Linux 6.6 vs. Linux 6.12 LTS Comparison

Linux 6.12 LTS benchmarks by Michael Larabel for a future article.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 2412221-PTS-LINUXLTS14
Jump To Table - Results

View

Do Not Show Noisy Results
Do Not Show Results With Incomplete Data
Do Not Show Results With Little Change/Spread
List Notable Results
Show Result Confidence Charts
Allow Limiting Results To Certain Suite(s)

Statistics

Show Overall Harmonic Mean(s)
Show Overall Geometric Mean
Show Wins / Losses Counts (Pie Chart)
Normalize Results
Remove Outliers Before Calculating Averages

Graph Settings

Force Line Graphs Where Applicable
Convert To Scalar Where Applicable
Prefer Vertical Bar Graphs

Multi-Way Comparison

Condense Multi-Option Tests Into Single Result Graphs

Table

Show Detailed System Result Table

Run Management

Highlight
Result
Toggle/Hide
Result
Result
Identifier
Performance Per
Dollar
Date
Run
  Test
  Duration
Linux 6.6 LTS
December 21
  6 Hours, 11 Minutes
Linux 6.12 LTS
December 22
  6 Hours, 14 Minutes
Invert Behavior (Only Show Selected Data)
  6 Hours, 13 Minutes
Only show results matching title/arguments (delimit multiple options with a comma):
Do not show results matching title/arguments (delimit multiple options with a comma):


Linux 6.6 vs. Linux 6.12 LTS ComparisonProcessorMotherboardChipsetMemoryDiskGraphicsAudioMonitorNetworkOSKernelDesktopDisplay ServerOpenGLCompilerFile-SystemScreen ResolutionLinux 6.6 LTSLinux 6.12 LTSAMD Ryzen Threadripper 7980X 64-Cores @ 5.37GHz (64 Cores / 128 Threads)System76 Thelio Major (FA Z5 BIOS)AMD Device 14a44 x 32GB DDR5-4800MT/s Micron MTC20F1045S1RC48BA21000GB CT1000T700SSD5AMD Radeon RX 6700 XT 12GBAMD Device 14ccDELL P2415QAquantia AQC113C NBase-T/IEEE + Realtek RTL8125 2.5GbE + Intel Wi-Fi 6EUbuntu 24.046.6.63-060663-generic (x86_64)GNOME Shell 46.0X Server + Wayland4.6 Mesa 24.0.9-0ubuntu0.2 (LLVM 17.0.6 DRM 3.54)GCC 13.3.0ext41920x12006.12.3-061203-generic (x86_64)4.6 Mesa 24.0.9-0ubuntu0.2 (LLVM 17.0.6 DRM 3.59)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-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-13-fG75Ri/gcc-13-13.3.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-fG75Ri/gcc-13-13.3.0/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- Linux 6.6 LTS: Scaling Governor: amd-pstate-epp powersave (EPP: balance_performance) - CPU Microcode: 0xa108105- Linux 6.12 LTS: Scaling Governor: amd-pstate-epp powersave (Boost: Enabled EPP: balance_performance) - CPU Microcode: 0xa108105 Graphics Details- Linux 6.6 LTS: BAR1 / Visible vRAM Size: 12272 MB - vBIOS Version: 113-D5121100-101- Linux 6.12 LTS: BAR1 / Visible vRAM Size: 12272 MBJava Details- OpenJDK Runtime Environment (build 21.0.5+11-Ubuntu-1ubuntu124.04)Python Details- Python 3.12.3Security Details- 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; BHI: Not affected + srbds: Not affected + tsx_async_abort: Not affected

Linux 6.6 LTS vs. Linux 6.12 LTS ComparisonPhoronix Test SuiteBaseline+22.2%+22.2%+44.4%+44.4%+66.6%+66.6%88.6%45.5%45.3%44.3%43.9%41.7%24.4%19.8%9.2%7.6%7%6.6%5.4%4.8%4.2%3.9%3.6%3.5%2.8%2.7%2.6%2.6%2.5%2.4%2.4%2.4%2.4%2.3%2.3%2.3%2.2%2%7.6%2.4%Context SwitchingA.X.5.DA.X.5.EA.X.2.DA.X.2.ESENDFILECPU BLAS - granite-3.0-3b-a800m-instruct-Q8_0 - P.P.1Z.1.2.B.I.PCPU BLAS - Llama-3.1-Tulu-3-8B-Q8_0 - P.P.1Wavelet Volume - 600 - 1920 x 1200Apache Tomcat7.6%A.L.S.ERead While Writing6.9%16 - ThreadTradesoap6%CPU BLAS - Mistral-7B-Instruct-v0.3-Q8_0 - P.P.1WritesSpring BootPreset 13 - Bosphorus 4K4%1:1003.8%Tradebeans3.7%T.X.5.E3.6%Timed Time - Size 1,000Socket Activity1:103.2%allmodconfig1920 x 1200 - Medium2.7%100 - 1000 - Read Write - Average Latency100 - 1000 - Read Write32 - ProcessA.w.3.5.AApache Xalan XSLT2.5%Many Spheres - 600 - 1920 x 1200defconfig100 - 800 - Read WriteT.X.2.E2.4%100 - 800 - Read Write - Average LatencyT.X.5.D2.4%GraphChi2.3%1920 x 1200 - Fullscreen - Low - OpenGL100 - 1000 - Read OnlyMixed Scheduler100 - 1000 - Read Only - Average Latency1920 x 1200 - Ultra2.2%F.P.F2.1%T.X.2.D2.1%H2 Database Engine2%100 - 800 - Read OnlyWavelet Volume - 600 - 1920 x 1200Many Spheres - 600 - 1920 x 1200Stress-NGCryptsetupCryptsetupCryptsetupCryptsetupStress-NGLlama.cppDaCapo BenchmarkLlama.cppParaViewDaCapo BenchmarkDaCapo BenchmarkRocksDBHackbenchDaCapo BenchmarkLlama.cppApache CassandraDaCapo BenchmarkSVT-AV1MemcachedWireGuard + Linux Networking Stack Stress TestDaCapo BenchmarkCryptsetupSQLite SpeedtestStress-NGMemcachedTimed Linux Kernel CompilationUnvanquishedPostgreSQLPostgreSQLHackbenchNAMDDaCapo BenchmarkParaViewTimed Linux Kernel CompilationPostgreSQLCryptsetupPostgreSQLCryptsetupDaCapo BenchmarkUnigine SuperpositionPostgreSQLStress-NGPostgreSQLUnvanquishedDaCapo BenchmarkCryptsetupDaCapo BenchmarkPostgreSQLParaViewParaViewLinux 6.6 LTSLinux 6.12 LTS

