0 tsc=nowatchdog 100K 200K 300K 400K 500K SE +/- 572.97, N = 3 SE +/- 1240.42, N = 3 465746.50 387212.68 1. (CC) gcc options: -O2 -std=gnu99 -lm -lc -lcrypt -ldl -ljpeg -lrt -lsctp -lz -pthread
Mobile Neural Network Model: MobileNetV2_224 OpenBenchmarking.org ms, Fewer Is Better Mobile Neural Network 1.2 Model: MobileNetV2_224 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 1.0307 2.0614 3.0921 4.1228 5.1535 SE +/- 0.067, N = 9 SE +/- 0.043, N = 12 4.581 3.821 MIN: 3.74 / MAX: 15.52 MIN: 3.28 / MAX: 10.26 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
Stress-NG Test: CPU Stress OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.14 Test: CPU Stress RHEL 8.6 RHEL 9.0 tsc=nowatchdog 60K 120K 180K 240K 300K SE +/- 403.46, N = 3 SE +/- 361.39, N = 3 216996.46 258331.25 1. (CC) gcc options: -O2 -std=gnu99 -lm -lc -lcrypt -ldl -ljpeg -lrt -lsctp -lz -pthread
Zstd Compression Compression Level: 3, Long Mode - Compression Speed OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.0 Compression Level: 3, Long Mode - Compression Speed RHEL 8.6 RHEL 9.0 tsc=nowatchdog 150 300 450 600 750 SE +/- 3.09, N = 3 SE +/- 7.09, N = 15 583.7 686.5 1. (CC) gcc options: -O3 -pthread -lz -llzma
Mobile Neural Network Model: inception-v3 OpenBenchmarking.org ms, Fewer Is Better Mobile Neural Network 1.2 Model: inception-v3 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 7 14 21 28 35 SE +/- 0.29, N = 9 SE +/- 0.29, N = 12 31.88 27.27 MIN: 27.64 / MAX: 165.4 MIN: 23.09 / MAX: 96.23 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
DaCapo Benchmark Java Test: H2 OpenBenchmarking.org msec, Fewer Is Better DaCapo Benchmark 9.12-MR1 Java Test: H2 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 900 1800 2700 3600 4500 SE +/- 46.50, N = 20 SE +/- 33.95, N = 20 4230 3656
JPEG XL Decoding libjxl CPU Threads: All OpenBenchmarking.org MP/s, More Is Better JPEG XL Decoding libjxl 0.6.1 CPU Threads: All RHEL 8.6 RHEL 9.0 tsc=nowatchdog 80 160 240 320 400 SE +/- 2.78, N = 7 SE +/- 2.79, N = 9 301.07 347.44
Timed MPlayer Compilation Time To Compile OpenBenchmarking.org Seconds, Fewer Is Better Timed MPlayer Compilation 1.5 Time To Compile RHEL 8.6 RHEL 9.0 tsc=nowatchdog 2 4 6 8 10 SE +/- 0.012, N = 3 SE +/- 0.036, N = 3 7.395 8.506
nginx Concurrent Requests: 500 OpenBenchmarking.org Requests Per Second, More Is Better nginx 1.21.1 Concurrent Requests: 500 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 20K 40K 60K 80K 100K SE +/- 1025.05, N = 3 SE +/- 25.79, N = 3 97314.91 84779.30 -ldl -lpthread 1. (CC) gcc options: -lcrypt -lz -O3 -march=native
PyPerformance Benchmark: nbody OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.0.0 Benchmark: nbody RHEL 8.6 RHEL 9.0 tsc=nowatchdog 40 80 120 160 200 SE +/- 0.33, N = 3 SE +/- 0.00, N = 3 143 164
Build2 Time To Compile OpenBenchmarking.org Seconds, Fewer Is Better Build2 0.13 Time To Compile RHEL 8.6 RHEL 9.0 tsc=nowatchdog 13 26 39 52 65 SE +/- 0.12, N = 3 SE +/- 0.23, N = 3 49.68 56.72
nginx Concurrent Requests: 1000 OpenBenchmarking.org Requests Per Second, More Is Better nginx 1.21.1 Concurrent Requests: 1000 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 20K 40K 60K 80K 100K SE +/- 144.70, N = 3 SE +/- 106.41, N = 3 97374.02 85299.53 -ldl -lpthread 1. (CC) gcc options: -lcrypt -lz -O3 -march=native
PHPBench PHP Benchmark Suite OpenBenchmarking.org Score, More Is Better PHPBench 0.8.1 PHP Benchmark Suite RHEL 8.6 RHEL 9.0 tsc=nowatchdog 130K 260K 390K 520K 650K SE +/- 697.48, N = 3 SE +/- 900.20, N = 3 525005 597949
nginx Concurrent Requests: 200 OpenBenchmarking.org Requests Per Second, More Is Better nginx 1.21.1 Concurrent Requests: 200 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 20K 40K 60K 80K 100K SE +/- 266.10, N = 3 SE +/- 409.39, N = 3 97400.21 85903.76 -ldl -lpthread 1. (CC) gcc options: -lcrypt -lz -O3 -march=native
Apache HTTP Server Concurrent Requests: 500 OpenBenchmarking.org Requests Per Second, More Is Better Apache HTTP Server 2.4.48 Concurrent Requests: 500 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 20K 40K 60K 80K 100K SE +/- 1069.05, N = 3 SE +/- 730.64, N = 3 111137.30 98091.14 -pthread 1. (CC) gcc options: -shared -fPIC -O2
srsRAN Test: OFDM_Test OpenBenchmarking.org Samples / Second, More Is Better srsRAN 21.10 Test: OFDM_Test RHEL 8.6 RHEL 9.0 tsc=nowatchdog 30M 60M 90M 120M 150M SE +/- 384418.75, N = 3 SE +/- 1123981.02, N = 3 108966667 123000000 1. (CXX) g++ options: -std=c++11 -fno-strict-aliasing -march=native -mfpmath=sse -mavx2 -fvisibility=hidden -O3 -fno-trapping-math -fno-math-errno -rdynamic -lpthread -lmbedcrypto -lconfig++ -lsctp -lm -lfftw3f
PyPerformance Benchmark: crypto_pyaes OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.0.0 Benchmark: crypto_pyaes RHEL 8.6 RHEL 9.0 tsc=nowatchdog 30 60 90 120 150 SE +/- 0.00, N = 3 116 130
LuaRadio Test: Five Back to Back FIR Filters OpenBenchmarking.org MiB/s, More Is Better LuaRadio 0.9.1 Test: Five Back to Back FIR Filters RHEL 8.6 RHEL 9.0 tsc=nowatchdog 100 200 300 400 500 SE +/- 3.83, N = 3 SE +/- 3.82, N = 3 461.1 415.4
Timed Apache Compilation Time To Compile OpenBenchmarking.org Seconds, Fewer Is Better Timed Apache Compilation 2.4.41 Time To Compile RHEL 8.6 RHEL 9.0 tsc=nowatchdog 5 10 15 20 25 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 19.88 22.06
Timed LLVM Compilation Build System: Unix Makefiles OpenBenchmarking.org Seconds, Fewer Is Better Timed LLVM Compilation 13.0 Build System: Unix Makefiles RHEL 8.6 RHEL 9.0 tsc=nowatchdog 40 80 120 160 200 SE +/- 0.19, N = 3 SE +/- 0.24, N = 3 162.68 180.46
PyPerformance Benchmark: pickle_pure_python OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.0.0 Benchmark: pickle_pure_python RHEL 8.6 RHEL 9.0 tsc=nowatchdog 130 260 390 520 650 SE +/- 1.15, N = 3 SE +/- 0.58, N = 3 587 532
