AMD EPYC and Intel Xeon benchmarks by Michael Larabel for a future article looking at Linux 6.6 (Git) vs. Linux 6.5 kernel performance.
Compare your own system(s) to this result file with the
Phoronix Test Suite by running the command:
phoronix-test-suite benchmark 2309189-NE-LINUX66AM62 Linux 6.6 Benchmarks AMD EPYC Genoa(X), Bergamo, and Xeon SPR - Phoronix Test Suite Linux 6.6 Benchmarks AMD EPYC Genoa(X), Bergamo, and Xeon SPR AMD EPYC and Intel Xeon benchmarks by Michael Larabel for a future article looking at Linux 6.6 (Git) vs. Linux 6.5 kernel performance.
HTML result view exported from: https://openbenchmarking.org/result/2309189-NE-LINUX66AM62&grr&export=pdf&sor .
Processor Motherboard Chipset Memory Disk Graphics Network OS Kernel Desktop Display Server Compiler File-System Screen Resolution EPYC 9654 EPYC 9684X EPYC 9754 Xeon Platinum 8490H Linux 6.5.1 Linux 6.6-rc1 Linux 6.5.1 Linux 6.6-rc1 Linux 6.5.1 Linux 6.6-rc1 Linux 6.5.1 Linux 6.6-rc1 AMD EPYC 9654 96-Core @ 2.40GHz (96 Cores / 192 Threads) AMD Titanite_4G (RTI1007B BIOS) AMD Device 14a4 768GB 3201GB Micron_7450_MTFDKCC3T2TFS ASPEED Broadcom NetXtreme BCM5720 PCIe Ubuntu 23.10 6.5.1-060501-generic (x86_64) GNOME Shell X Server 1.21.1.7 GCC 13.2.0 ext4 640x480 6.6.0-060600rc1-generic (x86_64) 1920x1200 AMD EPYC 9684X 96-Core @ 2.55GHz (96 Cores / 192 Threads) 6.5.1-060501-generic (x86_64) 640x480 6.6.0-060600rc1-generic (x86_64) 1920x1200 AMD EPYC 9754 128-Core @ 2.25GHz (128 Cores / 256 Threads) 6.5.1-060501-generic (x86_64) 640x480 6.6.0-060600rc1-generic (x86_64) 1920x1200 Intel Xeon Platinum 8490H @ 3.50GHz (60 Cores / 120 Threads) Quanta Cloud S6Q-MB-MPS (3A10.uh BIOS) Intel Device 1bce 512GB 6.5.1-060501-generic (x86_64) 640x480 6.6.0-060600rc1-generic (x86_64) 1920x1200 OpenBenchmarking.org Kernel Details - Transparent Huge Pages: madvise Compiler Details - --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --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-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-nEN1TP/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-nEN1TP/gcc-13-13.2.0/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-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v Disk Details - EPYC 9654: Linux 6.5.1, EPYC 9654: Linux 6.6-rc1, EPYC 9754: Linux 6.5.1, EPYC 9754: Linux 6.6-rc1: NONE / relatime,rw,stripe=64 / Block Size: 4096 Processor Details - EPYC 9654: Linux 6.5.1: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa10113e - EPYC 9654: Linux 6.6-rc1: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa10113e - EPYC 9684X: Linux 6.5.1: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa10113e - EPYC 9684X: Linux 6.6-rc1: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa10113e - EPYC 9754: Linux 6.5.1: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xaa00116 - EPYC 9754: Linux 6.6-rc1: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xaa00116 - Xeon Platinum 8490H: Linux 6.5.1: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x2b0004b1 - Xeon Platinum 8490H: Linux 6.6-rc1: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x2b0004b1 Java Details - OpenJDK Runtime Environment (build 11.0.20+8-post-Ubuntu-1ubuntu1) Python Details - Python 3.11.5 Security Details - EPYC 9654: Linux 6.5.1: 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: Mitigation of safe RET + 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 STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected - EPYC 9654: Linux 6.6-rc1: 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: Mitigation of safe RET + 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 STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected - EPYC 9684X: Linux 6.5.1: 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: Mitigation of safe RET + 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 STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected - EPYC 9684X: Linux 6.6-rc1: 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: Mitigation of safe RET + 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 STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected - EPYC 9754: Linux 6.5.1: 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: Mitigation of safe RET + 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 STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected - EPYC 9754: Linux 6.6-rc1: 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: Mitigation of safe RET + 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 STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected - Xeon Platinum 8490H: Linux 6.5.1: 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 - Xeon Platinum 8490H: Linux 6.6-rc1: 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
tidb: oltp_update_non_index - 256 tidb: oltp_update_index - 256 tidb: oltp_update_non_index - 128 build-linux-kernel: allmodconfig hadoop: Delete - 500 - 100000 pgbench: 100 - 1000 - Read Write - Average Latency pgbench: 100 - 1000 - Read Write clickhouse: 100M Rows Hits Dataset, Third Run clickhouse: 100M Rows Hits Dataset, Second Run clickhouse: 100M Rows Hits Dataset, First Run / Cold Cache pgbench: 100 - 800 - Read Write - Average Latency pgbench: 100 - 800 - Read Write blender: Barbershop - CPU-Only hadoop: Create - 100 - 100000 build-nodejs: Time To Compile cassandra: Writes hadoop: Create - 500 - 100000 rocksdb: Update Rand memtier-benchmark: Redis - 100 - 1:5 hadoop: Rename - 500 - 100000 renaissance: Apache Spark PageRank rocksdb: Read Rand Write Rand stress-ng: Context Switching nginx: 500 apache: 500 tensorflow: CPU - 64 - ResNet-50 sqlite-speedtest: Timed Time - Size 1,000 memcached: 1:100 apache-iotdb: 100 - 100 - 500 apache-iotdb: 100 - 100 - 500 renaissance: Apache Spark ALS tensorflow: CPU - 32 - ResNet-50 blender: Pabellon Barcelona - CPU-Only build-linux-kernel: defconfig apache-iotdb: 500 - 1 - 500 apache-iotdb: 500 - 1 - 500 blender: Classroom - CPU-Only tensorflow: CPU - 16 - ResNet-50 stress-ng: Pthread blender: Fishy Cat - CPU-Only blender: BMW27 - CPU-Only EPYC 9654 EPYC 9684X EPYC 9754 Xeon Platinum 8490H Linux 6.5.1 Linux 6.6-rc1 Linux 6.5.1 Linux 6.6-rc1 Linux 6.5.1 Linux 6.6-rc1 Linux 6.5.1 Linux 6.6-rc1 52333 30087 42617 282.696 45576 25.916 38592 556.04 550.24 538.09 18.298 43881 145.42 19667 119.798 298854 17775 426986 2875244.82 39717 2771.6 1946644 16242563.70 151141.19 98407.86 79.69 76.132 7727647.65 72.10 56532845.87 1581.6 59.43 49.16 36.284 25.78 1730485.73 40.63 42.93 73831.46 21.28 16.15 83040 40539 71410 270.329 61577 13.587 73638 559.44 551.95 531.47 10.050 79606 141.95 20817 118.576 293305 17986 582786 3046453.68 47095 2722.4 2464466 21629740.81 309505.55 119212.00 102.85 74.961 8278132.33 68.04 59531334.81 1527.2 79.93 48.66 35.381 25.06 1750353.96 39.97 55.00 96686.04 20.59 15.90 69410 36186 59326 270.067 55951 20.629 48496 574.52 565.25 546.48 14.023 57070 143.14 20242 118.111 286356 18408 560022 2976413.39 44701 2630.9 2408677 18877335.52 182923.75 134810.42 92.39 75.358 7481630.22 68.55 59064309.14 1568.4 74.95 49.93 35.635 30.12 1499845.25 40.59 54.53 94602.26 20.74 16.31 79091 40863 68832 271.014 58090 15.362 65181 575.41 574.44 544.55 11.466 69792 142.98 21764 117.585 278694 17661 567519 2956527.98 47576 2648.7 2334279 