Linux 6.6 vs. Linux 6.12 LTS Comparisonsrsran: PUSCH Processor Benchmark, Throughput Totalwireguard: llama-cpp: CPU BLAS - Llama-3.1-Tulu-3-8B-Q8_0 - Prompt Processing 1024build-linux-kernel: allmodconfigdacapobench: Tradebeansrocksdb: Read While Writingunigine-super: 1920 x 1200 - Fullscreen - High - OpenGLunigine-super: 1920 x 1200 - Fullscreen - Medium - OpenGLunigine-super: 1920 x 1200 - Fullscreen - Low - OpenGLllama-cpp: CPU BLAS - Mistral-7B-Instruct-v0.3-Q8_0 - Prompt Processing 1024svt-av1: Preset 3 - Bosphorus 4Kgravitymark: 1920 x 1200 - OpenGLgravitymark: 1920 x 1200 - Vulkanblender: Barbershop - CPU-Onlystress-ng: SENDFILEpgbench: 100 - 800 - Read Only - Average Latencypgbench: 100 - 800 - Read Onlypgbench: 100 - 1000 - Read Write - Average Latencypgbench: 100 - 1000 - Read Writepgbench: 100 - 1000 - Read Only - Average Latencypgbench: 100 - 1000 - Read Onlypgbench: 100 - 800 - Read Write - Average Latencypgbench: 100 - 800 - Read Writecassandra: Writesdacapobench: Tradesoapstress-ng: Context Switchingv-ray: CPUmemcached: 1:100memcached: 1:5memcached: 1:10indigobench: CPU - Bedroomindigobench: CPU - Supercarrocksdb: Update Randrocksdb: Read Rand Write Randrocksdb: Rand Readopenvino-genai: Gemma-7b-int4-ov - CPU - Time Per Output Tokenopenvino-genai: Gemma-7b-int4-ov - CPU - Time To First Tokenopenvino-genai: Gemma-7b-int4-ov - CPUblender: Pabellon Barcelona - CPU-Onlysvt-av1: Preset 5 - Bosphorus 4Klitert: Inception V4sqlite-speedtest: Timed Time - Size 1,000litert: Mobilenet Floatlitert: SqueezeNetopenvino-genai: Falcon-7b-instruct-int4-ov - CPU - Time Per Output Tokenopenvino-genai: Falcon-7b-instruct-int4-ov - CPU - Time To First Tokenopenvino-genai: Falcon-7b-instruct-int4-ov - CPUblender: Classroom - CPU-Onlystress-ng: Socket Activitybuild-linux-kernel: defconfigllama-cpp: CPU BLAS - Llama-3.1-Tulu-3-8B-Q8_0 - Text Generation 128cryptsetup: Twofish-XTS 512b Decryptioncryptsetup: Twofish-XTS 512b Encryptioncryptsetup: Serpent-XTS 512b Decryptioncryptsetup: Serpent-XTS 512b Encryptioncryptsetup: AES-XTS 512b Decryptioncryptsetup: AES-XTS 512b Encryptioncryptsetup: Twofish-XTS 256b Decryptioncryptsetup: Twofish-XTS 256b Encryptioncryptsetup: Serpent-XTS 256b Decryptioncryptsetup: Serpent-XTS 256b Encryptioncryptsetup: AES-XTS 256b Decryptioncryptsetup: AES-XTS 256b Encryptioncryptsetup: PBKDF2-whirlpoolcryptsetup: PBKDF2-sha512llama-cpp: CPU BLAS - Mistral-7B-Instruct-v0.3-Q8_0 - Text Generation 128namd: STMV with 1,066,628 Atomsnamd: ATPase with 327,506 Atomsgromacs: MPI CPU - water_GMX50_barellama-cpp: CPU BLAS - granite-3.0-3b-a800m-instruct-Q8_0 - Prompt Processing 1024stress-ng: Mixed Schedulerdacapobench: Eclipseparaview: Wavelet Volume - 600 - 1920 x 1200paraview: Wavelet Volume - 600 - 1920 x 1200openvino-genai: TinyLlama-1.1B-Chat-v1.0 - CPU - Time Per Output Tokenopenvino-genai: TinyLlama-1.1B-Chat-v1.0 - CPU - Time To First Tokenopenvino-genai: TinyLlama-1.1B-Chat-v1.0 - CPUunvanquished: 1920 x 1200 - Ultradacapobench: H2 Database Engineblender: Junkshop - CPU-Onlyblender: Fishy Cat - CPU-Onlydacapobench: jMonkeyEnginedacapobench: Jythonhackbench: 32 - Processdacapobench: Spring Bootparaview: Many Spheres - 600 - 1920 x 1200paraview: Many Spheres - 600 - 1920 x 1200svt-av1: Preset 8 - Bosphorus 4Kopenvino-genai: Phi-3-mini-128k-instruct-int4-ov - CPU - Time Per Output Tokenopenvino-genai: Phi-3-mini-128k-instruct-int4-ov - CPU - Time To First Tokenopenvino-genai: Phi-3-mini-128k-instruct-int4-ov - CPUdacapobench: Apache Kafkabreaking-limit: 1920 x 1200 - Offdacapobench: Apache Lucene Search Indexsvt-av1: Preset 13 - Bosphorus 4Kblender: BMW27 - CPU-Onlyunvanquished: 1920 x 1200 - Highdacapobench: BioJava Biological Data Frameworkdacapobench: Avrora AVR Simulation Frameworkdacapobench: PMD Source Code Analyzerunvanquished: 1920 x 1200 - Mediumx265: Bosphorus 4Kdacapobench: Batik SVG Toolkitdacapobench: Zxing 1D/2D Barcode Image Processingdacapobench: GraphChihackbench: 16 - Threadhackbench: 16 - Processdacapobench: Apache Lucene Search Enginellama-cpp: CPU BLAS - granite-3.0-3b-a800m-instruct-Q8_0 - Text Generation 128paraview: Wavelet Contour - 600 - 1920 x 1200paraview: Wavelet Contour - 600 - 1920 x 1200dacapobench: Apache Tomcatx265: Bosphorus 1080pdacapobench: Apache Xalan XSLTdacapobench: FOP Print Formatterhackbench: 32 - ThreadLinux 6.6 LTSLinux 6.12 LTS5811.7273.221110.82278.3221208111791885113.3157.1239.4111.1414.140198.2217.1167.97710089.350.22363806030.795324760.277361010924.33632873399076555011705033.611350339031077.633790466.166814926.216.14535.793531463363718844022822048.4257.0320.6654.2448.37526555.251.5632129.643230.9340.0049.8025.0046.4029743.6528.66115.81610.2597.7950.1958.43658.93662.2608.4592.3947.1946.03974.33972.8942440262702916.691.643746.272077.544186.8338895.0071696582.932411.4322.2224.7145.00687.2199424.4623.636863429520.335208511842.936118.13141.96928.0335.3735.685451132.372647232.45017.57691.0469648441155707.941.001024485240811.40311.138281777.763019.322289.731157116.927264305875.7283.732120.98270.6561252311028127111.7155.1244.8117.1113.971197.6218.4166.471006166.300.216370981230.000333340.271369149823.76533664418052588422070300.121357959379299.983741056.006600797.8316.16535.456526844358795243961448748.7057.1520.5354.1548.17626660.649.7632128.893287.5939.8749.7325.0846.0630773.3227.98615.51595.9576.8936.5946.55324.15320.7596.1578.4939.9946.25735.25714.9935732261177616.551.633246.427747.558232.3539770.2270637085.921442.8722.1524.5545.15672.7203424.1423.586880426719.817200112130.232120.99140.93828.0935.5735.605452131.342670223.54417.65678.6476849231173689.240.931012405246310.70110.975263276.813049.796292.651245116.41744439OpenBenchmarking.org