srsRAN Test: 5G PHY_DL_NR Test 52 PRB SISO 64-QAM OpenBenchmarking.org UE Mb/s, More Is Better srsRAN 21.10 Test: 5G PHY_DL_NR Test 52 PRB SISO 64-QAM RHEL 8.6 RHEL 9.0 tsc=nowatchdog 12 24 36 48 60 SE +/- 0.06, N = 3 SE +/- 0.03, N = 3 49.9 54.9 1. (CXX) g++ options: -std=c++11 -fno-strict-aliasing -march=native -mfpmath=sse -mavx2 -fvisibility=hidden -O3 -fno-trapping-math -fno-math-errno -rdynamic -lpthread -lmbedcrypto -lconfig++ -lsctp -lm -lfftw3f
PyPerformance Benchmark: regex_compile OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.0.0 Benchmark: regex_compile RHEL 8.6 RHEL 9.0 tsc=nowatchdog 50 100 150 200 250 SE +/- 0.00, N = 3 225 205
PyPerformance Benchmark: pathlib OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.0.0 Benchmark: pathlib RHEL 8.6 RHEL 9.0 tsc=nowatchdog 5 10 15 20 25 SE +/- 0.03, N = 3 SE +/- 0.03, N = 3 21.4 19.5
PyPerformance Benchmark: json_loads OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.0.0 Benchmark: json_loads RHEL 8.6 RHEL 9.0 tsc=nowatchdog 7 14 21 28 35 SE +/- 0.03, N = 3 SE +/- 0.03, N = 3 27.7 25.4
Radiance Benchmark Test: SMP Parallel OpenBenchmarking.org Seconds, Fewer Is Better Radiance Benchmark 5.0 Test: SMP Parallel RHEL 8.6 RHEL 9.0 tsc=nowatchdog 50 100 150 200 250 207.25 190.26
Apache HTTP Server Concurrent Requests: 200 OpenBenchmarking.org Requests Per Second, More Is Better Apache HTTP Server 2.4.48 Concurrent Requests: 200 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 30K 60K 90K 120K 150K SE +/- 1190.10, N = 3 SE +/- 1185.64, N = 3 121914.60 112402.32 -pthread 1. (CC) gcc options: -shared -fPIC -O2
Coremark CoreMark Size 666 - Iterations Per Second OpenBenchmarking.org Iterations/Sec, More Is Better Coremark 1.0 CoreMark Size 666 - Iterations Per Second RHEL 8.6 RHEL 9.0 tsc=nowatchdog 1000K 2000K 3000K 4000K 5000K SE +/- 3596.66, N = 3 SE +/- 3088.69, N = 3 4180906.33 4531044.49 1. (CC) gcc options: -O2 -lrt" -lrt
srsRAN Test: 4G PHY_DL_Test 100 PRB SISO 64-QAM OpenBenchmarking.org UE Mb/s, More Is Better srsRAN 21.10 Test: 4G PHY_DL_Test 100 PRB SISO 64-QAM RHEL 8.6 RHEL 9.0 tsc=nowatchdog 30 60 90 120 150 SE +/- 0.39, N = 3 SE +/- 0.27, N = 3 117.5 127.0 1. (CXX) g++ options: -std=c++11 -fno-strict-aliasing -march=native -mfpmath=sse -mavx2 -fvisibility=hidden -O3 -fno-trapping-math -fno-math-errno -rdynamic -lpthread -lmbedcrypto -lconfig++ -lsctp -lm -lfftw3f
Dolfyn Computational Fluid Dynamics OpenBenchmarking.org Seconds, Fewer Is Better Dolfyn 0.527 Computational Fluid Dynamics RHEL 8.6 RHEL 9.0 tsc=nowatchdog 5 10 15 20 25 SE +/- 0.02, N = 3 SE +/- 0.09, N = 3 20.68 19.19
ASKAP Test: tConvolve MPI - Gridding OpenBenchmarking.org Mpix/sec, More Is Better ASKAP 1.0 Test: tConvolve MPI - Gridding RHEL 8.6 RHEL 9.0 tsc=nowatchdog 15K 30K 45K 60K 75K SE +/- 0.00, N = 3 SE +/- 408.97, N = 3 66640.2 71567.1 1. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp
srsRAN Test: 4G PHY_DL_Test 100 PRB SISO 256-QAM OpenBenchmarking.org UE Mb/s, More Is Better srsRAN 21.10 Test: 4G PHY_DL_Test 100 PRB SISO 256-QAM RHEL 8.6 RHEL 9.0 tsc=nowatchdog 30 60 90 120 150 SE +/- 0.30, N = 3 SE +/- 0.23, N = 3 122.8 130.9 1. (CXX) g++ options: -std=c++11 -fno-strict-aliasing -march=native -mfpmath=sse -mavx2 -fvisibility=hidden -O3 -fno-trapping-math -fno-math-errno -rdynamic -lpthread -lmbedcrypto -lconfig++ -lsctp -lm -lfftw3f
libjpeg-turbo tjbench Test: Decompression Throughput OpenBenchmarking.org Megapixels/sec, More Is Better libjpeg-turbo tjbench 2.1.0 Test: Decompression Throughput RHEL 8.6 RHEL 9.0 tsc=nowatchdog 40 80 120 160 200 SE +/- 0.13, N = 3 SE +/- 0.25, N = 3 189.45 177.85 1. (CC) gcc options: -O3 -rdynamic
Etcpak Configuration: ETC1 OpenBenchmarking.org Mpx/s, More Is Better Etcpak 0.7 Configuration: ETC1 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 60 120 180 240 300 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 243.88 259.31 1. (CXX) g++ options: -O3 -march=native -std=c++11 -lpthread
Java JMH Throughput OpenBenchmarking.org Ops/s, More Is Better Java JMH Throughput RHEL 8.6 RHEL 9.0 tsc=nowatchdog 40000M 80000M 120000M 160000M 200000M 166378871533.27 156642355939.16
Timed LLVM Compilation Build System: Ninja OpenBenchmarking.org Seconds, Fewer Is Better Timed LLVM Compilation 13.0 Build System: Ninja RHEL 8.6 RHEL 9.0 tsc=nowatchdog 20 40 60 80 100 SE +/- 1.27, N = 3 SE +/- 0.25, N = 3 91.37 97.02
Zstd Compression Compression Level: 3, Long Mode - Decompression Speed OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.0 Compression Level: 3, Long Mode - Decompression Speed RHEL 8.6 RHEL 9.0 tsc=nowatchdog 800 1600 2400 3200 4000 SE +/- 11.93, N = 3 SE +/- 2.63, N = 15 3663.8 3885.9 1. (CC) gcc options: -O3 -pthread -lz -llzma
Pennant Test: sedovbig OpenBenchmarking.org Hydro Cycle Time - Seconds, Fewer Is Better Pennant 1.0.1 Test: sedovbig RHEL 8.6 RHEL 9.0 tsc=nowatchdog 1.2955 2.591 3.8865 5.182 6.4775 SE +/- 0.058271, N = 3 SE +/- 0.055204, N = 15 5.757579 5.429139 -pthread 1. (CXX) g++ options: -fopenmp -lmpi_cxx -lmpi
simdjson Throughput Test: LargeRandom OpenBenchmarking.org GB/s, More Is Better simdjson 1.0 Throughput Test: LargeRandom RHEL 8.6 RHEL 9.0 tsc=nowatchdog 0.198 0.396 0.594 0.792 0.99 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 0.83 0.88 -pthread 1. (CXX) g++ options: -O3
Timed PHP Compilation Time To Compile OpenBenchmarking.org Seconds, Fewer Is Better Timed PHP Compilation 7.4.2 Time To Compile RHEL 8.6 RHEL 9.0 tsc=nowatchdog 8 16 24 32 40 SE +/- 0.16, N = 3 SE +/- 0.07, N = 3 34.19 36.24
srsRAN Test: 4G PHY_DL_Test 100 PRB MIMO 256-QAM OpenBenchmarking.org UE Mb/s, More Is Better srsRAN 21.10 Test: 4G PHY_DL_Test 100 PRB MIMO 256-QAM RHEL 8.6 RHEL 9.0 tsc=nowatchdog 30 60 90 120 150 SE +/- 0.13, N = 3 SE +/- 0.15, N = 3 116.9 123.9 1. (CXX) g++ options: -std=c++11 -fno-strict-aliasing -march=native -mfpmath=sse -mavx2 -fvisibility=hidden -O3 -fno-trapping-math -fno-math-errno -rdynamic -lpthread -lmbedcrypto -lconfig++ -lsctp -lm -lfftw3f