20823667.30 290002.44 128556.25 92.22 74.371 7839246.78 66.93 59831087.46 1553.0 75.04 49.74 35.562 27.05 1636368.32 40.59 54.00 94337.45 20.71 16.32 38237 22715 30885 310.822 32964 24.053 41580 446.56 437.00 430.32 17.975 44510 120.24 14759 129.401 236822 14484 387449 2853269.76 26346 3294.4 2067895 19572583.51 133873.45 102203.23 78.87 91.876 7200983.53 83.66 49577934.35 1805.7 56.81 40.03 41.180 31.15 1445204.97 31.99 35.77 80145.36 17.13 13.17 66800 36795 58561 309.465 43753 14.603 68479 450.60 443.56 427.88 10.749 74426 117.60 16077 129.141 234456 14660 515261 2887591.22 35117 3265.8 2280220 25850837.13 324834.51 113440.94 94.01 90.579 7463997.96 83.94 49684489.24 1803.0 66.95 39.49 40.773 30.62 1447102.06 31.45 41.15 87300.83 16.76 13.04 96035 45081 78708 288.127 77811 14.815 67514 474.00 486.53 469.95 10.345 77336 273.61 49132 150.774 139942 43518 684747 2955837.33 66564 2917.3 3077247 12163462.02 279906.65 135243.78 110.95 53.827 4557102.26 68.91 58516616.33 1868.7 93.69 88.52 35.407 23.26 1892800.20 69.14 71.63 117098.86 35.46 25.91 96421 45316 78697 287.610 78750 14.768 67714 482.57 474.49 463.47 10.579 75625 273.57 48216 151.144 146888 42813 707640 2943924.78 66116 2931.3 3095641 10960102.41 373739.96 138016.99 110.48 54.054 4354637.29 66.91 60205390.81 1844.3 92.99 88.74 35.391 24.03 1829793.13 69.19 70.96 112147.13 35.42 25.85 OpenBenchmarking.org
TiDB Community Server Test: oltp_update_non_index - Threads: 256 Xeon Platinum 8490H EPYC 9654 EPYC 9684X EPYC 9754 OpenBenchmarking.org Queries Per Second, More Is Better TiDB Community Server 7.3 Test: oltp_update_non_index - Threads: 256 Linux 6.6-rc1 Linux 6.5.1 20K 40K 60K 80K 100K SE +/- 387.94, N = 3 SE +/- 289.09, N = 3 SE +/- 213.11, N = 3 SE +/- 446.22, N = 3 SE +/- 159.88, N = 3 SE +/- 310.05, N = 3 SE +/- 75.74, N = 3 SE +/- 384.35, N = 9 96421 96035 83040 52333 79091 69410 66800 38237
TiDB Community Server Test: oltp_update_index - Threads: 256 Xeon Platinum 8490H EPYC 9684X EPYC 9654 EPYC 9754 OpenBenchmarking.org Queries Per Second, More Is Better TiDB Community Server 7.3 Test: oltp_update_index - Threads: 256 Linux 6.6-rc1 Linux 6.5.1 10K 20K 30K 40K 50K SE +/- 214.83, N = 3 SE +/- 307.45, N = 3 SE +/- 337.03, N = 3 SE +/- 178.80, N = 3 SE +/- 250.46, N = 3 SE +/- 538.20, N = 9 SE +/- 149.16, N = 2 SE +/- 41.27, N = 3 45316 45081 40863 36186 40539 30087 36795 22715
TiDB Community Server Test: oltp_update_non_index - Threads: 128 Xeon Platinum 8490H EPYC 9654 EPYC 9684X EPYC 9754 OpenBenchmarking.org Queries Per Second, More Is Better TiDB Community Server 7.3 Test: oltp_update_non_index - Threads: 128 Linux 6.5.1 Linux 6.6-rc1 20K 40K 60K 80K 100K SE +/- 103.03, N = 3 SE +/- 384.95, N = 2 SE +/- 170.18, N = 3 SE +/- 287.78, N = 3 SE +/- 173.69, N = 3 SE +/- 203.37, N = 3 SE +/- 204.18, N = 3 SE +/- 312.34, N = 3 78708 78697 71410 42617 68832 59326 58561 30885
Timed Linux Kernel Compilation Build: allmodconfig EPYC 9684X EPYC 9654 Xeon Platinum 8490H EPYC 9754 OpenBenchmarking.org Seconds, Fewer Is Better Timed Linux Kernel Compilation 6.1 Build: allmodconfig Linux 6.5.1 Linux 6.6-rc1 70 140 210 280 350 SE +/- 0.67, N = 3 SE +/- 1.58, N = 3 SE +/- 0.61, N = 3 SE +/- 3.65, N = 3 SE +/- 0.64, N = 3 SE +/- 0.54, N = 3 SE +/- 1.57, N = 3 SE +/- 0.91, N = 3 270.07 271.01 270.33 282.70 287.61 288.13 309.47 310.82
Apache Hadoop Operation: Delete - Threads: 500 - Files: 100000 Xeon Platinum 8490H EPYC 9654 EPYC 9684X EPYC 9754 OpenBenchmarking.org Ops per sec, More Is Better Apache Hadoop 3.3.6 Operation: Delete - Threads: 500 - Files: 100000 Linux 6.6-rc1 Linux 6.5.1 20K 40K 60K 80K 100K SE +/- 1094.65, N = 3 SE +/- 713.93, N = 15 SE +/- 922.30, N = 15 SE +/- 1370.80, N = 15 SE +/- 1263.41, N = 15 SE +/- 945.00, N = 12 SE +/- 1071.18, N = 15 SE +/- 1173.88, N = 15 78750 77811 61577 45576 58090 55951 43753 32964