srsRAN Project

OpenBenchmarking.orgMbps, More Is BettersrsRAN Project 24.10Test: PUSCH Processor Benchmark, Throughput TotalLinux 6.12 LTSLinux 6.6 LTS13002600390052006500SE +/- 65.75, N = 14SE +/- 1.12, N = 35875.75811.71. (CXX) g++ options: -O3 -march=native -mtune=generic -fno-trapping-math -fno-math-errno -ldl

WireGuard + Linux Networking Stack Stress Test

This is a benchmark of the WireGuard secure VPN tunnel and Linux networking stack stress test. The test runs on the local host but does require root permissions to run. The way it works is it creates three namespaces. ns0 has a loopback device. ns1 and ns2 each have wireguard devices. Those two wireguard devices send traffic through the loopback device of ns0. The end result of this is that tests wind up testing encryption and decryption at the same time -- a pretty CPU and scheduler-heavy workflow. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterWireGuard + Linux Networking Stack Stress TestLinux 6.12 LTSLinux 6.6 LTS60120180240300SE +/- 0.34, N = 3SE +/- 0.02, N = 3283.73273.22

Llama.cpp

OpenBenchmarking.orgTokens Per Second, More Is BetterLlama.cpp b4154Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 1024Linux 6.12 LTSLinux 6.6 LTS306090120150SE +/- 1.07, N = 15SE +/- 1.50, N = 15120.98110.821. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -fopenmp -march=native -mtune=native -lopenblas

Timed Linux Kernel Compilation

This test times how long it takes to build the Linux kernel in a default configuration (defconfig) for the architecture being tested or alternatively an allmodconfig for building all possible kernel modules for the build. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 6.8Build: allmodconfigLinux 6.12 LTSLinux 6.6 LTS60120180240300SE +/- 0.19, N = 3SE +/- 0.29, N = 3270.66278.32

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: TradebeansLinux 6.12 LTSLinux 6.6 LTS3K6K9K12K15KSE +/- 396.80, N = 15SE +/- 385.34, N = 151252312081

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.orgOp/s, More Is BetterRocksDB 9.0Test: Read While WritingLinux 6.12 LTSLinux 6.6 LTS3M6M9M12M15MSE +/- 72612.28, N = 14SE +/- 117684.65, N = 611028127117918851. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti

Unigine Superposition

This test calculates the average frame-rate within the Superposition demo for the Unigine engine, released in 2017. This engine is extremely demanding on the system's graphics card. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterUnigine Superposition 1.0Resolution: 1920 x 1200 - Mode: Fullscreen - Quality: High - Renderer: OpenGLLinux 6.12 LTSLinux 6.6 LTS306090120150SE +/- 0.00, N = 3SE +/- 0.03, N = 3111.7113.3MAX: 137.2MAX: 138.7

OpenBenchmarking.orgFrames Per Second, More Is BetterUnigine Superposition 1.0Resolution: 1920 x 1200 - Mode: Fullscreen - Quality: Medium - Renderer: OpenGLLinux 6.12 LTSLinux 6.6 LTS306090120150SE +/- 0.06, N = 3SE +/- 0.12, N = 3155.1157.1MAX: 202.4MAX: 203.6

OpenBenchmarking.orgFrames Per Second, More Is BetterUnigine Superposition 1.0Resolution: 1920 x 1200 - Mode: Fullscreen - Quality: Low - Renderer: OpenGLLinux 6.12 LTSLinux 6.6 LTS50100150200250SE +/- 1.68, N = 3SE +/- 1.52, N = 3244.8239.4MAX: 374.8MAX: 350.7

Llama.cpp

OpenBenchmarking.orgTokens Per Second, More Is BetterLlama.cpp b4154Backend: CPU BLAS - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Prompt Processing 1024Linux 6.12 LTSLinux 6.6 LTS306090120150SE +/- 1.36, N = 4SE +/- 1.45, N = 15117.11111.141. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -fopenmp -march=native -mtune=native -lopenblas

SVT-AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.3Encoder Mode: Preset 3 - Input: Bosphorus 4KLinux 6.12 LTSLinux 6.6 LTS48121620SE +/- 0.03, N = 3SE +/- 0.03, N = 313.9714.141. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

GravityMark

OpenBenchmarking.orgFrames Per Second, More Is BetterGravityMark 1.87Resolution: 1920 x 1200 - Renderer: OpenGLLinux 6.12 LTSLinux 6.6 LTS4080120160200SE +/- 0.17, N = 3SE +/- 0.03, N = 3197.6198.2

OpenBenchmarking.orgFrames Per Second, More Is BetterGravityMark 1.87Resolution: 1920 x 1200 - Renderer: VulkanLinux 6.12 LTSLinux 6.6 LTS50100150200250SE +/- 0.03, N = 3SE +/- 0.23, N = 3218.4217.1

Blender

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.3Blend File: Barbershop - Compute: CPU-OnlyLinux 6.12 LTSLinux 6.6 LTS4080120160200SE +/- 0.23, N = 3SE +/- 0.62, N = 3166.47167.97

Stress-NG

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.17.08Test: SENDFILELinux 6.12 LTSLinux 6.6 LTS200K400K600K800K1000KSE +/- 10129.48, N = 15SE +/- 11884.90, N = 151006166.30710089.351. (CXX) g++ options: -O2 -std=gnu99 -lc -lm

