Benchmarks for a future article.
v6.13 Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --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,rust --enable-libphobos-checking=release --enable-libstdcxx-backtrace --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-14-zdkDXv/gcc-14-14.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-14-zdkDXv/gcc-14-14.2.0/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 -vProcessor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xb002116Java Notes: OpenJDK Runtime Environment (build 21.0.5+11-Ubuntu-1ubuntu124.10)Python Notes: Python 3.12.7Security Notes: 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: 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; STIBP: always-on; RSB filling; PBRSB-eIBRS: Not affected; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected
v6.14 29 Jan Processor: AMD EPYC 9575F 64-Core @ 3.30GHz (64 Cores / 128 Threads), Motherboard: Supermicro Super Server H13SSL-N v1.01 (3.0 BIOS), Chipset: AMD 1Ah, Memory: 12 x 64GB DDR5-6000MT/s Micron MTC40F2046S1RC64BDY QSFF, Disk: 3201GB Micron_7450_MTFDKCB3T2TFS, Graphics: ASPEED, Network: 2 x Broadcom NetXtreme BCM5720 PCIe
OS: Ubuntu 24.10, Kernel: 6.13.0-phx (x86_64), Desktop: GNOME Shell 47.0, Display Server: X Server, Compiler: GCC 14.2.0, File-System: ext4, Screen Resolution: 1024x768
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-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,rust --enable-libphobos-checking=release --enable-libstdcxx-backtrace --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-14-zdkDXv/gcc-14-14.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-14-zdkDXv/gcc-14-14.2.0/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 -vProcessor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xb002116Java Notes: OpenJDK Runtime Environment (build 21.0.5+11-Ubuntu-1ubuntu124.10)Python Notes: Python 3.12.7Security Notes: 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 IBPB on VMEXIT only + 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; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected
Linux 6.14 Early Benchmarks AMD EPYC OpenBenchmarking.org Phoronix Test Suite AMD EPYC 9575F 64-Core @ 3.30GHz (64 Cores / 128 Threads) Supermicro Super Server H13SSL-N v1.01 (3.0 BIOS) AMD 1Ah 12 x 64GB DDR5-6000MT/s Micron MTC40F2046S1RC64BDY QSFF 3201GB Micron_7450_MTFDKCB3T2TFS ASPEED 2 x Broadcom NetXtreme BCM5720 PCIe Ubuntu 24.10 6.13.0-phx (x86_64) GNOME Shell 47.0 X Server GCC 14.2.0 ext4 1024x768 Processor Motherboard Chipset Memory Disk Graphics Network OS Kernel Desktop Display Server Compiler File-System Screen Resolution Linux 6.14 Early Benchmarks AMD EPYC Performance System Logs - Transparent Huge Pages: madvise - --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,rust --enable-libphobos-checking=release --enable-libstdcxx-backtrace --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-14-zdkDXv/gcc-14-14.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-14-zdkDXv/gcc-14-14.2.0/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 - Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xb002116 - OpenJDK Runtime Environment (build 21.0.5+11-Ubuntu-1ubuntu124.10) - Python 3.12.7 - v6.13: 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: 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; STIBP: always-on; RSB filling; PBRSB-eIBRS: Not affected; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected - v6.14 29 Jan: 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 IBPB on VMEXIT only + 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; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected
v6.13 vs. v6.14 29 Jan Comparison Phoronix Test Suite Baseline +2.5% +2.5% +5% +5% +7.5% +7.5% +10% +10% 10.1% 7.3% 3% 2.1% 2% A.L.S.E Writes Basic - CPU 5.4% NUMA 4% CPU BLAS - granite-3.0-3b-a800m-instruct-Q8_0 - P.P.2 3.5% Read While Writing 100 - 800 - Read Write 2.8% 100 - 800 - Read Write - Average Latency 2.8% TinyLlama-1.1B-Chat-v1.0 - CPU TinyLlama-1.1B-Chat-v1.0 - CPU - T.P.O.T DaCapo Benchmark Apache Cassandra RELION Stress-NG Llama.cpp RocksDB PostgreSQL PostgreSQL OpenVINO GenAI OpenVINO GenAI v6.13 v6.14 29 Jan
Linux 6.14 Early Benchmarks AMD EPYC dacapobench: Apache Lucene Search Engine cassandra: Writes stress-ng: NUMA llama-cpp: CPU BLAS - granite-3.0-3b-a800m-instruct-Q8_0 - Prompt Processing 2048 rocksdb: Read While Writing pgbench: 100 - 800 - Read Write pgbench: 100 - 800 - Read Write - Average Latency openvino-genai: TinyLlama-1.1B-Chat-v1.0 - CPU stress-ng: Socket Activity memcached: 1:10 dacapobench: Tradesoap namd: ATPase with 327,506 Atoms dacapobench: Apache Tomcat svt-av1: Preset 8 - Bosphorus 4K openfoam: drivaerFastback, Medium Mesh Size - Execution Time openfoam: drivaerFastback, Medium Mesh Size - Mesh Time llama-cpp: CPU BLAS - Mistral-7B-Instruct-v0.3-Q8_0 - Prompt Processing 2048 dacapobench: Eclipse pgbench: 100 - 800 - Read Only - Average Latency stress-ng: Futex stress-ng: Mutex memcached: 1:100 pgbench: 100 - 800 - Read Only rocksdb: Read Rand Write Rand stress-ng: Context Switching memcached: 1:5 llama-cpp: CPU BLAS - Llama-3.1-Tulu-3-8B-Q8_0 - Prompt Processing 2048 llama-cpp: CPU BLAS - granite-3.0-3b-a800m-instruct-Q8_0 - Text Generation 128 blender: Pabellon Barcelona - CPU-Only build-llvm: Ninja svt-av1: Preset 3 - Bosphorus 4K clickhouse: 100M Rows Hits Dataset, Third Run stress-ng: SENDFILE svt-av1: Preset 5 - Bosphorus 4K blender: Barbershop - CPU-Only llama-cpp: CPU BLAS - Llama-3.1-Tulu-3-8B-Q8_0 - Text Generation 128 blender: Junkshop - CPU-Only dacapobench: Apache Lucene Search Index clickhouse: 100M Rows Hits Dataset, Second Run clickhouse: 100M Rows Hits Dataset, First Run / Cold Cache tensorflow: CPU - 512 - ResNet-50 rocksdb: Update Rand build-linux-kernel: allmodconfig astcenc: Exhaustive dacapobench: Avrora AVR Simulation Framework java-jmh: Throughput rocksdb: Rand Read svt-av1: Preset 13 - Bosphorus 4K build-nodejs: Time To Compile namd: STMV with 1,066,628 Atoms blender: Classroom - CPU-Only gromacs: MPI CPU - water_GMX50_bare build-linux-kernel: defconfig astcenc: Very Thorough llama-cpp: CPU BLAS - Mistral-7B-Instruct-v0.3-Q8_0 - Text Generation 128 openvino-genai: TinyLlama-1.1B-Chat-v1.0 - CPU - Time Per Output Token openvino-genai: TinyLlama-1.1B-Chat-v1.0 - CPU - Time To First Token stress-ng: Pipe dacapobench: Tradebeans relion: Basic - CPU v6.13 v6.14 29 Jan 3981 461247 2144.57 448.22 11734903 128762 6.216 76.57 47642.26 7054421.88 2977 13.07300 1043 200.461 230.22542 89.311917 143.26 6244 0.165 4271894.47 42972205.27 13460038.76 4850455 7398435 52538077.77 3888552.68 143.45 118.47 46.85 101.046 17.033 803.99 2105661.95 60.574 146.69 49.91 19.97 2236 802.09 778.02 253.76 692027 191.014 6.2336 2183 316508832117.05 535255862 471.679 124.240 3.75749 41.15 14.585 21.756 10.1750 52.55 13.06 16.16 82801426.52 4870 165.832 3615 494806 2062.27 432.94 12083935 125263 6.387 78.14 46759.61 7187440.12 3033 12.83401 1060 197.302 226.99365 88.166581 145.08 6321 0.167 4224604.07 42507064.95 13605595.65 4801097 7348485 52207921.65 3911714.46 144.23 117.88 47.08 101.413 17.094 801.13 2099036.56 60.392 147.05 50.00 19.94 2233 802.93 778.75 253.53 691421 190.882 6.2293 2184 316375911973.06 535465377 471.825 124.276 3.75641 41.16 14.583 21.758 10.1756 52.55 12.80 16.10 81959177.61 4872 174.729 OpenBenchmarking.org