x265 Video Input: Bosphorus 4K OpenBenchmarking.org Frames Per Second, More Is Better x265 3.4 Video Input: Bosphorus 4K RHEL 8.6 RHEL 9.0 tsc=nowatchdog 6 12 18 24 30 SE +/- 0.21, N = 15 SE +/- 0.29, N = 15 23.86 22.54 1. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl
SVT-VP9 Tuning: VMAF Optimized - Input: Bosphorus 1080p OpenBenchmarking.org Frames Per Second, More Is Better SVT-VP9 0.3 Tuning: VMAF Optimized - Input: Bosphorus 1080p RHEL 8.6 RHEL 9.0 tsc=nowatchdog 90 180 270 360 450 SE +/- 2.83, N = 15 SE +/- 4.24, N = 5 387.22 409.83 1. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm
ONNX Runtime Model: ArcFace ResNet-100 - Device: CPU - Executor: Standard OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.11 Model: ArcFace ResNet-100 - Device: CPU - Executor: Standard RHEL 8.6 RHEL 9.0 tsc=nowatchdog 200 400 600 800 1000 SE +/- 11.33, N = 5 SE +/- 9.94, N = 12 1081 1143 -pthread -lpthread 1. (CXX) g++ options: -ffunction-sections -fdata-sections -march=native -mtune=native -O3 -flto -fno-fat-lto-objects -ldl -lrt
srsRAN Test: 4G PHY_DL_Test 100 PRB MIMO 64-QAM OpenBenchmarking.org eNb Mb/s, More Is Better srsRAN 21.10 Test: 4G PHY_DL_Test 100 PRB MIMO 64-QAM RHEL 8.6 RHEL 9.0 tsc=nowatchdog 80 160 240 320 400 SE +/- 0.98, N = 3 SE +/- 0.50, N = 3 327.4 345.8 1. (CXX) g++ options: -std=c++11 -fno-strict-aliasing -march=native -mfpmath=sse -mavx2 -fvisibility=hidden -O3 -fno-trapping-math -fno-math-errno -rdynamic -lpthread -lmbedcrypto -lconfig++ -lsctp -lm -lfftw3f
Stress-NG Test: Glibc Qsort Data Sorting OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.14 Test: Glibc Qsort Data Sorting RHEL 8.6 RHEL 9.0 tsc=nowatchdog 300 600 900 1200 1500 SE +/- 4.51, N = 3 SE +/- 4.24, N = 3 1356.46 1285.16 1. (CC) gcc options: -O2 -std=gnu99 -lm -lc -lcrypt -ldl -ljpeg -lrt -lsctp -lz -pthread
Zstd Compression Compression Level: 19, Long Mode - Decompression Speed OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.0 Compression Level: 19, Long Mode - Decompression Speed RHEL 8.6 RHEL 9.0 tsc=nowatchdog 700 1400 2100 2800 3500 SE +/- 3.30, N = 3 SE +/- 2.48, N = 3 3155.0 3329.7 1. (CC) gcc options: -O3 -pthread -lz -llzma
srsRAN Test: 4G PHY_DL_Test 100 PRB SISO 64-QAM OpenBenchmarking.org eNb Mb/s, More Is Better srsRAN 21.10 Test: 4G PHY_DL_Test 100 PRB SISO 64-QAM RHEL 8.6 RHEL 9.0 tsc=nowatchdog 80 160 240 320 400 SE +/- 0.72, N = 3 SE +/- 0.53, N = 3 329.7 347.5 1. (CXX) g++ options: -std=c++11 -fno-strict-aliasing -march=native -mfpmath=sse -mavx2 -fvisibility=hidden -O3 -fno-trapping-math -fno-math-errno -rdynamic -lpthread -lmbedcrypto -lconfig++ -lsctp -lm -lfftw3f
Timed Node.js Compilation Time To Compile OpenBenchmarking.org Seconds, Fewer Is Better Timed Node.js Compilation 17.3 Time To Compile RHEL 8.6 RHEL 9.0 tsc=nowatchdog 20 40 60 80 100 SE +/- 0.29, N = 3 SE +/- 0.15, N = 3 96.03 101.11
srsRAN Test: 4G PHY_DL_Test 100 PRB MIMO 64-QAM OpenBenchmarking.org UE Mb/s, More Is Better srsRAN 21.10 Test: 4G PHY_DL_Test 100 PRB MIMO 64-QAM RHEL 8.6 RHEL 9.0 tsc=nowatchdog 30 60 90 120 150 SE +/- 0.18, N = 3 SE +/- 0.52, N = 3 110.7 116.4 1. (CXX) g++ options: -std=c++11 -fno-strict-aliasing -march=native -mfpmath=sse -mavx2 -fvisibility=hidden -O3 -fno-trapping-math -fno-math-errno -rdynamic -lpthread -lmbedcrypto -lconfig++ -lsctp -lm -lfftw3f
Stress-NG Test: Memory Copying OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.14 Test: Memory Copying RHEL 8.6 RHEL 9.0 tsc=nowatchdog 3K 6K 9K 12K 15K SE +/- 47.48, N = 3 SE +/- 110.20, N = 3 12203.20 12831.44 1. (CC) gcc options: -O2 -std=gnu99 -lm -lc -lcrypt -ldl -ljpeg -lrt -lsctp -lz -pthread
PyPerformance Benchmark: 2to3 OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.0.0 Benchmark: 2to3 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 90 180 270 360 450 SE +/- 0.33, N = 3 393 374
srsRAN Test: 4G PHY_DL_Test 100 PRB MIMO 256-QAM OpenBenchmarking.org eNb Mb/s, More Is Better srsRAN 21.10 Test: 4G PHY_DL_Test 100 PRB MIMO 256-QAM RHEL 8.6 RHEL 9.0 tsc=nowatchdog 80 160 240 320 400 SE +/- 0.35, N = 3 SE +/- 0.17, N = 3 361.7 378.5 1. (CXX) g++ options: -std=c++11 -fno-strict-aliasing -march=native -mfpmath=sse -mavx2 -fvisibility=hidden -O3 -fno-trapping-math -fno-math-errno -rdynamic -lpthread -lmbedcrypto -lconfig++ -lsctp -lm -lfftw3f
JPEG XL Decoding libjxl CPU Threads: 1 OpenBenchmarking.org MP/s, More Is Better JPEG XL Decoding libjxl 0.6.1 CPU Threads: 1 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 12 24 36 48 60 SE +/- 0.06, N = 3 SE +/- 0.06, N = 3 50.78 53.13
Zstd Compression Compression Level: 19 - Decompression Speed OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.0 Compression Level: 19 - Decompression Speed RHEL 8.6 RHEL 9.0 tsc=nowatchdog 700 1400 2100 2800 3500 SE +/- 12.54, N = 3 SE +/- 1.23, N = 3 3099.0 3240.2 1. (CC) gcc options: -O3 -pthread -lz -llzma
7-Zip Compression Test: Decompression Rating OpenBenchmarking.org MIPS, More Is Better 7-Zip Compression 21.06 Test: Decompression Rating RHEL 8.6 RHEL 9.0 tsc=nowatchdog 160K 320K 480K 640K 800K SE +/- 3414.73, N = 15 SE +/- 5726.36, N = 5 721671 754378 1. (CXX) g++ options: -lpthread -ldl -O2 -fPIC
Xcompact3d Incompact3d Input: input.i3d 193 Cells Per Direction OpenBenchmarking.org Seconds, Fewer Is Better Xcompact3d Incompact3d 2021-03-11 Input: input.i3d 193 Cells Per Direction RHEL 8.6 RHEL 9.0 tsc=nowatchdog 3 6 9 12 15 SE +/- 0.03, N = 3 SE +/- 0.06, N = 3 10.16 10.61 -pthread 1. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -lmpi_usempif08 -lmpi_mpifh -lmpi
