AMD EPYC Genoa-X Linux 6.9 Kernel Benchmarks

Benchmarks by Michael Larabel for a future article.

HTML result view exported from: https://openbenchmarking.org/result/2403282-NE-AMDEPYCGE98&hlc=1&sgm=1&ppt=D&grs&sor.

AMD EPYC Genoa-X Linux 6.9 Kernel BenchmarksProcessorMotherboardChipsetMemoryDiskGraphicsNetworkOSKernelCompilerFile-SystemScreen ResolutionLinux 6.8Linux 6.9 27 Mar2 x AMD EPYC 9684X 96-Core @ 2.55GHz (192 Cores / 384 Threads)AMD Titanite_4G (RTI1007B BIOS)AMD Device 14a41520GB3201GB Micron_7450_MTFDKCB3T2TFS + 257GB Flash DriveASPEEDBroadcom NetXtreme BCM5720 PCIeUbuntu 23.106.8.0-060800-generic (x86_64)GCC 13.2.0ext41920x12006.9.0-060900rc1daily20240327-generic (x86_64)OpenBenchmarking.orgKernel Details- Transparent Huge Pages: madviseCompiler 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-XYspKM/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-XYspKM/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 Processor Details- Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xa10113ePython Details- Python 3.11.6Security Details- Linux 6.8: 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 - Linux 6.9 27 Mar: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: 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

AMD EPYC Genoa-X Linux 6.9 Kernel Benchmarksstress-ng: Semaphoresrocksdb: Read Rand Write Randrocksdb: Overwriterocksdb: Update Randstress-ng: NUMAspeedb: Update Randllamafile: mistral-7b-instruct-v0.2.Q8_0 - CPUopenvino: Road Segmentation ADAS FP16-INT8 - CPUopenvino: Road Segmentation ADAS FP16-INT8 - CPUrocksdb: Rand Readopenvino: Vehicle Detection FP16-INT8 - CPUstress-ng: Context Switchingopenvino: Vehicle Detection FP16-INT8 - CPUhackbench: 32 - Processstress-ng: AVX-512 VNNIstress-ng: Vector Mathopenvino: Machine Translation EN To DE FP16 - CPUopenvino: Machine Translation EN To DE FP16 - CPUsvt-av1: Preset 12 - Bosphorus 4Kllamafile: llava-v1.5-7b-q4 - CPUrocksdb: Read While Writingstress-ng: System V Message PassingLinux 6.8Linux 6.9 27 Mar264112681.093182572381523375714646.3429374612.793864.6512.40104729086311218.1240075791.134.2610.83618968297.941306706.1944.131086.48159.51312.032066532425534145.87292541552.773484877416860406258691.5431379113.464063.1111.80109608195111719.9941854298.684.0810.42519594674.671349057.5142.821119.56164.30821.102241814028384870.15OpenBenchmarking.org

Stress-NG

Test: Semaphores

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: SemaphoresLinux 6.9 27 MarLinux 6.860M120M180M240M300MSE +/- 1842285.12, N = 3SE +/- 1107000.14, N = 3292541552.77264112681.091. (CXX) g++ options: -O2 -std=gnu99 -lc

RocksDB

Test: Read Random Write Random

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: Read Random Write RandomLinux 6.9 27 MarLinux 6.8700K1400K2100K2800K3500KSE +/- 3987.34, N = 3SE +/- 3761.31, N = 3348487731825721. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

RocksDB

Test: Overwrite

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: OverwriteLinux 6.9 27 MarLinux 6.890K180K270K360K450KSE +/- 1320.84, N = 3SE +/- 180.51, N = 34168603815231. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

RocksDB

Test: Update Random

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: Update RandomLinux 6.9 27 MarLinux 6.890K180K270K360K450KSE +/- 1155.95, N = 3SE +/- 379.66, N = 34062583757141. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

Stress-NG

Test: NUMA

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: NUMALinux 6.9 27 MarLinux 6.8150300450600750SE +/- 0.95, N = 3SE +/- 2.59, N = 3691.54646.341. (CXX) g++ options: -O2 -std=gnu99 -lc

Speedb

Test: Update Random

OpenBenchmarking.orgOp/s, More Is BetterSpeedb 2.7Test: Update RandomLinux 6.9 27 MarLinux 6.870K140K210K280K350KSE +/- 2172.75, N = 3SE +/- 945.90, N = 33137912937461. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

Llamafile

Test: mistral-7b-instruct-v0.2.Q8_0 - Acceleration: CPU

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.6Test: mistral-7b-instruct-v0.2.Q8_0 - Acceleration: CPULinux 6.9 27 MarLinux 6.83691215SE +/- 0.04, N = 3SE +/- 0.04, N = 313.4612.79

OpenVINO

Model: Road Segmentation ADAS FP16-INT8 - Device: CPU

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Road Segmentation ADAS FP16-INT8 - Device: CPULinux 6.9 27 MarLinux 6.89001800270036004500SE +/- 3.68, N = 3SE +/- 1.98, N = 34063.113864.651. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenVINO

Model: Road Segmentation ADAS FP16-INT8 - Device: CPU

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Road Segmentation ADAS FP16-INT8 - Device: CPULinux 6.9 27 MarLinux 6.83691215SE +/- 0.01, N = 3SE +/- 0.01, N = 311.8012.40MIN: 11.09 / MAX: 37.4MIN: 11.56 / MAX: 38.831. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

RocksDB

Test: Random Read

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: Random ReadLinux 6.9 27 MarLinux 6.8200M400M600M800M1000MSE +/- 2941599.90, N = 3SE +/- 4458137.36, N = 3109608195110472908631. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

OpenVINO

Model: Vehicle Detection FP16-INT8 - Device: CPU

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Vehicle Detection FP16-INT8 - Device: CPULinux 6.9 27 MarLinux 6.83K6K9K12K15KSE +/- 24.42, N = 3SE +/- 5.76, N = 311719.9911218.121. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Stress-NG

Test: Context Switching

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Context SwitchingLinux 6.9 27 MarLinux 6.89M18M27M36M45MSE +/- 110197.66, N = 3SE +/- 130933.36, N = 341854298.6840075791.131. (CXX) g++ options: -O2 -std=gnu99 -lc

OpenVINO

Model: Vehicle Detection FP16-INT8 - Device: CPU

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Vehicle Detection FP16-INT8 - Device: CPULinux 6.9 27 MarLinux 6.80.95851.9172.87553.8344.7925SE +/- 0.01, N = 3SE +/- 0.00, N = 34.084.26MIN: 3.64 / MAX: 20.85MIN: 3.76 / MAX: 20.91. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

Hackbench

Count: 32 - Type: Process

OpenBenchmarking.orgSeconds, Fewer Is BetterHackbenchCount: 32 - Type: ProcessLinux 6.9 27 MarLinux 6.83691215SE +/- 0.14, N = 3SE +/- 0.13, N = 410.4310.841. (CC) gcc options: -lpthread

Stress-NG

Test: AVX-512 VNNI

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: AVX-512 VNNILinux 6.9 27 MarLinux 6.84M8M12M16M20MSE +/- 25357.28, N = 3SE +/- 19220.66, N = 319594674.6718968297.941. (CXX) g++ options: -O2 -std=gnu99 -lc

Stress-NG

Test: Vector Math

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: Vector MathLinux 6.9 27 MarLinux 6.8300K600K900K1200K1500KSE +/- 1326.11, N = 3SE +/- 1312.02, N = 31349057.511306706.191. (CXX) g++ options: -O2 -std=gnu99 -lc

OpenVINO

Model: Machine Translation EN To DE FP16 - Device: CPU

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO 2024.0Model: Machine Translation EN To DE FP16 - Device: CPULinux 6.9 27 MarLinux 6.81020304050SE +/- 0.02, N = 3SE +/- 0.05, N = 342.8244.13MIN: 34.68 / MAX: 192.44MIN: 36.04 / MAX: 216.031. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

OpenVINO

Model: Machine Translation EN To DE FP16 - Device: CPU

OpenBenchmarking.orgFPS, More Is BetterOpenVINO 2024.0Model: Machine Translation EN To DE FP16 - Device: CPULinux 6.9 27 MarLinux 6.82004006008001000SE +/- 0.57, N = 3SE +/- 1.18, N = 31119.561086.481. (CXX) g++ options: -fPIC -fsigned-char -ffunction-sections -fdata-sections -O3 -fno-strict-overflow -fwrapv -shared -ldl

SVT-AV1

Encoder Mode: Preset 12 - Input: Bosphorus 4K

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.0Encoder Mode: Preset 12 - Input: Bosphorus 4KLinux 6.9 27 MarLinux 6.84080120160200SE +/- 0.71, N = 3SE +/- 1.75, N = 3164.31159.511. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

Llamafile

Test: llava-v1.5-7b-q4 - Acceleration: CPU

OpenBenchmarking.orgTokens Per Second, More Is BetterLlamafile 0.6Test: llava-v1.5-7b-q4 - Acceleration: CPULinux 6.9 27 MarLinux 6.8510152025SE +/- 0.49, N = 15SE +/- 0.42, N = 1521.1012.03

RocksDB

Test: Read While Writing

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: Read While WritingLinux 6.9 27 MarLinux 6.85M10M15M20M25MSE +/- 390827.28, N = 12SE +/- 155882.11, N = 322418140206653241. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread

Stress-NG

Test: System V Message Passing

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.16.04Test: System V Message PassingLinux 6.9 27 MarLinux 6.86M12M18M24M30MSE +/- 9139.64, N = 3SE +/- 770160.74, N = 1228384870.1525534145.871. (CXX) g++ options: -O2 -std=gnu99 -lc

Geometric Mean Of All Test Results

Result Composite - AMD EPYC Genoa-X Linux 6.9 Kernel Benchmarks

OpenBenchmarking.orgGeometric Mean, More Is BetterGeometric Mean Of All Test ResultsResult Composite - AMD EPYC Genoa-X Linux 6.9 Kernel BenchmarksLinux 6.9 27 MarLinux 6.86K12K18K24K30K30087.8927772.26

Phoronix Test Suite v10.8.4