PostgreSQL

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 17Scaling Factor: 100 - Clients: 800 - Mode: Read Only - Average LatencyLinux 6.12 LTSLinux 6.6 LTS0.04950.0990.14850.1980.2475SE +/- 0.000, N = 3SE +/- 0.000, N = 30.2160.2201. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpq -lpgcommon -lpgport -lm

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL 17Scaling Factor: 100 - Clients: 800 - Mode: Read OnlyLinux 6.12 LTSLinux 6.6 LTS800K1600K2400K3200K4000KSE +/- 7725.31, N = 3SE +/- 3948.30, N = 3370981236380601. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpq -lpgcommon -lpgport -lm

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 17Scaling Factor: 100 - Clients: 1000 - Mode: Read Write - Average LatencyLinux 6.12 LTSLinux 6.6 LTS714212835SE +/- 0.08, N = 3SE +/- 0.21, N = 330.0030.801. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpq -lpgcommon -lpgport -lm

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL 17Scaling Factor: 100 - Clients: 1000 - Mode: Read WriteLinux 6.12 LTSLinux 6.6 LTS7K14K21K28K35KSE +/- 89.94, N = 3SE +/- 219.18, N = 333334324761. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpq -lpgcommon -lpgport -lm

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 17Scaling Factor: 100 - Clients: 1000 - Mode: Read Only - Average LatencyLinux 6.12 LTSLinux 6.6 LTS0.06230.12460.18690.24920.3115SE +/- 0.001, N = 3SE +/- 0.001, N = 30.2710.2771. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpq -lpgcommon -lpgport -lm

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL 17Scaling Factor: 100 - Clients: 1000 - Mode: Read OnlyLinux 6.12 LTSLinux 6.6 LTS800K1600K2400K3200K4000KSE +/- 6510.05, N = 3SE +/- 6092.17, N = 3369149836101091. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpq -lpgcommon -lpgport -lm

OpenBenchmarking.orgms, Fewer Is BetterPostgreSQL 17Scaling Factor: 100 - Clients: 800 - Mode: Read Write - Average LatencyLinux 6.12 LTSLinux 6.6 LTS612182430SE +/- 0.06, N = 3SE +/- 0.01, N = 323.7724.341. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpq -lpgcommon -lpgport -lm

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL 17Scaling Factor: 100 - Clients: 800 - Mode: Read WriteLinux 6.12 LTSLinux 6.6 LTS7K14K21K28K35KSE +/- 81.42, N = 3SE +/- 7.12, N = 333664328731. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpq -lpgcommon -lpgport -lm

Apache Cassandra

OpenBenchmarking.orgOp/s, More Is BetterApache Cassandra 5.0Test: WritesLinux 6.12 LTSLinux 6.6 LTS90K180K270K360K450KSE +/- 4663.86, N = 3SE +/- 2581.70, N = 3418052399076

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: TradesoapLinux 6.12 LTSLinux 6.6 LTS13002600390052006500SE +/- 76.62, N = 15SE +/- 142.65, N = 1558845550

Stress-NG

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.17.08Test: Context SwitchingLinux 6.12 LTSLinux 6.6 LTS5M10M15M20M25MSE +/- 82130.66, N = 3SE +/- 129368.33, N = 1522070300.1211705033.611. (CXX) g++ options: -O2 -std=gnu99 -lc -lm

Chaos Group V-RAY

This is a test of Chaos Group's V-RAY benchmark. V-RAY is a commercial renderer that can integrate with various creator software products like SketchUp and 3ds Max. The V-RAY benchmark is standalone and supports CPU and NVIDIA CUDA/RTX based rendering. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgvsamples, More Is BetterChaos Group V-RAY 6.0Mode: CPULinux 6.12 LTSLinux 6.6 LTS30K60K90K120K150KSE +/- 649.45, N = 3SE +/- 165.30, N = 3135795135033

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.orgOps/sec, More Is BetterMemcached 1.6.19Set To Get Ratio: 1:100Linux 6.12 LTSLinux 6.6 LTS2M4M6M8M10MSE +/- 19495.81, N = 3SE +/- 27494.50, N = 39379299.989031077.631. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

OpenBenchmarking.orgOps/sec, More Is BetterMemcached 1.6.19Set To Get Ratio: 1:5Linux 6.12 LTSLinux 6.6 LTS800K1600K2400K3200K4000KSE +/- 2828.47, N = 3SE +/- 17586.46, N = 33741056.003790466.161. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

OpenBenchmarking.orgOps/sec, More Is BetterMemcached 1.6.19Set To Get Ratio: 1:10Linux 6.12 LTSLinux 6.6 LTS1.5M3M4.5M6M7.5MSE +/- 39216.77, N = 3SE +/- 49410.05, N = 36600797.836814926.201. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

IndigoBench

This is a test of Indigo Renderer's IndigoBench benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgM samples/s, More Is BetterIndigoBench 4.4Acceleration: CPU - Scene: BedroomLinux 6.12 LTSLinux 6.6 LTS48121620SE +/- 0.03, N = 3SE +/- 0.04, N = 316.1716.15

OpenBenchmarking.orgM samples/s, More Is BetterIndigoBench 4.4Acceleration: CPU - Scene: SupercarLinux 6.12 LTSLinux 6.6 LTS816243240SE +/- 0.18, N = 3SE +/- 0.10, N = 335.4635.79

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.orgOp/s, More Is BetterRocksDB 9.0Test: Update RandomLinux 6.12 LTSLinux 6.6 LTS110K220K330K440K550KSE +/- 306.96, N = 3SE +/- 646.08, N = 35268445314631. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: Read Random Write RandomLinux 6.12 LTSLinux 6.6 LTS800K1600K2400K3200K4000KSE +/- 17391.58, N = 3SE +/- 23703.72, N = 3358795236371881. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti

OpenBenchmarking.orgOp/s, More Is BetterRocksDB 9.0Test: Random ReadLinux 6.12 LTSLinux 6.6 LTS90M180M270M360M450MSE +/- 1998121.04, N = 3SE +/- 414934.54, N = 34396144874402282201. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti

OpenVINO GenAI

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: Gemma-7b-int4-ov - Device: CPU - Time Per Output TokenLinux 6.12 LTSLinux 6.6 LTS1122334455SE +/- 0.18, N = 3SE +/- 0.14, N = 348.7048.42

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: Gemma-7b-int4-ov - Device: CPU - Time To First TokenLinux 6.12 LTSLinux 6.6 LTS1326395265SE +/- 0.28, N = 3SE +/- 0.29, N = 357.1557.03