srsRAN Test: 4G PHY_DL_Test 100 PRB SISO 256-QAM OpenBenchmarking.org eNb Mb/s, More Is Better srsRAN 21.10 Test: 4G PHY_DL_Test 100 PRB SISO 256-QAM RHEL 8.6 RHEL 9.0 tsc=nowatchdog 80 160 240 320 400 SE +/- 0.59, N = 3 SE +/- 1.07, N = 3 360.8 375.0 1. (CXX) g++ options: -std=c++11 -fno-strict-aliasing -march=native -mfpmath=sse -mavx2 -fvisibility=hidden -O3 -fno-trapping-math -fno-math-errno -rdynamic -lpthread -lmbedcrypto -lconfig++ -lsctp -lm -lfftw3f
Stress-NG Test: Semaphores OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.14 Test: Semaphores RHEL 8.6 RHEL 9.0 tsc=nowatchdog 4M 8M 12M 16M 20M SE +/- 24969.77, N = 3 SE +/- 230054.48, N = 3 18815432.41 18135356.63 1. (CC) gcc options: -O2 -std=gnu99 -lm -lc -lcrypt -ldl -ljpeg -lrt -lsctp -lz -pthread
Graph500 Scale: 26 OpenBenchmarking.org sssp median_TEPS, More Is Better Graph500 3.0 Scale: 26 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 80M 160M 240M 320M 400M 360311000 373684000 -pthread 1. (CC) gcc options: -fcommon -O3 -lpthread -lm -lmpi
Etcpak Configuration: DXT1 OpenBenchmarking.org Mpx/s, More Is Better Etcpak 0.7 Configuration: DXT1 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 200 400 600 800 1000 SE +/- 0.43, N = 3 SE +/- 8.56, N = 3 965.99 1001.43 1. (CXX) g++ options: -O3 -march=native -std=c++11 -lpthread
Ngspice Circuit: C7552 OpenBenchmarking.org Seconds, Fewer Is Better Ngspice 34 Circuit: C7552 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 30 60 90 120 150 SE +/- 1.44, N = 3 SE +/- 0.14, N = 3 112.60 108.69 1. (CC) gcc options: -O0 -fopenmp -lm -lstdc++ -lfftw3 -lXaw -lXmu -lXt -lXext -lX11 -lSM -lICE
Timed GDB GNU Debugger Compilation Time To Compile OpenBenchmarking.org Seconds, Fewer Is Better Timed GDB GNU Debugger Compilation 10.2 Time To Compile RHEL 8.6 RHEL 9.0 tsc=nowatchdog 12 24 36 48 60 SE +/- 0.39, N = 3 SE +/- 0.56, N = 5 51.77 53.58
JPEG XL libjxl Input: PNG - Encode Speed: 8 OpenBenchmarking.org MP/s, More Is Better JPEG XL libjxl 0.6.1 Input: PNG - Encode Speed: 8 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 0.207 0.414 0.621 0.828 1.035 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 0.89 0.92 -pthread 1. (CXX) g++ options: -funwind-tables -O3 -O2 -fPIE -pie
PyPerformance Benchmark: go OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.0.0 Benchmark: go RHEL 8.6 RHEL 9.0 tsc=nowatchdog 70 140 210 280 350 SE +/- 0.58, N = 3 316 306
Apache HTTP Server Concurrent Requests: 1000 OpenBenchmarking.org Requests Per Second, More Is Better Apache HTTP Server 2.4.48 Concurrent Requests: 1000 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 20K 40K 60K 80K 100K SE +/- 942.90, N = 6 SE +/- 1177.39, N = 3 101529.21 104832.79 -pthread 1. (CC) gcc options: -shared -fPIC -O2
ASTC Encoder Preset: Medium OpenBenchmarking.org Seconds, Fewer Is Better ASTC Encoder 3.2 Preset: Medium RHEL 8.6 RHEL 9.0 tsc=nowatchdog 0.9432 1.8864 2.8296 3.7728 4.716 SE +/- 0.0078, N = 3 SE +/- 0.0170, N = 3 4.0684 4.1918 1. (CXX) g++ options: -O3 -flto -pthread
ASTC Encoder Preset: Thorough OpenBenchmarking.org Seconds, Fewer Is Better ASTC Encoder 3.2 Preset: Thorough RHEL 8.6 RHEL 9.0 tsc=nowatchdog 1.2386 2.4772 3.7158 4.9544 6.193 SE +/- 0.0042, N = 3 SE +/- 0.0070, N = 3 5.3483 5.5049 1. (CXX) g++ options: -O3 -flto -pthread
PyPerformance Benchmark: chaos OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.0.0 Benchmark: chaos RHEL 8.6 RHEL 9.0 tsc=nowatchdog 30 60 90 120 150 SE +/- 0.33, N = 3 138 142
Appleseed Scene: Material Tester OpenBenchmarking.org Seconds, Fewer Is Better Appleseed 2.0 Beta Scene: Material Tester RHEL 8.6 RHEL 9.0 tsc=nowatchdog 70 140 210 280 350 340.50 330.97
Numpy Benchmark OpenBenchmarking.org Score, More Is Better Numpy Benchmark RHEL 8.6 RHEL 9.0 tsc=nowatchdog 100 200 300 400 500 SE +/- 0.57, N = 3 SE +/- 1.56, N = 3 431.91 444.26
Stress-NG Test: Glibc C String Functions OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.14 Test: Glibc C String Functions RHEL 8.6 RHEL 9.0 tsc=nowatchdog 3M 6M 9M 12M 15M SE +/- 93884.01, N = 3 SE +/- 84803.96, N = 3 13787576.86 14151608.76 1. (CC) gcc options: -O2 -std=gnu99 -lm -lc -lcrypt -ldl -ljpeg -lrt -lsctp -lz -pthread
Renaissance Test: ALS Movie Lens OpenBenchmarking.org ms, Fewer Is Better Renaissance 0.14 Test: ALS Movie Lens RHEL 8.6 RHEL 9.0 tsc=nowatchdog 2K 4K 6K 8K 10K SE +/- 43.47, N = 3 SE +/- 120.52, N = 4 9758.4 10002.0 MIN: 9580.74 / MAX: 10422.67 MIN: 9086.74 / MAX: 11113.36
DaCapo Benchmark Java Test: Jython OpenBenchmarking.org msec, Fewer Is Better DaCapo Benchmark 9.12-MR1 Java Test: Jython RHEL 8.6 RHEL 9.0 tsc=nowatchdog 900 1800 2700 3600 4500 SE +/- 39.75, N = 4 SE +/- 36.84, N = 4 4316 4420
Graph500 Scale: 26 OpenBenchmarking.org bfs median_TEPS, More Is Better Graph500 3.0 Scale: 26 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 200M 400M 600M 800M 1000M 952659000 973443000 -pthread 1. (CC) gcc options: -fcommon -O3 -lpthread -lm -lmpi
Etcpak Configuration: ETC1 + Dithering OpenBenchmarking.org Mpx/s, More Is Better Etcpak 0.7 Configuration: ETC1 + Dithering RHEL 8.6 RHEL 9.0 tsc=nowatchdog 50 100 150 200 250 SE +/- 0.08, N = 3 SE +/- 0.10, N = 3 225.76 230.59 1. (CXX) g++ options: -O3 -march=native -std=c++11 -lpthread
MariaDB Clients: 256 OpenBenchmarking.org Queries Per Second, More Is Better MariaDB 10.8.2 Clients: 256 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 80 160 240 320 400 SE +/- 3.62, N = 3 SE +/- 1.39, N = 3 387 379 -pthread 1. (CXX) g++ options: -pie -fPIC -fstack-protector -O3 -lpcre2-8 -lcrypt -lz -lm -lssl -lcrypto -lpthread -ldl
Etcpak Configuration: ETC2 OpenBenchmarking.org Mpx/s, More Is Better Etcpak 0.7 Configuration: ETC2 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 40 80 120 160 200 SE +/- 0.14, N = 3 SE +/- 0.01, N = 3 162.67 159.46 1. (CXX) g++ options: -O3 -march=native -std=c++11 -lpthread