Stress-NG OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.18.09 Test: NUMA v6.13 v6.14 29 Jan 500 1000 1500 2000 2500 SE +/- 3.79, N = 3 SE +/- 4.06, N = 3 2144.57 2062.27 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lcrypt -ldl -ljpeg -lEGL -lGLESv2 -lmpfr -lgmp -lsctp -lz -lrt -lpthread -lc -std=gnu99 -O2 -fipa-pta -fivopts -fmodulo-sched
Llama.cpp OpenBenchmarking.org Tokens Per Second, More Is Better Llama.cpp b4397 Backend: CPU BLAS - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Prompt Processing 2048 v6.13 v6.14 29 Jan 100 200 300 400 500 SE +/- 3.77, N = 3 SE +/- 2.05, N = 3 448.22 432.94 1. (CXX) g++ options: -O3
RocksDB This is a benchmark of Meta/Facebook's RocksDB as an embeddable persistent key-value store for fast storage based on Google's LevelDB. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Op/s, More Is Better RocksDB 9.0 Test: Read While Writing v6.13 v6.14 29 Jan 3M 6M 9M 12M 15M SE +/- 94771.95, N = 3 SE +/- 116959.75, N = 3 11734903 12083935 1. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti
PostgreSQL OpenBenchmarking.org TPS, More Is Better PostgreSQL 17 Scaling Factor: 100 - Clients: 800 - Mode: Read Write v6.13 v6.14 29 Jan 30K 60K 90K 120K 150K SE +/- 895.52, N = 12 SE +/- 608.11, N = 3 128762 125263 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpq -lpgcommon -lpgport -lm
OpenBenchmarking.org ms, Fewer Is Better PostgreSQL 17 Scaling Factor: 100 - Clients: 800 - Mode: Read Write - Average Latency v6.13 v6.14 29 Jan 2 4 6 8 10 SE +/- 0.042, N = 12 SE +/- 0.031, N = 3 6.216 6.387 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpq -lpgcommon -lpgport -lm
Stress-NG OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.18.09 Test: Socket Activity v6.13 v6.14 29 Jan 10K 20K 30K 40K 50K SE +/- 61.51, N = 3 SE +/- 50.78, N = 3 47642.26 46759.61 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lcrypt -ldl -ljpeg -lEGL -lGLESv2 -lmpfr -lgmp -lsctp -lz -lrt -lpthread -lc -std=gnu99 -O2 -fipa-pta -fivopts -fmodulo-sched
Memcached Memcached is a high performance, distributed memory object caching system. This Memcached test profiles makes use of memtier_benchmark for excuting this CPU/memory-focused server benchmark. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Ops/sec, More Is Better Memcached 1.6.19 Set To Get Ratio: 1:10 v6.13 v6.14 29 Jan 1.5M 3M 4.5M 6M 7.5M SE +/- 11715.29, N = 3 SE +/- 21122.44, N = 3 7054421.88 7187440.12 1. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre
NAMD OpenBenchmarking.org ns/day, More Is Better NAMD 3.0 Input: ATPase with 327,506 Atoms v6.13 v6.14 29 Jan 3 6 9 12 15 SE +/- 0.09, N = 3 SE +/- 0.16, N = 4 13.07 12.83
SVT-AV1 OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 2.3 Encoder Mode: Preset 8 - Input: Bosphorus 4K v6.13 v6.14 29 Jan 40 80 120 160 200 SE +/- 0.02, N = 3 SE +/- 1.79, N = 3 200.46 197.30 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenFOAM OpenFOAM is the leading free, open-source software for computational fluid dynamics (CFD). This test profile currently uses the drivaerFastback test case for analyzing automotive aerodynamics or alternatively the older motorBike input. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better OpenFOAM 10 Input: drivaerFastback, Medium Mesh Size - Execution Time v6.13 v6.14 29 Jan 50 100 150 200 250 230.23 226.99 1. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfiniteVolume -lmeshTools -lparallel -llagrangian -lregionModels -lgenericPatchFields -lOpenFOAM -ldl -lm