OpenBenchmarking.orgtokens/s, More Is BetterOpenVINO GenAI 2024.5Model: Gemma-7b-int4-ov - Device: CPULinux 6.12 LTSLinux 6.6 LTS510152025SE +/- 0.07, N = 3SE +/- 0.06, N = 320.5320.66

Blender

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.3Blend File: Pabellon Barcelona - Compute: CPU-OnlyLinux 6.12 LTSLinux 6.6 LTS1224364860SE +/- 0.17, N = 3SE +/- 0.06, N = 354.1554.24

SVT-AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.3Encoder Mode: Preset 5 - Input: Bosphorus 4KLinux 6.12 LTSLinux 6.6 LTS1122334455SE +/- 0.07, N = 3SE +/- 0.13, N = 348.1848.381. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

LiteRT

OpenBenchmarking.orgMicroseconds, Fewer Is BetterLiteRT 2024-10-15Model: Inception V4Linux 6.12 LTSLinux 6.6 LTS6K12K18K24K30KSE +/- 20.95, N = 3SE +/- 67.75, N = 326660.626555.2

SQLite Speedtest

This is a benchmark of SQLite's speedtest1 benchmark program with an increased problem size of 1,000. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterSQLite Speedtest 3.30Timed Time - Size 1,000Linux 6.12 LTSLinux 6.6 LTS1224364860SE +/- 0.04, N = 3SE +/- 0.15, N = 349.7651.561. (CC) gcc options: -O2 -lz

LiteRT

OpenBenchmarking.orgMicroseconds, Fewer Is BetterLiteRT 2024-10-15Model: Mobilenet FloatLinux 6.12 LTSLinux 6.6 LTS5001000150020002500SE +/- 12.65, N = 3SE +/- 8.56, N = 32128.892129.64

OpenBenchmarking.orgMicroseconds, Fewer Is BetterLiteRT 2024-10-15Model: SqueezeNetLinux 6.12 LTSLinux 6.6 LTS7001400210028003500SE +/- 16.68, N = 3SE +/- 27.31, N = 33287.593230.93

OpenVINO GenAI

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: Falcon-7b-instruct-int4-ov - Device: CPU - Time Per Output TokenLinux 6.12 LTSLinux 6.6 LTS918273645SE +/- 0.15, N = 3SE +/- 0.08, N = 339.8740.00

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: Falcon-7b-instruct-int4-ov - Device: CPU - Time To First TokenLinux 6.12 LTSLinux 6.6 LTS1122334455SE +/- 0.32, N = 3SE +/- 0.58, N = 349.7349.80

OpenBenchmarking.orgtokens/s, More Is BetterOpenVINO GenAI 2024.5Model: Falcon-7b-instruct-int4-ov - Device: CPULinux 6.12 LTSLinux 6.6 LTS612182430SE +/- 0.10, N = 3SE +/- 0.05, N = 325.0825.00

Blender

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.3Blend File: Classroom - Compute: CPU-OnlyLinux 6.12 LTSLinux 6.6 LTS1122334455SE +/- 0.03, N = 3SE +/- 0.06, N = 346.0646.40

Stress-NG

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.17.08Test: Socket ActivityLinux 6.12 LTSLinux 6.6 LTS7K14K21K28K35KSE +/- 83.00, N = 3SE +/- 284.16, N = 630773.3229743.651. (CXX) g++ options: -O2 -std=gnu99 -lc -lm

Timed Linux Kernel Compilation

This test times how long it takes to build the Linux kernel in a default configuration (defconfig) for the architecture being tested or alternatively an allmodconfig for building all possible kernel modules for the build. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 6.8Build: defconfigLinux 6.12 LTSLinux 6.6 LTS714212835SE +/- 0.37, N = 3SE +/- 0.29, N = 627.9928.66

Llama.cpp

OpenBenchmarking.orgTokens Per Second, More Is BetterLlama.cpp b4154Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Text Generation 128Linux 6.12 LTSLinux 6.6 LTS48121620SE +/- 0.09, N = 3SE +/- 0.03, N = 315.5115.811. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -fopenmp -march=native -mtune=native -lopenblas

Cryptsetup

This is a test profile for running the cryptsetup benchmark to report on the system's cryptography performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterCryptsetupTwofish-XTS 512b DecryptionLinux 6.12 LTSLinux 6.6 LTS130260390520650SE +/- 4.61, N = 3SE +/- 0.30, N = 2595.9610.2

OpenBenchmarking.orgMiB/s, More Is BetterCryptsetupTwofish-XTS 512b EncryptionLinux 6.12 LTSLinux 6.6 LTS130260390520650SE +/- 2.60, N = 2SE +/- 0.97, N = 3576.8597.7

OpenBenchmarking.orgMiB/s, More Is BetterCryptsetupSerpent-XTS 512b DecryptionLinux 6.12 LTSLinux 6.6 LTS2004006008001000SE +/- 5.34, N = 3SE +/- 1.22, N = 3936.5950.1

OpenBenchmarking.orgMiB/s, More Is BetterCryptsetupSerpent-XTS 512b EncryptionLinux 6.12 LTSLinux 6.6 LTS2004006008001000SE +/- 5.35, N = 3SE +/- 1.11, N = 3946.5958.4

OpenBenchmarking.orgMiB/s, More Is BetterCryptsetupAES-XTS 512b DecryptionLinux 6.12 LTSLinux 6.6 LTS11002200330044005500SE +/- 19.22, N = 3SE +/- 12.68, N = 35324.13658.9

OpenBenchmarking.orgMiB/s, More Is BetterCryptsetupAES-XTS 512b EncryptionLinux 6.12 LTSLinux 6.6 LTS11002200330044005500SE +/- 18.18, N = 3SE +/- 13.94, N = 35320.73662.2

OpenBenchmarking.orgMiB/s, More Is BetterCryptsetupTwofish-XTS 256b DecryptionLinux 6.12 LTSLinux 6.6 LTS130260390520650SE +/- 4.98, N = 3SE +/- 3.03, N = 3596.1608.4

OpenBenchmarking.orgMiB/s, More Is BetterCryptsetupTwofish-XTS 256b EncryptionLinux 6.12 LTSLinux 6.6 LTS130260390520650SE +/- 6.39, N = 3SE +/- 3.72, N = 3578.4592.3

OpenBenchmarking.orgMiB/s, More Is BetterCryptsetupSerpent-XTS 256b DecryptionLinux 6.12 LTSLinux 6.6 LTS2004006008001000SE +/- 2.40, N = 3SE +/- 3.12, N = 3939.9947.1