Stress-NG Test: Forking OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.14 Test: Forking RHEL 8.6 RHEL 9.0 tsc=nowatchdog 7K 14K 21K 28K 35K SE +/- 49.04, N = 3 SE +/- 259.01, N = 3 33665.32 34328.25 1. (CC) gcc options: -O2 -std=gnu99 -lm -lc -lcrypt -ldl -ljpeg -lrt -lsctp -lz -pthread
ASKAP Test: tConvolve OpenMP - Gridding OpenBenchmarking.org Million Grid Points Per Second, More Is Better ASKAP 1.0 Test: tConvolve OpenMP - Gridding RHEL 8.6 RHEL 9.0 tsc=nowatchdog 2K 4K 6K 8K 10K SE +/- 108.24, N = 15 SE +/- 147.93, N = 3 10750.9 10946.1 1. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp
Appleseed Scene: Emily OpenBenchmarking.org Seconds, Fewer Is Better Appleseed 2.0 Beta Scene: Emily RHEL 8.6 RHEL 9.0 tsc=nowatchdog 40 80 120 160 200 166.81 163.84
Graph500 Scale: 26 OpenBenchmarking.org sssp max_TEPS, More Is Better Graph500 3.0 Scale: 26 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 110M 220M 330M 440M 550M 492999000 501887000 -pthread 1. (CC) gcc options: -fcommon -O3 -lpthread -lm -lmpi
asmFish 1024 Hash Memory, 26 Depth OpenBenchmarking.org Nodes/second, More Is Better asmFish 2018-07-23 1024 Hash Memory, 26 Depth RHEL 8.6 RHEL 9.0 tsc=nowatchdog 60M 120M 180M 240M 300M SE +/- 2264642.00, N = 10 SE +/- 3888047.47, N = 3 295354804 290367032
m-queens Time To Solve OpenBenchmarking.org Seconds, Fewer Is Better m-queens 1.2 Time To Solve RHEL 8.6 RHEL 9.0 tsc=nowatchdog 1.28 2.56 3.84 5.12 6.4 SE +/- 0.027, N = 3 SE +/- 0.023, N = 3 5.689 5.595 1. (CXX) g++ options: -fopenmp -O2 -march=native
ASKAP Test: tConvolve MPI - Degridding OpenBenchmarking.org Mpix/sec, More Is Better ASKAP 1.0 Test: tConvolve MPI - Degridding RHEL 8.6 RHEL 9.0 tsc=nowatchdog 14K 28K 42K 56K 70K SE +/- 0.00, N = 3 SE +/- 556.62, N = 3 64589.7 63620.8 1. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp
SVT-VP9 Tuning: PSNR/SSIM Optimized - Input: Bosphorus 1080p OpenBenchmarking.org Frames Per Second, More Is Better SVT-VP9 0.3 Tuning: PSNR/SSIM Optimized - Input: Bosphorus 1080p RHEL 8.6 RHEL 9.0 tsc=nowatchdog 90 180 270 360 450 SE +/- 4.61, N = 15 SE +/- 3.05, N = 3 399.83 405.69 1. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm
SVT-VP9 Tuning: Visual Quality Optimized - Input: Bosphorus 1080p OpenBenchmarking.org Frames Per Second, More Is Better SVT-VP9 0.3 Tuning: Visual Quality Optimized - Input: Bosphorus 1080p RHEL 8.6 RHEL 9.0 tsc=nowatchdog 70 140 210 280 350 SE +/- 2.18, N = 3 SE +/- 2.82, N = 3 316.95 320.93 1. (CC) gcc options: -O3 -fcommon -fPIE -fPIC -fvisibility=hidden -pie -rdynamic -lpthread -lrt -lm
Xcompact3d Incompact3d Input: X3D-benchmarking input.i3d OpenBenchmarking.org Seconds, Fewer Is Better Xcompact3d Incompact3d 2021-03-11 Input: X3D-benchmarking input.i3d RHEL 8.6 RHEL 9.0 tsc=nowatchdog 70 140 210 280 350 SE +/- 2.67, N = 8 SE +/- 1.21, N = 3 306.67 303.14 -pthread 1. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -lmpi_usempif08 -lmpi_mpifh -lmpi
Crafty Elapsed Time OpenBenchmarking.org Nodes Per Second, More Is Better Crafty 25.2 Elapsed Time RHEL 8.6 RHEL 9.0 tsc=nowatchdog 2M 4M 6M 8M 10M SE +/- 13943.85, N = 3 SE +/- 18282.66, N = 3 8615936 8518882 1. (CC) gcc options: -pthread -lstdc++ -fprofile-use -lm
Renaissance Test: Random Forest OpenBenchmarking.org ms, Fewer Is Better Renaissance 0.14 Test: Random Forest RHEL 8.6 RHEL 9.0 tsc=nowatchdog 160 320 480 640 800 SE +/- 7.49, N = 3 SE +/- 7.51, N = 15 762.8 754.3 MIN: 661 / MAX: 1026.39 MIN: 664.49 / MAX: 1068.59
Blender Blend File: BMW27 - Compute: CPU-Only OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.0 Blend File: BMW27 - Compute: CPU-Only RHEL 8.6 RHEL 9.0 tsc=nowatchdog 4 8 12 16 20 SE +/- 0.13, N = 3 SE +/- 0.06, N = 3 14.98 14.82
Intel Open Image Denoise Run: RTLightmap.hdr.4096x4096 OpenBenchmarking.org Images / Sec, More Is Better Intel Open Image Denoise 1.4.0 Run: RTLightmap.hdr.4096x4096 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 0.2138 0.4276 0.6414 0.8552 1.069 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 0.95 0.94
Intel Open Image Denoise Run: RT.ldr_alb_nrm.3840x2160 OpenBenchmarking.org Images / Sec, More Is Better Intel Open Image Denoise 1.4.0 Run: RT.ldr_alb_nrm.3840x2160 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 0.4388 0.8776 1.3164 1.7552 2.194 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 1.93 1.95
Algebraic Multi-Grid Benchmark OpenBenchmarking.org Figure Of Merit, More Is Better Algebraic Multi-Grid Benchmark 1.2 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 400M 800M 1200M 1600M 2000M SE +/- 4985377.50, N = 3 SE +/- 1350756.87, N = 3 2051105333 2032026667 -pthread 1. (CC) gcc options: -lparcsr_ls -lparcsr_mv -lseq_mv -lIJ_mv -lkrylov -lHYPRE_utilities -lm -fopenmp -lmpi
OSPray Benchmark: gravity_spheres_volume/dim_512/ao/real_time OpenBenchmarking.org Items Per Second, More Is Better OSPray 2.9 Benchmark: gravity_spheres_volume/dim_512/ao/real_time RHEL 8.6 RHEL 9.0 tsc=nowatchdog 4 8 12 16 20 SE +/- 0.01, N = 3 SE +/- 0.03, N = 3 16.58 16.74
srsRAN Test: 5G PHY_DL_NR Test 52 PRB SISO 64-QAM OpenBenchmarking.org eNb Mb/s, More Is Better srsRAN 21.10 Test: 5G PHY_DL_NR Test 52 PRB SISO 64-QAM RHEL 8.6 RHEL 9.0 tsc=nowatchdog 30 60 90 120 150 SE +/- 0.12, N = 3 SE +/- 0.20, N = 3 110.9 111.9 1. (CXX) g++ options: -std=c++11 -fno-strict-aliasing -march=native -mfpmath=sse -mavx2 -fvisibility=hidden -O3 -fno-trapping-math -fno-math-errno -rdynamic -lpthread -lmbedcrypto -lconfig++ -lsctp -lm -lfftw3f
OSPray Benchmark: gravity_spheres_volume/dim_512/scivis/real_time OpenBenchmarking.org Items Per Second, More Is Better OSPray 2.9 Benchmark: gravity_spheres_volume/dim_512/scivis/real_time RHEL 8.6 RHEL 9.0 tsc=nowatchdog 4 8 12 16 20 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 15.92 16.06