PostgreSQL Scaling Factor: 100 - Clients: 1000 - Mode: Read Write - Average Latency EPYC 9654 EPYC 9754 Xeon Platinum 8490H EPYC 9684X OpenBenchmarking.org ms, Fewer Is Better PostgreSQL 15 Scaling Factor: 100 - Clients: 1000 - Mode: Read Write - Average Latency Linux 6.6-rc1 Linux 6.5.1 6 12 18 24 30 SE +/- 0.15, N = 5 SE +/- 0.22, N = 3 SE +/- 0.04, N = 3 SE +/- 0.17, N = 3 SE +/- 0.05, N = 3 SE +/- 0.16, N = 3 SE +/- 0.17, N = 12 SE +/- 0.29, N = 3 13.59 25.92 14.60 24.05 14.77 14.82 15.36 20.63 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
PostgreSQL Scaling Factor: 100 - Clients: 1000 - Mode: Read Write EPYC 9654 EPYC 9754 Xeon Platinum 8490H EPYC 9684X OpenBenchmarking.org TPS, More Is Better PostgreSQL 15 Scaling Factor: 100 - Clients: 1000 - Mode: Read Write Linux 6.6-rc1 Linux 6.5.1 16K 32K 48K 64K 80K SE +/- 805.64, N = 5 SE +/- 329.37, N = 3 SE +/- 185.20, N = 3 SE +/- 291.85, N = 3 SE +/- 209.58, N = 3 SE +/- 720.22, N = 3 SE +/- 731.64, N = 12 SE +/- 687.05, N = 3 73638 38592 68479 41580 67714 67514 65181 48496 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
ClickHouse 100M Rows Hits Dataset, Third Run EPYC 9684X EPYC 9654 Xeon Platinum 8490H EPYC 9754 OpenBenchmarking.org Queries Per Minute, Geo Mean, More Is Better ClickHouse 22.12.3.5 100M Rows Hits Dataset, Third Run Linux 6.6-rc1 Linux 6.5.1 120 240 360 480 600 SE +/- 5.74, N = 3 SE +/- 3.01, N = 3 SE +/- 2.23, N = 3 SE +/- 5.36, N = 3 SE +/- 8.15, N = 3 SE +/- 3.58, N = 3 SE +/- 1.27, N = 3 SE +/- 3.03, N = 3 575.41 574.52 559.44 556.04 482.57 474.00 450.60 446.56
ClickHouse 100M Rows Hits Dataset, Second Run EPYC 9684X EPYC 9654 Xeon Platinum 8490H EPYC 9754 OpenBenchmarking.org Queries Per Minute, Geo Mean, More Is Better ClickHouse 22.12.3.5 100M Rows Hits Dataset, Second Run Linux 6.6-rc1 Linux 6.5.1 120 240 360 480 600 SE +/- 6.48, N = 3 SE +/- 3.48, N = 3 SE +/- 5.49, N = 3 SE +/- 0.51, N = 3 SE +/- 2.66, N = 3 SE +/- 1.02, N = 3 SE +/- 3.18, N = 3 SE +/- 3.52, N = 3 574.44 565.25 551.95 550.24 486.53 474.49 443.56 437.00
ClickHouse 100M Rows Hits Dataset, First Run / Cold Cache EPYC 9684X EPYC 9654 Xeon Platinum 8490H EPYC 9754 OpenBenchmarking.org Queries Per Minute, Geo Mean, More Is Better ClickHouse 22.12.3.5 100M Rows Hits Dataset, First Run / Cold Cache Linux 6.5.1 Linux 6.6-rc1 120 240 360 480 600 SE +/- 7.07, N = 3 SE +/- 6.05, N = 3 SE +/- 7.76, N = 3 SE +/- 1.79, N = 3 SE +/- 1.55, N = 3 SE +/- 3.61, N = 3 SE +/- 1.84, N = 3 SE +/- 4.13, N = 3 546.48 544.55 538.09 531.47 469.95 463.47 430.32 427.88
PostgreSQL Scaling Factor: 100 - Clients: 800 - Mode: Read Write - Average Latency EPYC 9654 Xeon Platinum 8490H EPYC 9754 EPYC 9684X OpenBenchmarking.org ms, Fewer Is Better PostgreSQL 15 Scaling Factor: 100 - Clients: 800 - Mode: Read Write - Average Latency Linux 6.6-rc1 Linux 6.5.1 5 10 15 20 25 SE +/- 0.07, N = 3 SE +/- 0.32, N = 12 SE +/- 0.06, N = 3 SE +/- 0.06, N = 3 SE +/- 0.04, N = 3 SE +/- 0.11, N = 3 SE +/- 0.15, N = 3 SE +/- 0.19, N = 3 10.05 18.30 10.35 10.58 10.75 17.98 11.47 14.02 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
PostgreSQL Scaling Factor: 100 - Clients: 800 - Mode: Read Write EPYC 9654 Xeon Platinum 8490H EPYC 9754 EPYC 9684X OpenBenchmarking.org TPS, More Is Better PostgreSQL 15 Scaling Factor: 100 - Clients: 800 - Mode: Read Write Linux 6.6-rc1 Linux 6.5.1 20K 40K 60K 80K 100K SE +/- 532.86, N = 3 SE +/- 826.44, N = 12 SE +/- 423.16, N = 3 SE +/- 398.61, N = 3 SE +/- 244.17, N = 3 SE +/- 273.63, N = 3 SE +/- 896.96, N = 3 SE +/- 766.03, N = 3 79606 43881 77336 75625 74426 44510 69792 57070 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