OpenBenchmarking.orgMiB/s, More Is BetterCryptsetupSerpent-XTS 256b EncryptionLinux 6.12 LTSLinux 6.6 LTS2004006008001000SE +/- 5.03, N = 3SE +/- 9.38, N = 3946.2946.0

OpenBenchmarking.orgMiB/s, More Is BetterCryptsetupAES-XTS 256b DecryptionLinux 6.12 LTSLinux 6.6 LTS12002400360048006000SE +/- 4.24, N = 3SE +/- 16.03, N = 35735.23974.3

OpenBenchmarking.orgMiB/s, More Is BetterCryptsetupAES-XTS 256b EncryptionLinux 6.12 LTSLinux 6.6 LTS12002400360048006000SE +/- 14.10, N = 3SE +/- 17.91, N = 35714.93972.8

OpenBenchmarking.orgIterations Per Second, More Is BetterCryptsetupPBKDF2-whirlpoolLinux 6.12 LTSLinux 6.6 LTS200K400K600K800K1000KSE +/- 5339.48, N = 3SE +/- 4406.63, N = 3935732942440

OpenBenchmarking.orgIterations Per Second, More Is BetterCryptsetupPBKDF2-sha512Linux 6.12 LTSLinux 6.6 LTS600K1200K1800K2400K3000KSE +/- 13075.33, N = 3SE +/- 12354.07, N = 326117762627029

Llama.cpp

OpenBenchmarking.orgTokens Per Second, More Is BetterLlama.cpp b4154Backend: CPU BLAS - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Text Generation 128Linux 6.12 LTSLinux 6.6 LTS48121620SE +/- 0.06, N = 3SE +/- 0.01, N = 316.5516.691. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -fopenmp -march=native -mtune=native -lopenblas

NAMD

OpenBenchmarking.orgns/day, More Is BetterNAMD 3.0Input: STMV with 1,066,628 AtomsLinux 6.12 LTSLinux 6.6 LTS0.36980.73961.10941.47921.849SE +/- 0.00261, N = 3SE +/- 0.01023, N = 31.633241.64374

OpenBenchmarking.orgns/day, More Is BetterNAMD 3.0Input: ATPase with 327,506 AtomsLinux 6.12 LTSLinux 6.6 LTS246810SE +/- 0.01151, N = 3SE +/- 0.04086, N = 146.427746.27207

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.orgNs Per Day, More Is BetterGROMACS 2024Implementation: MPI CPU - Input: water_GMX50_bareLinux 6.12 LTSLinux 6.6 LTS246810SE +/- 0.007, N = 3SE +/- 0.013, N = 37.5587.5441. (CXX) g++ options: -O3 -lm

Llama.cpp

OpenBenchmarking.orgTokens Per Second, More Is BetterLlama.cpp b4154Backend: CPU BLAS - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Prompt Processing 1024Linux 6.12 LTSLinux 6.6 LTS50100150200250SE +/- 2.50, N = 3SE +/- 1.22, N = 3232.35186.831. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -fopenmp -march=native -mtune=native -lopenblas

Stress-NG

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.17.08Test: Mixed SchedulerLinux 6.12 LTSLinux 6.6 LTS9K18K27K36K45KSE +/- 90.61, N = 3SE +/- 38.24, N = 339770.2238895.001. (CXX) g++ options: -O2 -std=gnu99 -lc -lm

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: EclipseLinux 6.12 LTSLinux 6.6 LTS15003000450060007500SE +/- 36.56, N = 3SE +/- 30.14, N = 370637169

ParaView

OpenBenchmarking.orgMiVoxels / Sec, More Is BetterParaView 5.13Test: Wavelet Volume - Frames: 600 - Resolution: 1920 x 1200Linux 6.12 LTSLinux 6.6 LTS15003000450060007500SE +/- 199.16, N = 12SE +/- 203.86, N = 147085.926582.93

OpenBenchmarking.orgFrames / Sec, More Is BetterParaView 5.13Test: Wavelet Volume - Frames: 600 - Resolution: 1920 x 1200Linux 6.12 LTSLinux 6.6 LTS100200300400500SE +/- 12.45, N = 12SE +/- 12.74, N = 14442.87411.43

OpenVINO GenAI

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: TinyLlama-1.1B-Chat-v1.0 - Device: CPU - Time Per Output TokenLinux 6.12 LTSLinux 6.6 LTS510152025SE +/- 0.01, N = 3SE +/- 0.06, N = 322.1522.22

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: TinyLlama-1.1B-Chat-v1.0 - Device: CPU - Time To First TokenLinux 6.12 LTSLinux 6.6 LTS612182430SE +/- 0.03, N = 3SE +/- 0.09, N = 324.5524.71

OpenBenchmarking.orgtokens/s, More Is BetterOpenVINO GenAI 2024.5Model: TinyLlama-1.1B-Chat-v1.0 - Device: CPULinux 6.12 LTSLinux 6.6 LTS1020304050SE +/- 0.02, N = 3SE +/- 0.12, N = 345.1545.00

Unvanquished

OpenBenchmarking.orgFrames Per Second, More Is BetterUnvanquished 0.55Resolution: 1920 x 1200 - Effects Quality: UltraLinux 6.12 LTSLinux 6.6 LTS150300450600750SE +/- 6.32, N = 6SE +/- 2.62, N = 3672.7687.2

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: H2 Database EngineLinux 6.12 LTSLinux 6.6 LTS400800120016002000SE +/- 12.68, N = 3SE +/- 12.33, N = 320341994

Blender

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.3Blend File: Junkshop - Compute: CPU-OnlyLinux 6.12 LTSLinux 6.6 LTS612182430SE +/- 0.03, N = 3SE +/- 0.05, N = 324.1424.46

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.3Blend File: Fishy Cat - Compute: CPU-OnlyLinux 6.12 LTSLinux 6.6 LTS612182430SE +/- 0.14, N = 3SE +/- 0.08, N = 323.5823.63

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: jMonkeyEngineLinux 6.12 LTSLinux 6.6 LTS15003000450060007500SE +/- 2.19, N = 3SE +/- 0.88, N = 368806863

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: JythonLinux 6.12 LTSLinux 6.6 LTS9001800270036004500SE +/- 47.04, N = 5SE +/- 26.50, N = 342674295

Hackbench

This is a benchmark of Hackbench, a test of the Linux kernel scheduler. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterHackbenchCount: 32 - Type: ProcessLinux 6.12 LTSLinux 6.6 LTS510152025SE +/- 0.05, N = 3SE +/- 0.13, N = 319.8220.341. (CC) gcc options: -lpthread