OpenBenchmarking.org Seconds, Fewer Is Better OpenFOAM 10 Input: drivaerFastback, Medium Mesh Size - Mesh Time v6.13 v6.14 29 Jan 20 40 60 80 100 89.31 88.17 1. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfiniteVolume -lmeshTools -lparallel -llagrangian -lregionModels -lgenericPatchFields -lOpenFOAM -ldl -lm
Llama.cpp OpenBenchmarking.org Tokens Per Second, More Is Better Llama.cpp b4397 Backend: CPU BLAS - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Prompt Processing 2048 v6.13 v6.14 29 Jan 30 60 90 120 150 SE +/- 0.13, N = 3 SE +/- 0.59, N = 3 143.26 145.08 1. (CXX) g++ options: -O3
PostgreSQL OpenBenchmarking.org ms, Fewer Is Better PostgreSQL 17 Scaling Factor: 100 - Clients: 800 - Mode: Read Only - Average Latency v6.13 v6.14 29 Jan 0.0376 0.0752 0.1128 0.1504 0.188 SE +/- 0.000, N = 3 SE +/- 0.000, N = 3 0.165 0.167 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpq -lpgcommon -lpgport -lm
Stress-NG OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.18.09 Test: Futex v6.13 v6.14 29 Jan 900K 1800K 2700K 3600K 4500K SE +/- 31436.04, N = 3 SE +/- 47235.90, N = 3 4271894.47 4224604.07 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lcrypt -ldl -ljpeg -lEGL -lGLESv2 -lmpfr -lgmp -lsctp -lz -lrt -lpthread -lc -std=gnu99 -O2 -fipa-pta -fivopts -fmodulo-sched
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.18.09 Test: Mutex v6.13 v6.14 29 Jan 9M 18M 27M 36M 45M SE +/- 125947.68, N = 3 SE +/- 317020.74, N = 3 42972205.27 42507064.95 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lcrypt -ldl -ljpeg -lEGL -lGLESv2 -lmpfr -lgmp -lsctp -lz -lrt -lpthread -lc -std=gnu99 -O2 -fipa-pta -fivopts -fmodulo-sched
Memcached Memcached is a high performance, distributed memory object caching system. This Memcached test profiles makes use of memtier_benchmark for excuting this CPU/memory-focused server benchmark. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Ops/sec, More Is Better Memcached 1.6.19 Set To Get Ratio: 1:100 v6.13 v6.14 29 Jan 3M 6M 9M 12M 15M SE +/- 65975.03, N = 3 SE +/- 9305.87, N = 3 13460038.76 13605595.65 1. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre
PostgreSQL OpenBenchmarking.org TPS, More Is Better PostgreSQL 17 Scaling Factor: 100 - Clients: 800 - Mode: Read Only v6.13 v6.14 29 Jan 1000K 2000K 3000K 4000K 5000K SE +/- 12547.95, N = 3 SE +/- 2758.89, N = 3 4850455 4801097 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpq -lpgcommon -lpgport -lm
RocksDB This is a benchmark of Meta/Facebook's RocksDB as an embeddable persistent key-value store for fast storage based on Google's LevelDB. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Op/s, More Is Better RocksDB 9.0 Test: Read Random Write Random v6.13 v6.14 29 Jan 1.6M 3.2M 4.8M 6.4M 8M SE +/- 47503.65, N = 3 SE +/- 69319.41, N = 3 7398435 7348485 1. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti
Stress-NG OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.18.09 Test: Context Switching v6.13 v6.14 29 Jan 11M 22M 33M 44M 55M SE +/- 424783.11, N = 3 SE +/- 58765.40, N = 3 52538077.77 52207921.65 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lcrypt -ldl -ljpeg -lEGL -lGLESv2 -lmpfr -lgmp -lsctp -lz -lrt -lpthread -lc -std=gnu99 -O2 -fipa-pta -fivopts -fmodulo-sched
Memcached Memcached is a high performance, distributed memory object caching system. This Memcached test profiles makes use of memtier_benchmark for excuting this CPU/memory-focused server benchmark. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Ops/sec, More Is Better Memcached 1.6.19 Set To Get Ratio: 1:5 v6.13 v6.14 29 Jan 800K 1600K 2400K 3200K 4000K SE +/- 987.88, N = 3 SE +/- 5274.38, N = 3 3888552.68 3911714.46 1. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre
Llama.cpp OpenBenchmarking.org Tokens Per Second, More Is Better Llama.cpp b4397 Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 2048 v6.13 v6.14 29 Jan 30 60 90 120 150 SE +/- 0.48, N = 3 SE +/- 0.77, N = 3 143.45 144.23 1. (CXX) g++ options: -O3
OpenBenchmarking.org Tokens Per Second, More Is Better Llama.cpp b4397 Backend: CPU BLAS - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Text Generation 128 v6.13 v6.14 29 Jan 30 60 90 120 150 SE +/- 0.52, N = 3 SE +/- 0.46, N = 3 118.47 117.88 1. (CXX) g++ options: -O3
Blender OpenBenchmarking.org Seconds, Fewer Is Better Blender 4.3 Blend File: Pabellon Barcelona - Compute: CPU-Only v6.13 v6.14 29 Jan 11 22 33 44 55 SE +/- 0.07, N = 3 SE +/- 0.09, N = 3 46.85 47.08
SVT-AV1 OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 2.3 Encoder Mode: Preset 3 - Input: Bosphorus 4K v6.13 v6.14 29 Jan 4 8 12 16 20 SE +/- 0.03, N = 3 SE +/- 0.05, N = 3 17.03 17.09 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
ClickHouse ClickHouse is an open-source, high performance OLAP data management system. This test profile uses ClickHouse's standard benchmark recommendations per https://clickhouse.com/docs/en/operations/performance-test/ / https://github.com/ClickHouse/ClickBench/tree/main/clickhouse with the 100 million rows web analytics dataset. The reported value is the query processing time using the geometric mean of all separate queries performed as an aggregate. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Queries Per Minute, Geo Mean, More Is Better ClickHouse 22.12.3.5 100M Rows Hits Dataset, Third Run v6.13 v6.14 29 Jan 200 400 600 800 1000 SE +/- 1.04, N = 3 SE +/- 5.26, N = 3 803.99 801.13 MIN: 66.08 / MAX: 10000 MIN: 65.01 / MAX: 8571.43
Stress-NG OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.18.09 Test: SENDFILE v6.13 v6.14 29 Jan 500K 1000K 1500K 2000K 2500K SE +/- 3455.33, N = 3 SE +/- 4114.32, N = 3 2105661.95 2099036.56 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lcrypt -ldl -ljpeg -lEGL -lGLESv2 -lmpfr -lgmp -lsctp -lz -lrt -lpthread -lc -std=gnu99 -O2 -fipa-pta -fivopts -fmodulo-sched
SVT-AV1 OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 2.3 Encoder Mode: Preset 5 - Input: Bosphorus 4K v6.13 v6.14 29 Jan 14 28 42 56 70 SE +/- 0.27, N = 3 SE +/- 0.05, N = 3 60.57 60.39 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
Blender OpenBenchmarking.org Seconds, Fewer Is Better Blender 4.3 Blend File: Barbershop - Compute: CPU-Only v6.13 v6.14 29 Jan 30 60 90 120 150 SE +/- 0.07, N = 3 SE +/- 0.22, N = 3 146.69 147.05
Llama.cpp OpenBenchmarking.org Tokens Per Second, More Is Better Llama.cpp b4397 Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Text Generation 128 v6.13 v6.14 29 Jan 11 22 33 44 55 SE +/- 0.09, N = 3 SE +/- 0.03, N = 3 49.91 50.00 1. (CXX) g++ options: -O3
Blender OpenBenchmarking.org Seconds, Fewer Is Better Blender 4.3 Blend File: Junkshop - Compute: CPU-Only v6.13 v6.14 29 Jan 5 10 15 20 25 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 19.97 19.94
ClickHouse ClickHouse is an open-source, high performance OLAP data management system. This test profile uses ClickHouse's standard benchmark recommendations per https://clickhouse.com/docs/en/operations/performance-test/ / https://github.com/ClickHouse/ClickBench/tree/main/clickhouse with the 100 million rows web analytics dataset. The reported value is the query processing time using the geometric mean of all separate queries performed as an aggregate. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Queries Per Minute, Geo Mean, More Is Better ClickHouse 22.12.3.5 100M Rows Hits Dataset, Second Run v6.13 v6.14 29 Jan 200 400 600 800 1000 SE +/- 7.48, N = 3 SE +/- 5.08, N = 3 802.09 802.93 MIN: 65.43 / MAX: 8571.43 MIN: 66.01 / MAX: 8571.43