Blender Blend File: Barbershop - Compute: CPU-Only EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8490H OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.6 Blend File: Barbershop - Compute: CPU-Only Linux 6.6-rc1 Linux 6.5.1 60 120 180 240 300 SE +/- 0.14, N = 3 SE +/- 0.32, N = 3 SE +/- 0.21, N = 3 SE +/- 0.28, N = 3 SE +/- 0.05, N = 3 SE +/- 0.13, N = 3 SE +/- 0.34, N = 3 SE +/- 0.33, N = 3 117.60 120.24 141.95 145.42 142.98 143.14 273.57 273.61
Apache Hadoop Operation: Create - Threads: 100 - Files: 100000 Xeon Platinum 8490H EPYC 9684X EPYC 9654 EPYC 9754 OpenBenchmarking.org Ops per sec, More Is Better Apache Hadoop 3.3.6 Operation: Create - Threads: 100 - Files: 100000 Linux 6.5.1 Linux 6.6-rc1 11K 22K 33K 44K 55K SE +/- 432.37, N = 3 SE +/- 530.32, N = 4 SE +/- 1237.07, N = 12 SE +/- 763.53, N = 15 SE +/- 716.26, N = 15 SE +/- 1058.51, N = 15 SE +/- 39.40, N = 3 SE +/- 455.77, N = 12 49132 48216 21764 20242 20817 19667 16077 14759
Timed Node.js Compilation Time To Compile EPYC 9684X EPYC 9654 EPYC 9754 Xeon Platinum 8490H OpenBenchmarking.org Seconds, Fewer Is Better Timed Node.js Compilation 19.8.1 Time To Compile Linux 6.6-rc1 Linux 6.5.1 30 60 90 120 150 SE +/- 0.33, N = 3 SE +/- 0.23, N = 3 SE +/- 0.05, N = 3 SE +/- 0.29, N = 3 SE +/- 0.28, N = 3 SE +/- 0.03, N = 3 SE +/- 0.10, N = 3 SE +/- 0.17, N = 3 117.59 118.11 118.58 119.80 129.14 129.40 150.77 151.14
Apache Cassandra Test: Writes EPYC 9654 EPYC 9684X EPYC 9754 Xeon Platinum 8490H OpenBenchmarking.org Op/s, More Is Better Apache Cassandra 4.1.3 Test: Writes Linux 6.5.1 Linux 6.6-rc1 60K 120K 180K 240K 300K SE +/- 850.68, N = 3 SE +/- 925.21, N = 3 SE +/- 1077.34, N = 3 SE +/- 978.81, N = 3 SE +/- 2388.51, N = 3 SE +/- 2557.09, N = 3 SE +/- 536.40, N = 3 SE +/- 426.76, N = 3 298854 293305 286356 278694 236822 234456 146888 139942
Apache Hadoop Operation: Create - Threads: 500 - Files: 100000 Xeon Platinum 8490H EPYC 9684X EPYC 9654 EPYC 9754 OpenBenchmarking.org Ops per sec, More Is Better Apache Hadoop 3.3.6 Operation: Create - Threads: 500 - Files: 100000 Linux 6.5.1 Linux 6.6-rc1 9K 18K 27K 36K 45K SE +/- 418.80, N = 3 SE +/- 304.04, N = 3 SE +/- 194.81, N = 3 SE +/- 380.81, N = 15 SE +/- 95.32, N = 3 SE +/- 581.69, N = 12 SE +/- 120.56, N = 12 SE +/- 211.36, N = 15 43518 42813 18408 17661 17986 17775 14660 14484
RocksDB Test: Update Random Xeon Platinum 8490H EPYC 9654 EPYC 9684X EPYC 9754 OpenBenchmarking.org Op/s, More Is Better RocksDB 8.0 Test: Update Random Linux 6.6-rc1 Linux 6.5.1 150K 300K 450K 600K 750K SE +/- 3476.68, N = 3 SE +/- 1930.54, N = 3 SE +/- 589.75, N = 3 SE +/- 9700.53, N = 15 SE +/- 417.45, N = 3 SE +/- 996.65, N = 3 SE +/- 708.03, N = 3 SE +/- 11096.16, N = 15 707640 684747 582786 426986 567519 560022 515261 387449 1. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread
Redis 7.0.12 + memtier_benchmark Protocol: Redis - Clients: 100 - Set To Get Ratio: 1:5 EPYC 9654 EPYC 9684X Xeon Platinum 8490H EPYC 9754 OpenBenchmarking.org Ops/sec, More Is Better Redis 7.0.12 + memtier_benchmark 2.0 Protocol: Redis - Clients: 100 - Set To Get Ratio: 1:5 Linux 6.6-rc1 Linux 6.5.1 700K 1400K 2100K 2800K 3500K SE +/- 28761.65, N = 3 SE +/- 60151.15, N = 15 SE +/- 11128.04, N = 3 SE +/- 13409.46, N = 3 SE +/- 25445.36, N = 3 SE +/- 16135.74, N = 3 SE +/- 21084.09, N = 3 SE +/- 36417.01, N = 3 3046453.68 2875244.82 2976413.39 2956527.98 2955837.33 2943924.78 2887591.22 2853269.76 1. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre
Apache Hadoop Operation: Rename - Threads: 500 - Files: 100000 Xeon Platinum 8490H EPYC 9684X EPYC 9654 EPYC 9754 OpenBenchmarking.org Ops per sec, More Is Better Apache Hadoop 3.3.6 Operation: Rename - Threads: 500 - Files: 100000 Linux 6.5.1 Linux 6.6-rc1 14K 28K 42K 56K 70K SE +/- 125.96, N = 3 SE +/- 689.95, N = 4 SE +/- 250.61, N = 3 SE +/- 344.46, N = 3 SE +/- 534.89, N = 15 SE +/- 519.86, N = 3 SE +/- 386.00, N = 3 SE +/- 666.53, N = 15 66564 66116 47576 44701 47095 39717 35117 26346
Renaissance Test: Apache Spark PageRank EPYC 9684X EPYC 9654 Xeon Platinum 8490H EPYC 9754 OpenBenchmarking.org ms, Fewer Is Better Renaissance 0.14 Test: Apache Spark PageRank Linux 6.5.1 Linux 6.6-rc1 700 1400 2100 2800 3500 SE +/- 22.71, N = 3 SE +/- 20.86, N = 3 SE +/- 7.73, N = 3 SE +/- 14.62, N = 3 SE +/- 32.63, N = 3 SE +/- 18.67, N = 3 SE +/- 30.24, N = 3 SE +/- 35.18, N = 3 2630.9 2648.7 2722.4 2771.6 2917.3 2931.3 3265.8 3294.4
RocksDB Test: Read Random Write Random Xeon Platinum 8490H EPYC 9654 EPYC 9684X EPYC 9754 OpenBenchmarking.org Op/s, More Is Better RocksDB 8.0 Test: Read Random Write Random Linux 6.6-rc1 Linux 6.5.1 700K 1400K 2100K 2800K 3500K SE +/- 4001.86, N = 3 SE +/- 3155.99, N = 3 SE +/- 2627.72, N = 3 SE +/- 15269.98, N = 3 SE +/- 7550.49, N = 3 SE +/- 2183.63, N = 3 SE +/- 16611.36, N = 15 SE +/- 19389.11, N = 6 3095641 3077247 2464466 1946644 2408677 2334279 2280220 2067895 1. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread
Stress-NG Test: Context Switching EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8490H OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: Context Switching Linux 6.6-rc1 Linux 6.5.1 6M 12M 18M 24M 30M SE +/- 99220.16, N = 3 SE +/- 317478.56, N = 15 SE +/- 35574.92, N = 3 SE +/- 244007.46, N = 15 SE +/- 43524.32, N = 3 SE +/- 52908.90, N = 3 SE +/- 1907162.84, N = 15 SE +/- 2178551.01, N = 15 25850837.13 19572583.51 21629740.81 16242563.70 20823667.30 18877335.52 12163462.02 10960102.41 1. (CXX) g++ options: -O2 -std=gnu99 -lc
nginx Connections: 500 Xeon Platinum 8490H EPYC 9754 EPYC 9654 EPYC 9684X OpenBenchmarking.org Requests Per Second, More Is Better nginx 1.23.2 Connections: 500 Linux 6.6-rc1 Linux 6.5.1 80K 160K 240K 320K 400K SE +/- 732.01, N = 3 SE +/- 609.97, N = 3 SE +/- 1624.12, N = 3 SE +/- 212.94, N = 3 SE +/- 578.66, N = 3 SE +/- 308.46, N = 3 SE +/- 312.37, N = 3 SE +/- 99.67, N = 3 373739.96 279906.65 324834.51 133873.45 309505.55 151141.19 290002.44 182923.75 1. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2
Apache HTTP Server Concurrent Requests: 500 Xeon Platinum 8490H EPYC 9684X EPYC 9654 EPYC 9754 OpenBenchmarking.org Requests Per Second, More Is Better Apache HTTP Server 2.4.56 Concurrent Requests: 500 Linux 6.6-rc1 Linux 6.5.1 30K 60K 90K 120K 150K SE +/- 125.55, N = 3 SE +/- 120.97, N = 3 SE +/- 1448.64, N = 3 SE +/- 293.11, N = 3 SE +/- 322.13, N = 3 SE +/- 217.59, N = 3 SE +/- 472.34, N = 3 SE +/- 107.86, N = 3 138016.99 135243.78 134810.42 128556.25 119212.00 98407.86 113440.94 102203.23 1. (CC) gcc options: -lluajit-5.1 -lm -lssl -lcrypto -lpthread -ldl -std=c99 -O2
TensorFlow Device: CPU - Batch Size: 64 - Model: ResNet-50 Xeon Platinum 8490H EPYC 9654 EPYC 9754 EPYC 9684X OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.12 Device: CPU - Batch Size: 64 - Model: ResNet-50 Linux 6.5.1 Linux 6.6-rc1 20 40 60 80 100 SE +/- 0.03, N = 3 SE +/- 0.03, N = 3 SE +/- 0.37, N = 3 SE +/- 0.17, N = 3 SE +/- 0.49, N = 3 SE +/- 0.51, N = 3 SE +/- 0.19, N = 3 SE +/- 0.03, N = 3 110.95 110.48 102.85 79.69 94.01 78.87 92.39 92.22