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Spring BootLinux 6.12 LTSLinux 6.6 LTS400800120016002000SE +/- 24.47, N = 4SE +/- 19.17, N = 720012085

ParaView

OpenBenchmarking.orgMiPolys / Sec, More Is BetterParaView 5.13Test: Many Spheres - Frames: 600 - Resolution: 1920 x 1200Linux 6.12 LTSLinux 6.6 LTS3K6K9K12K15KSE +/- 28.78, N = 3SE +/- 106.65, N = 312130.2311842.94

OpenBenchmarking.orgFrames / Sec, More Is BetterParaView 5.13Test: Many Spheres - Frames: 600 - Resolution: 1920 x 1200Linux 6.12 LTSLinux 6.6 LTS306090120150SE +/- 0.29, N = 3SE +/- 1.06, N = 3120.99118.13

SVT-AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.3Encoder Mode: Preset 8 - Input: Bosphorus 4KLinux 6.12 LTSLinux 6.6 LTS306090120150SE +/- 0.78, N = 3SE +/- 0.17, N = 3140.94141.971. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

OpenVINO GenAI

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: Phi-3-mini-128k-instruct-int4-ov - Device: CPU - Time Per Output TokenLinux 6.12 LTSLinux 6.6 LTS714212835SE +/- 0.17, N = 3SE +/- 0.12, N = 328.0928.03

OpenBenchmarking.orgms, Fewer Is BetterOpenVINO GenAI 2024.5Model: Phi-3-mini-128k-instruct-int4-ov - Device: CPU - Time To First TokenLinux 6.12 LTSLinux 6.6 LTS816243240SE +/- 0.06, N = 3SE +/- 0.14, N = 335.5735.37

OpenBenchmarking.orgtokens/s, More Is BetterOpenVINO GenAI 2024.5Model: Phi-3-mini-128k-instruct-int4-ov - Device: CPULinux 6.12 LTSLinux 6.6 LTS816243240SE +/- 0.21, N = 3SE +/- 0.16, N = 335.6035.68

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Apache KafkaLinux 6.12 LTSLinux 6.6 LTS12002400360048006000SE +/- 4.04, N = 3SE +/- 0.58, N = 354525451

GPUScore: Breaking Limit

OpenBenchmarking.orgFPS, More Is BetterGPUScore: Breaking Limit 1.0Resolution: 1920 x 1200 - Ray-Tracing: OffLinux 6.12 LTSLinux 6.6 LTS306090120150SE +/- 0.05, N = 3SE +/- 0.08, N = 3131.34132.37MIN: 94.33 / MAX: 317.71MIN: 95.51 / MAX: 317.39

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Apache Lucene Search IndexLinux 6.12 LTSLinux 6.6 LTS6001200180024003000SE +/- 37.99, N = 3SE +/- 10.17, N = 326702647

SVT-AV1

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 2.3Encoder Mode: Preset 13 - Input: Bosphorus 4KLinux 6.12 LTSLinux 6.6 LTS50100150200250SE +/- 1.14, N = 3SE +/- 2.36, N = 6223.54232.451. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq

Blender

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 4.3Blend File: BMW27 - Compute: CPU-OnlyLinux 6.12 LTSLinux 6.6 LTS48121620SE +/- 0.03, N = 3SE +/- 0.04, N = 317.6517.57

Unvanquished

OpenBenchmarking.orgFrames Per Second, More Is BetterUnvanquished 0.55Resolution: 1920 x 1200 - Effects Quality: HighLinux 6.12 LTSLinux 6.6 LTS150300450600750SE +/- 3.84, N = 3SE +/- 2.80, N = 3678.6691.0

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: BioJava Biological Data FrameworkLinux 6.12 LTSLinux 6.6 LTS10002000300040005000SE +/- 15.03, N = 3SE +/- 11.98, N = 347684696

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Avrora AVR Simulation FrameworkLinux 6.12 LTSLinux 6.6 LTS11002200330044005500SE +/- 8.19, N = 3SE +/- 29.92, N = 349234844

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: PMD Source Code AnalyzerLinux 6.12 LTSLinux 6.6 LTS30060090012001500SE +/- 23.46, N = 12SE +/- 12.80, N = 411731155

Unvanquished

OpenBenchmarking.orgFrames Per Second, More Is BetterUnvanquished 0.55Resolution: 1920 x 1200 - Effects Quality: MediumLinux 6.12 LTSLinux 6.6 LTS150300450600750SE +/- 1.99, N = 3SE +/- 5.36, N = 3689.2707.9

x265

OpenBenchmarking.orgFrames Per Second, More Is Betterx265 4.1Video Input: Bosphorus 4KLinux 6.12 LTSLinux 6.6 LTS918273645SE +/- 0.04, N = 3SE +/- 0.12, N = 340.9341.001. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Batik SVG ToolkitLinux 6.12 LTSLinux 6.6 LTS2004006008001000SE +/- 9.69, N = 15SE +/- 12.22, N = 310121024

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Zxing 1D/2D Barcode Image ProcessingLinux 6.12 LTSLinux 6.6 LTS100200300400500SE +/- 3.12, N = 15SE +/- 5.05, N = 15405485

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: GraphChiLinux 6.12 LTSLinux 6.6 LTS5001000150020002500SE +/- 19.86, N = 3SE +/- 33.01, N = 324632408

Hackbench

This is a benchmark of Hackbench, a test of the Linux kernel scheduler. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterHackbenchCount: 16 - Type: ThreadLinux 6.12 LTSLinux 6.6 LTS3691215SE +/- 0.07, N = 3SE +/- 0.02, N = 310.7011.401. (CC) gcc options: -lpthread

OpenBenchmarking.orgSeconds, Fewer Is BetterHackbenchCount: 16 - Type: ProcessLinux 6.12 LTSLinux 6.6 LTS3691215SE +/- 0.10, N = 3SE +/- 0.04, N = 310.9811.141. (CC) gcc options: -lpthread

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Apache Lucene Search EngineLinux 6.12 LTSLinux 6.6 LTS6001200180024003000SE +/- 8.62, N = 3SE +/- 37.69, N = 326322817

Llama.cpp

OpenBenchmarking.orgTokens Per Second, More Is BetterLlama.cpp b4154Backend: CPU BLAS - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Text Generation 128Linux 6.12 LTSLinux 6.6 LTS20406080100SE +/- 0.17, N = 3SE +/- 0.28, N = 376.8177.761. (CXX) g++ options: -std=c++11 -fPIC -O3 -pthread -fopenmp -march=native -mtune=native -lopenblas