SecureMark Benchmark: SecureMark-TLS OpenBenchmarking.org marks, More Is Better SecureMark 1.0.4 Benchmark: SecureMark-TLS RHEL 8.6 RHEL 9.0 tsc=nowatchdog 50K 100K 150K 200K 250K SE +/- 755.69, N = 3 SE +/- 451.57, N = 3 218474 220380 1. (CC) gcc options: -pedantic -O3
Graph500 Scale: 26 OpenBenchmarking.org bfs max_TEPS, More Is Better Graph500 3.0 Scale: 26 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 200M 400M 600M 800M 1000M 987760000 996357000 -pthread 1. (CC) gcc options: -fcommon -O3 -lpthread -lm -lmpi
Stress-NG Test: Crypto OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.14 Test: Crypto RHEL 8.6 RHEL 9.0 tsc=nowatchdog 40K 80K 120K 160K 200K SE +/- 343.26, N = 3 SE +/- 297.03, N = 3 162556.21 163933.11 1. (CC) gcc options: -O2 -std=gnu99 -lm -lc -lcrypt -ldl -ljpeg -lrt -lsctp -lz -pthread
High Performance Conjugate Gradient OpenBenchmarking.org GFLOP/s, More Is Better High Performance Conjugate Gradient 3.1 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 9 18 27 36 45 SE +/- 0.17, N = 3 SE +/- 0.05, N = 3 38.29 37.97 -pthread 1. (CXX) g++ options: -O3 -ffast-math -ftree-vectorize -lmpi_cxx -lmpi
Blender Blend File: Pabellon Barcelona - Compute: CPU-Only OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.0 Blend File: Pabellon Barcelona - Compute: CPU-Only RHEL 8.6 RHEL 9.0 tsc=nowatchdog 10 20 30 40 50 SE +/- 0.16, N = 3 SE +/- 0.14, N = 3 45.30 45.67
Appleseed Scene: Disney Material OpenBenchmarking.org Seconds, Fewer Is Better Appleseed 2.0 Beta Scene: Disney Material RHEL 8.6 RHEL 9.0 tsc=nowatchdog 11 22 33 44 55 47.23 47.61
Stress-NG Test: Malloc OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.14 Test: Malloc RHEL 8.6 RHEL 9.0 tsc=nowatchdog 130M 260M 390M 520M 650M SE +/- 334851.99, N = 3 SE +/- 1538588.29, N = 3 601718674.55 596924219.93 1. (CC) gcc options: -O2 -std=gnu99 -lm -lc -lcrypt -ldl -ljpeg -lrt -lsctp -lz -pthread
OSPray Benchmark: particle_volume/ao/real_time OpenBenchmarking.org Items Per Second, More Is Better OSPray 2.9 Benchmark: particle_volume/ao/real_time RHEL 8.6 RHEL 9.0 tsc=nowatchdog 30 60 90 120 150 SE +/- 0.51, N = 3 SE +/- 0.43, N = 3 128.06 129.07
OSPray Benchmark: particle_volume/scivis/real_time OpenBenchmarking.org Items Per Second, More Is Better OSPray 2.9 Benchmark: particle_volume/scivis/real_time RHEL 8.6 RHEL 9.0 tsc=nowatchdog 30 60 90 120 150 SE +/- 0.21, N = 3 SE +/- 0.74, N = 3 123.18 124.14
Liquid-DSP Threads: 64 - Buffer Length: 256 - Filter Length: 57 OpenBenchmarking.org samples/s, More Is Better Liquid-DSP 2021.01.31 Threads: 64 - Buffer Length: 256 - Filter Length: 57 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 700M 1400M 2100M 2800M 3500M SE +/- 22742544.57, N = 3 SE +/- 12470899.46, N = 3 3054800000 3075900000 1. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
SQLite Speedtest Timed Time - Size 1,000 OpenBenchmarking.org Seconds, Fewer Is Better SQLite Speedtest 3.30 Timed Time - Size 1,000 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 14 28 42 56 70 SE +/- 0.25, N = 3 SE +/- 0.03, N = 3 62.68 63.06 -ldl -lpthread 1. (CC) gcc options: -O2 -lz
Ngspice Circuit: C2670 OpenBenchmarking.org Seconds, Fewer Is Better Ngspice 34 Circuit: C2670 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 30 60 90 120 150 SE +/- 0.37, N = 3 SE +/- 0.45, N = 3 147.73 146.87 1. (CC) gcc options: -O0 -fopenmp -lm -lstdc++ -lfftw3 -lXaw -lXmu -lXt -lXext -lX11 -lSM -lICE
JPEG XL libjxl Input: JPEG - Encode Speed: 8 OpenBenchmarking.org MP/s, More Is Better JPEG XL libjxl 0.6.1 Input: JPEG - Encode Speed: 8 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 6 12 18 24 30 SE +/- 0.04, N = 3 SE +/- 0.07, N = 3 26.56 26.71 -pthread 1. (CXX) g++ options: -funwind-tables -O3 -O2 -fPIE -pie
ASTC Encoder Preset: Exhaustive OpenBenchmarking.org Seconds, Fewer Is Better ASTC Encoder 3.2 Preset: Exhaustive RHEL 8.6 RHEL 9.0 tsc=nowatchdog 1.3463 2.6926 4.0389 5.3852 6.7315 SE +/- 0.0055, N = 3 SE +/- 0.0186, N = 3 5.9835 5.9508 1. (CXX) g++ options: -O3 -flto -pthread
Intel Open Image Denoise Run: RT.hdr_alb_nrm.3840x2160 OpenBenchmarking.org Images / Sec, More Is Better Intel Open Image Denoise 1.4.0 Run: RT.hdr_alb_nrm.3840x2160 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 0.4343 0.8686 1.3029 1.7372 2.1715 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 1.92 1.93
Timed Linux Kernel Compilation Build: defconfig OpenBenchmarking.org Seconds, Fewer Is Better Timed Linux Kernel Compilation 5.16 Build: defconfig RHEL 8.6 RHEL 9.0 tsc=nowatchdog 5 10 15 20 25 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 19.72 19.82
ASKAP Test: tConvolve OpenMP - Degridding OpenBenchmarking.org Million Grid Points Per Second, More Is Better ASKAP 1.0 Test: tConvolve OpenMP - Degridding RHEL 8.6 RHEL 9.0 tsc=nowatchdog 3K 6K 9K 12K 15K SE +/- 150.54, N = 15 SE +/- 0.00, N = 3 14868.3 14792.0 1. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp
Zstd Compression Compression Level: 19, Long Mode - Compression Speed OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.0 Compression Level: 19, Long Mode - Compression Speed RHEL 8.6 RHEL 9.0 tsc=nowatchdog 14 28 42 56 70 SE +/- 0.33, N = 3 SE +/- 0.12, N = 3 60.5 60.2 1. (CC) gcc options: -O3 -pthread -lz -llzma
OpenSSL OpenBenchmarking.org verify/s, More Is Better OpenSSL RHEL 8.6 RHEL 9.0 tsc=nowatchdog 400K 800K 1200K 1600K 2000K SE +/- 1189.19, N = 3 SE +/- 1359.58, N = 3 1763867.7 1771726.6 1. RHEL 8.6: OpenSSL 1.1.1k FIPS 25 Mar 2021 2. RHEL 9.0 tsc=nowatchdog: OpenSSL 3.0.1 14 Dec 2021 (Library: OpenSSL 3.0.1 14 Dec 2021)
Blender Blend File: Classroom - Compute: CPU-Only OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.0 Blend File: Classroom - Compute: CPU-Only RHEL 8.6 RHEL 9.0 tsc=nowatchdog 9 18 27 36 45 SE +/- 0.09, N = 3 SE +/- 0.06, N = 3 37.24 37.08