SQLite Speedtest Timed Time - Size 1,000 Xeon Platinum 8490H EPYC 9684X EPYC 9654 EPYC 9754 OpenBenchmarking.org Seconds, Fewer Is Better SQLite Speedtest 3.30 Timed Time - Size 1,000 Linux 6.5.1 Linux 6.6-rc1 20 40 60 80 100 SE +/- 0.13, N = 3 SE +/- 0.18, N = 3 SE +/- 0.04, N = 3 SE +/- 0.01, N = 3 SE +/- 0.04, N = 3 SE +/- 0.32, N = 3 SE +/- 0.19, N = 3 SE +/- 0.02, N = 3 53.83 54.05 74.37 75.36 74.96 76.13 90.58 91.88 1. (CC) gcc options: -O2 -lz
Memcached Set To Get Ratio: 1:100 EPYC 9654 EPYC 9684X EPYC 9754 Xeon Platinum 8490H OpenBenchmarking.org Ops/sec, More Is Better Memcached 1.6.19 Set To Get Ratio: 1:100 Linux 6.6-rc1 Linux 6.5.1 2M 4M 6M 8M 10M SE +/- 39410.84, N = 3 SE +/- 31537.67, N = 3 SE +/- 24338.08, N = 3 SE +/- 23233.41, N = 3 SE +/- 25008.72, N = 3 SE +/- 18708.63, N = 3 SE +/- 9057.27, N = 3 SE +/- 51449.56, N = 4 8278132.33 7727647.65 7839246.78 7481630.22 7463997.96 7200983.53 4557102.26 4354637.29 1. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre
Apache IoTDB Device Count: 100 - Batch Size Per Write: 100 - Sensor Count: 500 Xeon Platinum 8490H EPYC 9684X EPYC 9654 EPYC 9754 OpenBenchmarking.org Average Latency, Fewer Is Better Apache IoTDB 1.1.2 Device Count: 100 - Batch Size Per Write: 100 - Sensor Count: 500 Linux 6.6-rc1 Linux 6.5.1 20 40 60 80 100 SE +/- 0.88, N = 12 SE +/- 0.29, N = 3 SE +/- 0.47, N = 3 SE +/- 0.68, N = 15 SE +/- 1.00, N = 3 SE +/- 0.84, N = 15 SE +/- 0.28, N = 3 SE +/- 1.15, N = 3 66.91 68.91 66.93 68.55 68.04 72.10 83.66 83.94
Apache IoTDB Device Count: 100 - Batch Size Per Write: 100 - Sensor Count: 500 Xeon Platinum 8490H EPYC 9684X EPYC 9654 EPYC 9754 OpenBenchmarking.org point/sec, More Is Better Apache IoTDB 1.1.2 Device Count: 100 - Batch Size Per Write: 100 - Sensor Count: 500 Linux 6.6-rc1 Linux 6.5.1 13M 26M 39M 52M 65M SE +/- 655518.44, N = 12 SE +/- 129652.43, N = 3 SE +/- 283687.36, N = 3 SE +/- 477800.57, N = 15 SE +/- 772180.53, N = 3 SE +/- 513129.74, N = 15 SE +/- 575371.81, N = 3 SE +/- 214551.39, N = 3 60205390.81 58516616.33 59831087.46 59064309.14 59531334.81 56532845.87 49684489.24 49577934.35
Renaissance Test: Apache Spark ALS EPYC 9654 EPYC 9684X EPYC 9754 Xeon Platinum 8490H OpenBenchmarking.org ms, Fewer Is Better Renaissance 0.14 Test: Apache Spark ALS Linux 6.6-rc1 Linux 6.5.1 400 800 1200 1600 2000 SE +/- 2.46, N = 3 SE +/- 9.89, N = 3 SE +/- 8.63, N = 3 SE +/- 8.03, N = 3 SE +/- 6.78, N = 3 SE +/- 2.05, N = 3 SE +/- 2.69, N = 3 SE +/- 13.41, N = 3 1527.2 1581.6 1553.0 1568.4 1803.0 1805.7 1844.3 1868.7
TensorFlow Device: CPU - Batch Size: 32 - Model: ResNet-50 Xeon Platinum 8490H EPYC 9654 EPYC 9684X EPYC 9754 OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.12 Device: CPU - Batch Size: 32 - Model: ResNet-50 Linux 6.5.1 Linux 6.6-rc1 20 40 60 80 100 SE +/- 0.07, N = 3 SE +/- 0.19, N = 3 SE +/- 0.07, N = 3 SE +/- 0.06, N = 3 SE +/- 0.23, N = 3 SE +/- 0.37, N = 3 SE +/- 0.40, N = 3 SE +/- 0.63, N = 5 93.69 92.99 79.93 59.43 75.04 74.95 66.95 56.81
Blender Blend File: Pabellon Barcelona - Compute: CPU-Only EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8490H OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.6 Blend File: Pabellon Barcelona - Compute: CPU-Only Linux 6.6-rc1 Linux 6.5.1 20 40 60 80 100 SE +/- 0.12, N = 3 SE +/- 0.08, N = 3 SE +/- 0.14, N = 3 SE +/- 0.12, N = 3 SE +/- 0.08, N = 3 SE +/- 0.01, N = 3 SE +/- 0.10, N = 3 SE +/- 0.20, N = 3 39.49 40.03 48.66 49.16 49.74 49.93 88.52 88.74