ParaView

OpenBenchmarking.orgMiPolys / Sec, More Is BetterParaView 5.13Test: Wavelet Contour - Frames: 600 - Resolution: 1920 x 1200Linux 6.12 LTSLinux 6.6 LTS7001400210028003500SE +/- 2.54, N = 3SE +/- 12.53, N = 33049.803019.32

OpenBenchmarking.orgFrames / Sec, More Is BetterParaView 5.13Test: Wavelet Contour - Frames: 600 - Resolution: 1920 x 1200Linux 6.12 LTSLinux 6.6 LTS60120180240300SE +/- 0.24, N = 3SE +/- 1.20, N = 3292.65289.73

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Apache TomcatLinux 6.12 LTSLinux 6.6 LTS30060090012001500SE +/- 7.54, N = 3SE +/- 8.17, N = 312451157

x265

OpenBenchmarking.orgFrames Per Second, More Is Betterx265 4.1Video Input: Bosphorus 1080pLinux 6.12 LTSLinux 6.6 LTS306090120150SE +/- 0.34, N = 3SE +/- 0.41, N = 3116.41116.921. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance of various popular real-world Java workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: Apache Xalan XSLTLinux 6.12 LTSLinux 6.6 LTS160320480640800SE +/- 1.86, N = 3SE +/- 5.90, N = 3744726

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 23.11Java Test: FOP Print FormatterLinux 6.12 LTSLinux 6.6 LTS100200300400500SE +/- 4.18, N = 3SE +/- 2.85, N = 3439430

Java Test: Apache Cassandra

Linux 6.6 LTS: The test quit with a non-zero exit status.

Linux 6.12 LTS: The test quit with a non-zero exit status.

Hackbench

This is a benchmark of Hackbench, a test of the Linux kernel scheduler. Learn more via the OpenBenchmarking.org test page.

Count: 32 - Type: Thread

Linux 6.6 LTS: The test quit with a non-zero exit status. E: Creating fdpair (error: Too many open files)

Linux 6.12 LTS: The test quit with a non-zero exit status. E: Creating fdpair (error: Too many open files)

115 Results Shown

srsRAN Project
WireGuard + Linux Networking Stack Stress Test
Llama.cpp
Timed Linux Kernel Compilation
DaCapo Benchmark
RocksDB
Unigine Superposition:
  1920 x 1200 - Fullscreen - High - OpenGL
  1920 x 1200 - Fullscreen - Medium - OpenGL
  1920 x 1200 - Fullscreen - Low - OpenGL
Llama.cpp
SVT-AV1
GravityMark:
  1920 x 1200 - OpenGL
  1920 x 1200 - Vulkan
Blender
Stress-NG
PostgreSQL:
  100 - 800 - Read Only - Average Latency
  100 - 800 - Read Only
  100 - 1000 - Read Write - Average Latency
  100 - 1000 - Read Write
  100 - 1000 - Read Only - Average Latency
  100 - 1000 - Read Only
  100 - 800 - Read Write - Average Latency
  100 - 800 - Read Write
Apache Cassandra
DaCapo Benchmark
Stress-NG
Chaos Group V-RAY
Memcached:
  1:100
  1:5
  1:10
IndigoBench:
  CPU - Bedroom
  CPU - Supercar
RocksDB:
  Update Rand
  Read Rand Write Rand
  Rand Read
OpenVINO GenAI:
  Gemma-7b-int4-ov - CPU - Time Per Output Token
  Gemma-7b-int4-ov - CPU - Time To First Token
  Gemma-7b-int4-ov - CPU
Blender
SVT-AV1
LiteRT
SQLite Speedtest
LiteRT:
  Mobilenet Float
  SqueezeNet
OpenVINO GenAI:
  Falcon-7b-instruct-int4-ov - CPU - Time Per Output Token
  Falcon-7b-instruct-int4-ov - CPU - Time To First Token
  Falcon-7b-instruct-int4-ov - CPU
Blender
Stress-NG
Timed Linux Kernel Compilation
Llama.cpp
Cryptsetup:
  Twofish-XTS 512b Decryption
  Twofish-XTS 512b Encryption
  Serpent-XTS 512b Decryption
  Serpent-XTS 512b Encryption
  AES-XTS 512b Decryption
  AES-XTS 512b Encryption
  Twofish-XTS 256b Decryption
  Twofish-XTS 256b Encryption
  Serpent-XTS 256b Decryption
  Serpent-XTS 256b Encryption
  AES-XTS 256b Decryption
  AES-XTS 256b Encryption
  PBKDF2-whirlpool
  PBKDF2-sha512
Llama.cpp
NAMD:
  STMV with 1,066,628 Atoms
  ATPase with 327,506 Atoms
GROMACS
Llama.cpp
Stress-NG
DaCapo Benchmark
ParaView:
  Wavelet Volume - 600 - 1920 x 1200:
    MiVoxels / Sec
    Frames / Sec
OpenVINO GenAI:
  TinyLlama-1.1B-Chat-v1.0 - CPU - Time Per Output Token
  TinyLlama-1.1B-Chat-v1.0 - CPU - Time To First Token
  TinyLlama-1.1B-Chat-v1.0 - CPU
Unvanquished
DaCapo Benchmark
Blender:
  Junkshop - CPU-Only
  Fishy Cat - CPU-Only
DaCapo Benchmark:
  jMonkeyEngine
  Jython
Hackbench
DaCapo Benchmark
ParaView:
  Many Spheres - 600 - 1920 x 1200:
    MiPolys / Sec
    Frames / Sec
SVT-AV1
OpenVINO GenAI:
  Phi-3-mini-128k-instruct-int4-ov - CPU - Time Per Output Token
  Phi-3-mini-128k-instruct-int4-ov - CPU - Time To First Token
  Phi-3-mini-128k-instruct-int4-ov - CPU
DaCapo Benchmark
GPUScore: Breaking Limit
DaCapo Benchmark
SVT-AV1
Blender
Unvanquished
DaCapo Benchmark:
  BioJava Biological Data Framework
  Avrora AVR Simulation Framework
  PMD Source Code Analyzer
Unvanquished
x265
DaCapo Benchmark:
  Batik SVG Toolkit
  Zxing 1D/2D Barcode Image Processing
  GraphChi
Hackbench:
  16 - Thread
  16 - Process
DaCapo Benchmark
Llama.cpp
ParaView:
  Wavelet Contour - 600 - 1920 x 1200:
    MiPolys / Sec
    Frames / Sec
DaCapo Benchmark
x265
DaCapo Benchmark:
  Apache Xalan XSLT
  FOP Print Formatter