OpenBenchmarking.org Queries Per Minute, Geo Mean, More Is Better ClickHouse 22.12.3.5 100M Rows Hits Dataset, First Run / Cold Cache v6.13 v6.14 29 Jan 200 400 600 800 1000 SE +/- 6.64, N = 3 SE +/- 4.26, N = 3 778.02 778.75 MIN: 65.22 / MAX: 8571.43 MIN: 65.57 / MAX: 7500
TensorFlow This is a benchmark of the TensorFlow deep learning framework using the TensorFlow reference benchmarks (tensorflow/benchmarks with tf_cnn_benchmarks.py). Note with the Phoronix Test Suite there is also pts/tensorflow-lite for benchmarking the TensorFlow Lite binaries if desired for complementary metrics. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org images/sec, More Is Better TensorFlow 2.16.1 Device: CPU - Batch Size: 512 - Model: ResNet-50 v6.13 v6.14 29 Jan 60 120 180 240 300 SE +/- 0.26, N = 3 SE +/- 0.19, N = 3 253.76 253.53
RocksDB This is a benchmark of Meta/Facebook's RocksDB as an embeddable persistent key-value store for fast storage based on Google's LevelDB. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Op/s, More Is Better RocksDB 9.0 Test: Update Random v6.13 v6.14 29 Jan 150K 300K 450K 600K 750K SE +/- 200.83, N = 3 SE +/- 1034.41, N = 3 692027 691421 1. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti
ASTC Encoder OpenBenchmarking.org MT/s, More Is Better ASTC Encoder 5.0 Preset: Exhaustive v6.13 v6.14 29 Jan 2 4 6 8 10 SE +/- 0.0010, N = 3 SE +/- 0.0014, N = 3 6.2336 6.2293 1. (CXX) g++ options: -O3 -flto -pthread
RocksDB This is a benchmark of Meta/Facebook's RocksDB as an embeddable persistent key-value store for fast storage based on Google's LevelDB. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Op/s, More Is Better RocksDB 9.0 Test: Random Read v6.13 v6.14 29 Jan 110M 220M 330M 440M 550M SE +/- 1082500.13, N = 3 SE +/- 657385.47, N = 3 535255862 535465377 1. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti
SVT-AV1 OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 2.3 Encoder Mode: Preset 13 - Input: Bosphorus 4K v6.13 v6.14 29 Jan 100 200 300 400 500 SE +/- 4.76, N = 3 SE +/- 2.61, N = 3 471.68 471.83 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
NAMD OpenBenchmarking.org ns/day, More Is Better NAMD 3.0 Input: STMV with 1,066,628 Atoms v6.13 v6.14 29 Jan 0.8454 1.6908 2.5362 3.3816 4.227 SE +/- 0.01116, N = 3 SE +/- 0.00440, N = 3 3.75749 3.75641
Blender OpenBenchmarking.org Seconds, Fewer Is Better Blender 4.3 Blend File: Classroom - Compute: CPU-Only v6.13 v6.14 29 Jan 9 18 27 36 45 SE +/- 0.05, N = 3 SE +/- 0.01, N = 3 41.15 41.16
GROMACS The GROMACS (GROningen MAchine for Chemical Simulations) molecular dynamics package testing with the water_GMX50 data. This test profile allows selecting between CPU and GPU-based GROMACS builds. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Ns Per Day, More Is Better GROMACS 2024 Implementation: MPI CPU - Input: water_GMX50_bare v6.13 v6.14 29 Jan 4 8 12 16 20 SE +/- 0.08, N = 3 SE +/- 0.16, N = 3 14.59 14.58 1. (CXX) g++ options: -O3 -lm
ASTC Encoder OpenBenchmarking.org MT/s, More Is Better ASTC Encoder 5.0 Preset: Very Thorough v6.13 v6.14 29 Jan 3 6 9 12 15 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 10.18 10.18 1. (CXX) g++ options: -O3 -flto -pthread
Llama.cpp OpenBenchmarking.org Tokens Per Second, More Is Better Llama.cpp b4397 Backend: CPU BLAS - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Text Generation 128 v6.13 v6.14 29 Jan 12 24 36 48 60 SE +/- 0.15, N = 3 SE +/- 0.02, N = 3 52.55 52.55 1. (CXX) g++ options: -O3
OpenVINO GenAI Model: Phi-3-mini-128k-instruct-int4-ov - Device: CPU