Timed Linux Kernel Compilation Build: defconfig EPYC 9654 Xeon Platinum 8490H EPYC 9684X EPYC 9754 OpenBenchmarking.org Seconds, Fewer Is Better Timed Linux Kernel Compilation 6.1 Build: defconfig Linux 6.6-rc1 Linux 6.5.1 9 18 27 36 45 SE +/- 0.36, N = 5 SE +/- 0.44, N = 4 SE +/- 0.50, N = 3 SE +/- 0.38, N = 4 SE +/- 0.43, N = 4 SE +/- 0.36, N = 5 SE +/- 0.45, N = 5 SE +/- 0.43, N = 5 35.38 36.28 35.39 35.41 35.56 35.64 40.77 41.18
Apache IoTDB Device Count: 500 - Batch Size Per Write: 1 - Sensor Count: 500 Xeon Platinum 8490H EPYC 9654 EPYC 9684X EPYC 9754 OpenBenchmarking.org Average Latency, Fewer Is Better Apache IoTDB 1.1.2 Device Count: 500 - Batch Size Per Write: 1 - Sensor Count: 500 Linux 6.5.1 Linux 6.6-rc1 7 14 21 28 35 SE +/- 0.10, N = 3 SE +/- 0.32, N = 4 SE +/- 0.20, N = 3 SE +/- 0.16, N = 3 SE +/- 0.43, N = 3 SE +/- 0.25, N = 3 SE +/- 0.35, N = 3 SE +/- 0.37, N = 3 23.26 24.03 25.06 25.78 27.05 30.12 30.62 31.15
Apache IoTDB Device Count: 500 - Batch Size Per Write: 1 - Sensor Count: 500 Xeon Platinum 8490H EPYC 9654 EPYC 9684X EPYC 9754 OpenBenchmarking.org point/sec, More Is Better Apache IoTDB 1.1.2 Device Count: 500 - Batch Size Per Write: 1 - Sensor Count: 500 Linux 6.5.1 Linux 6.6-rc1 400K 800K 1200K 1600K 2000K SE +/- 3938.02, N = 3 SE +/- 21947.33, N = 4 SE +/- 5825.59, N = 3 SE +/- 5822.63, N = 3 SE +/- 4016.65, N = 3 SE +/- 877.17, N = 3 SE +/- 16428.27, N = 3 SE +/- 8733.96, N = 3 1892800.20 1829793.13 1750353.96 1730485.73 1636368.32 1499845.25 1447102.06 1445204.97
Blender Blend File: Classroom - Compute: CPU-Only EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8490H OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.6 Blend File: Classroom - Compute: CPU-Only Linux 6.6-rc1 Linux 6.5.1 15 30 45 60 75 SE +/- 0.01, N = 3 SE +/- 0.08, N = 3 SE +/- 0.05, N = 3 SE +/- 0.05, N = 3 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 SE +/- 0.09, N = 3 SE +/- 0.11, N = 3 31.45 31.99 39.97 40.63 40.59 40.59 69.14 69.19
TensorFlow Device: CPU - Batch Size: 16 - Model: ResNet-50 Xeon Platinum 8490H EPYC 9654 EPYC 9684X EPYC 9754 OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.12 Device: CPU - Batch Size: 16 - Model: ResNet-50 Linux 6.5.1 Linux 6.6-rc1 16 32 48 64 80 SE +/- 0.03, N = 3 SE +/- 0.05, N = 3 SE +/- 0.30, N = 3 SE +/- 0.14, N = 3 SE +/- 0.05, N = 3 SE +/- 0.21, N = 3 SE +/- 0.04, N = 3 SE +/- 0.16, N = 3 71.63 70.96 55.00 42.93 54.53 54.00 41.15 35.77
Stress-NG Test: Pthread Xeon Platinum 8490H EPYC 9654 EPYC 9684X EPYC 9754 OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.16.04 Test: Pthread Linux 6.5.1 Linux 6.6-rc1 30K 60K 90K 120K 150K SE +/- 707.60, N = 3 SE +/- 98.27, N = 3 SE +/- 50.81, N = 3 SE +/- 350.35, N = 3 SE +/- 168.38, N = 3 SE +/- 100.58, N = 3 SE +/- 44.92, N = 3 SE +/- 167.92, N = 3 117098.86 112147.13 96686.04 73831.46 94602.26 94337.45 87300.83 80145.36 1. (CXX) g++ options: -O2 -std=gnu99 -lc
Blender Blend File: Fishy Cat - Compute: CPU-Only EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8490H OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.6 Blend File: Fishy Cat - Compute: CPU-Only Linux 6.6-rc1 Linux 6.5.1 8 16 24 32 40 SE +/- 0.04, N = 3 SE +/- 0.09, N = 3 SE +/- 0.18, N = 3 SE +/- 0.04, N = 3 SE +/- 0.07, N = 3 SE +/- 0.09, N = 3 SE +/- 0.23, N = 3 SE +/- 0.10, N = 3 16.76 17.13 20.59 21.28 20.71 20.74 35.42 35.46
Blender Blend File: BMW27 - Compute: CPU-Only EPYC 9754 EPYC 9654 EPYC 9684X Xeon Platinum 8490H OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.6 Blend File: BMW27 - Compute: CPU-Only Linux 6.6-rc1 Linux 6.5.1 6 12 18 24 30 SE +/- 0.05, N = 3 SE +/- 0.03, N = 3 SE +/- 0.06, N = 3 SE +/- 0.09, N = 3 SE +/- 0.08, N = 3 SE +/- 0.04, N = 3 SE +/- 0.03, N = 3 SE +/- 0.04, N = 3 13.04 13.17 15.90 16.15 16.31 16.32 25.85 25.91
Phoronix Test Suite v10.8.4