Blender Blend File: Fishy Cat - Compute: CPU-Only OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.0 Blend File: Fishy Cat - Compute: CPU-Only RHEL 8.6 RHEL 9.0 tsc=nowatchdog 5 10 15 20 25 SE +/- 0.03, N = 3 SE +/- 0.10, N = 3 19.11 19.19
Liquid-DSP Threads: 128 - Buffer Length: 256 - Filter Length: 57 OpenBenchmarking.org samples/s, More Is Better Liquid-DSP 2021.01.31 Threads: 128 - Buffer Length: 256 - Filter Length: 57 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 1200M 2400M 3600M 4800M 6000M SE +/- 4886830.37, N = 3 SE +/- 8248097.42, N = 3 5653166667 5673033333 1. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
Liquid-DSP Threads: 160 - Buffer Length: 256 - Filter Length: 57 OpenBenchmarking.org samples/s, More Is Better Liquid-DSP 2021.01.31 Threads: 160 - Buffer Length: 256 - Filter Length: 57 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 1200M 2400M 3600M 4800M 6000M SE +/- 6357235.60, N = 3 SE +/- 4812830.08, N = 3 5710066667 5729800000 1. (CC) gcc options: -O3 -pthread -lm -lc -lliquid
PyBench Total For Average Test Times OpenBenchmarking.org Milliseconds, Fewer Is Better PyBench 2018-02-16 Total For Average Test Times RHEL 8.6 RHEL 9.0 tsc=nowatchdog 300 600 900 1200 1500 SE +/- 3.84, N = 3 SE +/- 4.67, N = 3 1235 1239
OpenSSL OpenBenchmarking.org sign/s, More Is Better OpenSSL RHEL 8.6 RHEL 9.0 tsc=nowatchdog 6K 12K 18K 24K 30K SE +/- 23.12, N = 3 SE +/- 8.07, N = 3 27000.2 27060.3 1. RHEL 8.6: OpenSSL 1.1.1k FIPS 25 Mar 2021 2. RHEL 9.0 tsc=nowatchdog: OpenSSL 3.0.1 14 Dec 2021 (Library: OpenSSL 3.0.1 14 Dec 2021)
NAMD ATPase Simulation - 327,506 Atoms OpenBenchmarking.org days/ns, Fewer Is Better NAMD 2.14 ATPase Simulation - 327,506 Atoms RHEL 8.6 RHEL 9.0 tsc=nowatchdog 0.0509 0.1018 0.1527 0.2036 0.2545 SE +/- 0.00125, N = 3 SE +/- 0.00055, N = 3 0.22581 0.22630
ONNX Runtime Model: GPT-2 - Device: CPU - Executor: Standard OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.11 Model: GPT-2 - Device: CPU - Executor: Standard RHEL 8.6 RHEL 9.0 tsc=nowatchdog 1300 2600 3900 5200 6500 SE +/- 89.00, N = 12 SE +/- 31.09, N = 3 6187 6200 -pthread -lpthread 1. (CXX) g++ options: -ffunction-sections -fdata-sections -march=native -mtune=native -O3 -flto -fno-fat-lto-objects -ldl -lrt
OSPray Benchmark: gravity_spheres_volume/dim_512/pathtracer/real_time OpenBenchmarking.org Items Per Second, More Is Better OSPray 2.9 Benchmark: gravity_spheres_volume/dim_512/pathtracer/real_time RHEL 8.6 RHEL 9.0 tsc=nowatchdog 6 12 18 24 30 SE +/- 0.06, N = 3 SE +/- 0.09, N = 3 25.62 25.56
Zstd Compression Compression Level: 19 - Compression Speed OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.0 Compression Level: 19 - Compression Speed RHEL 8.6 RHEL 9.0 tsc=nowatchdog 20 40 60 80 100 SE +/- 0.63, N = 3 SE +/- 1.24, N = 3 109.2 109.4 1. (CC) gcc options: -O3 -pthread -lz -llzma
Stress-NG Test: CPU Cache OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.14 Test: CPU Cache RHEL 8.6 RHEL 9.0 tsc=nowatchdog 4 8 12 16 20 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 16.95 16.94 1. (CC) gcc options: -O2 -std=gnu99 -lm -lc -lcrypt -ldl -ljpeg -lrt -lsctp -lz -pthread
LuaJIT Test: Composite OpenBenchmarking.org Mflops, More Is Better LuaJIT 2.1-git Test: Composite RHEL 8.6 RHEL 9.0 tsc=nowatchdog 300 600 900 1200 1500 SE +/- 0.39, N = 3 SE +/- 0.80, N = 3 1295.56 1295.01 1. (CC) gcc options: -lm -ldl -O2 -fomit-frame-pointer -U_FORTIFY_SOURCE -fno-stack-protector
Stress-NG Test: Atomic OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.14 Test: Atomic RHEL 8.6 RHEL 9.0 tsc=nowatchdog 30K 60K 90K 120K 150K SE +/- 833.42, N = 3 SE +/- 518.53, N = 3 123081.86 123128.59 1. (CC) gcc options: -O2 -std=gnu99 -lm -lc -lcrypt -ldl -ljpeg -lrt -lsctp -lz -pthread
Blender Blend File: Barbershop - Compute: CPU-Only OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.0 Blend File: Barbershop - Compute: CPU-Only RHEL 8.6 RHEL 9.0 tsc=nowatchdog 30 60 90 120 150 SE +/- 0.41, N = 3 SE +/- 0.43, N = 3 142.62 142.61
ONNX Runtime Model: yolov4 - Device: CPU - Executor: Standard OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.11 Model: yolov4 - Device: CPU - Executor: Standard RHEL 8.6 RHEL 9.0 tsc=nowatchdog 60 120 180 240 300 SE +/- 1.53, N = 3 SE +/- 2.40, N = 3 265 265 -pthread -lpthread 1. (CXX) g++ options: -ffunction-sections -fdata-sections -march=native -mtune=native -O3 -flto -fno-fat-lto-objects -ldl -lrt
PyPerformance Benchmark: float OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.0.0 Benchmark: float RHEL 8.6 RHEL 9.0 tsc=nowatchdog 30 60 90 120 150 SE +/- 0.33, N = 3 133 133
simdjson Throughput Test: Kostya OpenBenchmarking.org GB/s, More Is Better simdjson 1.0 Throughput Test: Kostya RHEL 8.6 RHEL 9.0 tsc=nowatchdog 0.5805 1.161 1.7415 2.322 2.9025 SE +/- 0.01, N = 3 SE +/- 0.00, N = 3 2.58 2.58 -pthread 1. (CXX) g++ options: -O3
ONNX Runtime Model: super-resolution-10 - Device: CPU - Executor: Standard OpenBenchmarking.org Inferences Per Minute, More Is Better ONNX Runtime 1.11 Model: super-resolution-10 - Device: CPU - Executor: Standard RHEL 8.6 RHEL 9.0 tsc=nowatchdog 1300 2600 3900 5200 6500 SE +/- 176.83, N = 12 SE +/- 107.76, N = 12 4826 5878 -pthread -lpthread 1. (CXX) g++ options: -ffunction-sections -fdata-sections -march=native -mtune=native -O3 -flto -fno-fat-lto-objects -ldl -lrt