v6.13: The test quit with a non-zero exit status. E: RuntimeError: Exception from src/inference/src/cpp/core.cpp:90:
v6.14 29 Jan: The test quit with a non-zero exit status. E: RuntimeError: Exception from src/inference/src/cpp/core.cpp:90:
Model: Falcon-7b-instruct-int4-ov - Device: CPU
v6.13: The test quit with a non-zero exit status. E: RuntimeError: Exception from src/inference/src/cpp/core.cpp:90:
v6.14 29 Jan: The test quit with a non-zero exit status. E: RuntimeError: Exception from src/inference/src/cpp/core.cpp:90:
Model: Gemma-7b-int4-ov - Device: CPU
v6.13: The test quit with a non-zero exit status. E: RuntimeError: Exception from src/inference/src/cpp/core.cpp:90:
v6.14 29 Jan: The test quit with a non-zero exit status. E: RuntimeError: Exception from src/inference/src/cpp/core.cpp:90:
Stress-NG OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.18.09 Test: Pipe v6.13 v6.14 29 Jan 20M 40M 60M 80M 100M SE +/- 1013892.33, N = 15 SE +/- 1868912.16, N = 15 82801426.52 81959177.61 1. (CXX) g++ options: -lm -laio -lapparmor -latomic -lcrypt -ldl -ljpeg -lEGL -lGLESv2 -lmpfr -lgmp -lsctp -lz -lrt -lpthread -lc -std=gnu99 -O2 -fipa-pta -fivopts -fmodulo-sched
RELION OpenBenchmarking.org Seconds, Fewer Is Better RELION 5.0 Test: Basic - Device: CPU v6.13 v6.14 29 Jan 40 80 120 160 200 SE +/- 3.02, N = 12 SE +/- 1.63, N = 12 165.83 174.73 1. (CXX) g++ options: -fPIC -std=c++14 -fopenmp -O3 -rdynamic -ldl -ltiff -lfftw3f -lfftw3 -lpng -ljpeg -lmpi_cxx -lmpi
v6.13 Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --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,rust --enable-libphobos-checking=release --enable-libstdcxx-backtrace --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-14-zdkDXv/gcc-14-14.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-14-zdkDXv/gcc-14-14.2.0/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 -vProcessor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xb002116Java Notes: OpenJDK Runtime Environment (build 21.0.5+11-Ubuntu-1ubuntu124.10)Python Notes: Python 3.12.7Security Notes: 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: 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; STIBP: always-on; RSB filling; PBRSB-eIBRS: Not affected; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 29 January 2025 19:58 by user phoronix.
v6.14 29 Jan Processor: AMD EPYC 9575F 64-Core @ 3.30GHz (64 Cores / 128 Threads), Motherboard: Supermicro Super Server H13SSL-N v1.01 (3.0 BIOS), Chipset: AMD 1Ah, Memory: 12 x 64GB DDR5-6000MT/s Micron MTC40F2046S1RC64BDY QSFF, Disk: 3201GB Micron_7450_MTFDKCB3T2TFS, Graphics: ASPEED, Network: 2 x Broadcom NetXtreme BCM5720 PCIe
OS: Ubuntu 24.10, Kernel: 6.13.0-phx (x86_64), Desktop: GNOME Shell 47.0, Display Server: X Server, Compiler: GCC 14.2.0, File-System: ext4, Screen Resolution: 1024x768
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-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,rust --enable-libphobos-checking=release --enable-libstdcxx-backtrace --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-14-zdkDXv/gcc-14-14.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-14-zdkDXv/gcc-14-14.2.0/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 -vProcessor Notes: Scaling Governor: acpi-cpufreq performance (Boost: Enabled) - CPU Microcode: 0xb002116Java Notes: OpenJDK Runtime Environment (build 21.0.5+11-Ubuntu-1ubuntu124.10)Python Notes: Python 3.12.7Security Notes: 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 IBPB on VMEXIT only + 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; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 30 January 2025 01:11 by user phoronix.