Mobile Neural Network Model: mobilenet-v1-1.0 OpenBenchmarking.org ms, Fewer Is Better Mobile Neural Network 1.2 Model: mobilenet-v1-1.0 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 1.0595 2.119 3.1785 4.238 5.2975 SE +/- 0.162, N = 9 SE +/- 0.172, N = 12 4.709 3.900 MIN: 3.49 / MAX: 15.34 MIN: 2.58 / MAX: 10.35 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 OpenBenchmarking.org ms, Fewer Is Better Mobile Neural Network 1.2 Model: SqueezeNetV1.0 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 2 4 6 8 10 SE +/- 0.490, N = 9 SE +/- 1.258, N = 12 8.485 8.714 MIN: 6.29 / MAX: 44.32 MIN: 5.68 / MAX: 49.6 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 OpenBenchmarking.org ms, Fewer Is Better Mobile Neural Network 1.2 Model: resnet-v2-50 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 7 14 21 28 35 SE +/- 0.50, N = 9 SE +/- 0.93, N = 12 28.92 24.40 MIN: 24.45 / MAX: 82.5 MIN: 18.61 / MAX: 56.64 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 OpenBenchmarking.org ms, Fewer Is Better Mobile Neural Network 1.2 Model: squeezenetv1.1 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 2 4 6 8 10 SE +/- 0.065, N = 9 SE +/- 0.768, N = 12 6.053 6.744 MIN: 5.61 / MAX: 19.47 MIN: 4.42 / MAX: 18.72 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 OpenBenchmarking.org ms, Fewer Is Better Mobile Neural Network 1.2 Model: mobilenetV3 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 2 4 6 8 10 SE +/- 1.129, N = 9 SE +/- 0.963, N = 12 6.003 4.924 MIN: 3.77 / MAX: 30.6 MIN: 2.84 / MAX: 24.76 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
Stress-NG Test: Futex OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.14 Test: Futex RHEL 8.6 RHEL 9.0 tsc=nowatchdog 400K 800K 1200K 1600K 2000K SE +/- 7833.12, N = 15 SE +/- 17824.22, N = 7 192662.22 1910034.62 1. (CC) gcc options: -O2 -std=gnu99 -lm -lc -lcrypt -ldl -ljpeg -lrt -lsctp -lz -pthread
MariaDB Clients: 128 OpenBenchmarking.org Queries Per Second, More Is Better MariaDB 10.8.2 Clients: 128 RHEL 8.6 RHEL 9.0 tsc=nowatchdog 100 200 300 400 500 SE +/- 2.62, N = 3 SE +/- 13.28, N = 9 479 416 -pthread 1. (CXX) g++ options: -pie -fPIC -fstack-protector -O3 -lpcre2-8 -lcrypt -lz -lm -lssl -lcrypto -lpthread -ldl
ASKAP Test: Hogbom Clean OpenMP OpenBenchmarking.org Iterations Per Second, More Is Better ASKAP 1.0 Test: Hogbom Clean OpenMP RHEL 8.6 RHEL 9.0 tsc=nowatchdog 80 160 240 320 400 SE +/- 19.19, N = 13 SE +/- 4.34, N = 4 319.78 361.82 1. (CXX) g++ options: -O3 -fstrict-aliasing -fopenmp
7-Zip Compression Test: Compression Rating OpenBenchmarking.org MIPS, More Is Better 7-Zip Compression 21.06 Test: Compression Rating RHEL 8.6 RHEL 9.0 tsc=nowatchdog 160K 320K 480K 640K 800K SE +/- 14990.19, N = 15 SE +/- 7637.62, N = 5 737301 706479 1. (CXX) g++ options: -lpthread -ldl -O2 -fPIC
Himeno Benchmark Poisson Pressure Solver OpenBenchmarking.org MFLOPS, More Is Better Himeno Benchmark 3.0 Poisson Pressure Solver RHEL 8.6 RHEL 9.0 tsc=nowatchdog 800 1600 2400 3200 4000 SE +/- 45.70, N = 15 SE +/- 60.44, N = 15 3547.51 3563.32 1. (CC) gcc options: -O3 -mavx2
OSPray Benchmark: particle_volume/pathtracer/real_time OpenBenchmarking.org Items Per Second, More Is Better OSPray 2.9 Benchmark: particle_volume/pathtracer/real_time RHEL 8.6 RHEL 9.0 tsc=nowatchdog 40 80 120 160 200 SE +/- 2.66, N = 9 SE +/- 4.03, N = 9 164.56 170.11
LuaRadio Test: Complex Phase OpenBenchmarking.org MiB/s, More Is Better LuaRadio 0.9.1 Test: Complex Phase RHEL 8.6 RHEL 9.0 tsc=nowatchdog 60 120 180 240 300 SE +/- 11.14, N = 3 SE +/- 6.83, N = 3 297.9 230.4
Zstd Compression Compression Level: 3, Long Mode - Decompression Speed OpenBenchmarking.org MB/s, More Is Better Zstd Compression Compression Level: 3, Long Mode - Decompression Speed RHEL 8.6 RHEL 9.0 tsc=nowatchdog 900 1800 2700 3600 4500 SE +/- 196.46, N = 5 SE +/- 2.43, N = 15 3182.7 3974.8 1. RHEL 8.6: *** zstd command line interface 64-bits v1.4.4, by Yann Collet *** 2. RHEL 9.0 tsc=nowatchdog: *** zstd command line interface 64-bits v1.5.1, by Yann Collet ***
Zstd Compression Compression Level: 19 - Decompression Speed OpenBenchmarking.org MB/s, More Is Better Zstd Compression Compression Level: 19 - Decompression Speed RHEL 8.6 RHEL 9.0 tsc=nowatchdog 700 1400 2100 2800 3500 SE +/- 218.57, N = 3 SE +/- 7.53, N = 3 2675.0 3251.0 1. RHEL 8.6: *** zstd command line interface 64-bits v1.4.4, by Yann Collet *** 2. RHEL 9.0 tsc=nowatchdog: *** zstd command line interface 64-bits v1.5.1, by Yann Collet ***
Zstd Compression Compression Level: 3 - Compression Speed OpenBenchmarking.org MB/s, More Is Better Zstd Compression 1.5.0 Compression Level: 3 - Compression Speed RHEL 8.6 RHEL 9.0 tsc=nowatchdog 1500 3000 4500 6000 7500 SE +/- 123.19, N = 14 SE +/- 85.87, N = 3 6998.9 6839.4 1. (CC) gcc options: -O3 -pthread -lz -llzma
Renaissance Test: Apache Spark ALS OpenBenchmarking.org ms, Fewer Is Better Renaissance 0.14 Test: Apache Spark ALS RHEL 8.6 RHEL 9.0 tsc=nowatchdog 600 1200 1800 2400 3000 SE +/- 158.91, N = 15 SE +/- 20.87, N = 3 2695.0 1857.2 MIN: 1905.3 / MAX: 3701.67 MIN: 1691.15 / MAX: 2053.68
Renaissance Test: Scala Dotty OpenBenchmarking.org ms, Fewer Is Better Renaissance 0.14 Test: Scala Dotty RHEL 8.6 RHEL 9.0 tsc=nowatchdog 200 400 600 800 1000 SE +/- 14.51, N = 12 SE +/- 15.08, N = 15 866.5 897.0 MIN: 670.92 / MAX: 1407.72 MIN: 692.9 / MAX: 1420.43
DaCapo Benchmark Java Test: Tradebeans OpenBenchmarking.org msec, Fewer Is Better DaCapo Benchmark 9.12-MR1 Java Test: Tradebeans RHEL 8.6 RHEL 9.0 tsc=nowatchdog 2K 4K 6K 8K 10K SE +/- 621.37, N = 4 SE +/- 97.65, N = 5 8248 9859
Xcompact3d Incompact3d Input: input.i3d 129 Cells Per Direction OpenBenchmarking.org Seconds, Fewer Is Better Xcompact3d Incompact3d 2021-03-11 Input: input.i3d 129 Cells Per Direction RHEL 8.6 RHEL 9.0 tsc=nowatchdog 0.7121 1.4242 2.1363 2.8484 3.5605 SE +/- 0.06100083, N = 15 SE +/- 0.07931796, N = 15 3.16484340 2.50625911 -pthread 1. (F9X) gfortran options: -cpp -O2 -funroll-loops -floop-optimize -fcray-pointer -fbacktrace -lmpi_usempif08 -lmpi_mpifh -lmpi
Pennant Test: leblancbig OpenBenchmarking.org Hydro Cycle Time - Seconds, Fewer Is Better Pennant 1.0.1 Test: leblancbig RHEL 8.6 RHEL 9.0 tsc=nowatchdog 0.9518 1.9036 2.8554 3.8072 4.759 SE +/- 0.069100, N = 15 SE +/- 0.081186, N = 15 4.230145 3.772760 -pthread 1. (CXX) g++ options: -fopenmp -lmpi_cxx -lmpi
Phoronix Test Suite v10.8.4