Extra EPYC Rome Benchmarks

AMD Ryzen 9 3900X 12-Core testing with a Gigabyte X570 AORUS MASTER (F5k BIOS) and Gigabyte NVIDIA GeForce RTX 2080 Ti Rev. A 11GB on Fedora 30 via the Phoronix Test Suite.

Compare your own system(s) to this result file with the Phoronix Test Suite by running the command: phoronix-test-suite benchmark 1908150-SP-1908130AS62
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

Limit displaying results to tests within:

Audio Encoding 3 Tests
AV1 2 Tests
Bioinformatics 2 Tests
C++ Boost Tests 2 Tests
Chess Test Suite 4 Tests
Timed Code Compilation 5 Tests
C/C++ Compiler Tests 25 Tests
Compression Tests 4 Tests
CPU Massive 64 Tests
Creator Workloads 22 Tests
Cryptography 5 Tests
Database Test Suite 3 Tests
Encoding 10 Tests
Fortran Tests 3 Tests
HPC - High Performance Computing 12 Tests
Imaging 4 Tests
Java 5 Tests
Common Kernel Benchmarks 7 Tests
Memory Test Suite 4 Tests
Molecular Dynamics 2 Tests
MPI Benchmarks 3 Tests
Multi-Core 34 Tests
NVIDIA GPU Compute 4 Tests
OpenCL 3 Tests
OpenMPI Tests 5 Tests
Programmer / Developer System Benchmarks 7 Tests
Python 2 Tests
Raytracing 2 Tests
Renderers 7 Tests
Scientific Computing 7 Tests
Server 9 Tests
Server CPU Tests 42 Tests
Single-Threaded 16 Tests
Video Encoding 7 Tests
Common Workstation Benchmarks 5 Tests

Statistics

Show Overall Harmonic Mean(s)
Show Overall Geometric Mean
Show Geometric Means Per-Suite/Category
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
Disable Color Branding
Prefer Vertical Bar Graphs

Additional Graphs

Show Perf Per Core/Thread Calculation Graphs Where Applicable
Show Perf Per Clock Calculation Graphs Where Applicable

Multi-Way Comparison

Condense Multi-Option Tests Into Single Result Graphs

Table

Show Detailed System Result Table

Run Management

Highlight
Result
Hide
Result
Result
Identifier
Performance Per
Dollar
Date
Run
  Test
  Duration
2 x EPYC 7502
July 25 2019
  2 Days, 7 Hours, 4 Minutes
2 x EPYC 7742
July 28 2019
  4 Hours, 46 Minutes
2 x Xeon Platinum 8280
August 11 2019
  1 Day, 21 Hours, 18 Minutes
Ryzen 3900X
August 14 2019
  1 Day, 3 Hours, 30 Minutes
Invert Hiding All Results Option
  1 Day, 9 Hours, 10 Minutes

Only show results where is faster than
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):

2 x EPYC 7502

Processor: 2 x AMD EPYC 7502 32-Core @ 2.50GHz (64 Cores / 128 Threads), Motherboard: AMD DAYTONA_X (RDY1001C BIOS), Chipset: AMD Device 1480, Memory: 516096MB, Disk: 280GB INTEL SSDPE21D280GA + 280GB INTEL SSDPED1D280GA + 5 x 3841GB Micron_9300_MTFDHAL3T8TDP + 256GB Micron_1100_MTFD, Graphics: llvmpipe 504GB, Monitor: VE228, Network: 2 x Mellanox MT27710

OS: Ubuntu 19.04, Kernel: 5.2.0-050200rc7-generic (x86_64) 20190630, Desktop: GNOME Shell 3.32.1, Display Server: X Server 1.20.4, Display Driver: modesetting 1.20.4, OpenGL: 3.3 Mesa 19.0.2 (LLVM 8.0 128 bits), Compiler: GCC 8.3.0, File-System: ext4, Screen Resolution: 1920x1080

Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++ --enable-libmpx --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none --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-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib --with-tune=generic --without-cuda-driver -v
Disk Notes: NONE / errors=remount-ro,relatime,rw
Processor Notes: Scaling Governor: acpi-cpufreq ondemand
Java Notes: OpenJDK Runtime Environment (build 11.0.3+7-Ubuntu-1ubuntu219.04.1)
Python Notes: Python 2.7.16 + Python 3.7.3
Security Notes: l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional IBRS_FW STIBP: conditional RSB filling

2 x EPYC 7742

Processor: 2 x AMD EPYC 7742 64-Core @ 2.25GHz (128 Cores / 256 Threads), Motherboard: AMD DAYTONA_X (RDY1001C BIOS), Chipset: AMD Device 1480, Memory: 516096MB, Disk: 280GB INTEL SSDPE21D280GA + 280GB INTEL SSDPED1D280GA + 5 x 3841GB Micron_9300_MTFDHAL3T8TDP + 256GB Micron_1100_MTFD, Graphics: ASPEED, Monitor: VE228, Network: 2 x Mellanox MT27710

OS: Ubuntu 19.04, Kernel: 5.3.0-rc1-power (x86_64) 20190728, Desktop: GNOME Shell 3.32.1, Display Server: X Server 1.20.4, Display Driver: modesetting 1.20.4, Compiler: GCC 8.3.0, File-System: ext4, Screen Resolution: 1920x1080

2 x Xeon Platinum 8280

Processor: 2 x Intel Xeon Platinum 8280 @ 4.00GHz (56 Cores / 112 Threads), Motherboard: GIGABYTE MD61-SC2-00 v01000100 (T15 BIOS), Chipset: Intel Sky Lake-E DMI3 Registers, Memory: 386048MB, Disk: 280GB INTEL SSDPED1D280GA + 32GB Flash Disk, Graphics: llvmpipe 377GB, Monitor: VE228, Network: 2 x Intel X722 for 1GbE + 2 x QLogic FastLinQ QL41000 10/25/40/50GbE

OS: Ubuntu 19.04, Kernel: 5.2.0-050200rc7-generic (x86_64) 20190630, Desktop: GNOME Shell 3.32.2, Display Server: X Server 1.20.4, Display Driver: modesetting 1.20.4, OpenGL: 3.3 Mesa 19.0.2 (LLVM 8.0 256 bits), Compiler: GCC 8.3.0, File-System: ext4, Screen Resolution: 1920x1080

Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++ --enable-libmpx --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none --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-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib --with-tune=generic --without-cuda-driver -v
Disk Notes: NONE / errors=remount-ro,relatime,rw
Processor Notes: Scaling Governor: intel_pstate powersave
Java Notes: OpenJDK Runtime Environment (build 11.0.4+11-post-Ubuntu-1ubuntu219.04)
Python Notes: Python 2.7.16 + Python 3.7.3
Security Notes: l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling

Ryzen 3900X

Processor: AMD Ryzen 9 3900X 12-Core @ 3.80GHz (12 Cores / 24 Threads), Motherboard: Gigabyte X570 AORUS MASTER (F5k BIOS), Chipset: AMD Starship/Matisse, Memory: 16384MB, Disk: 2 x 2000GB GIGABYTE GP-ASM2NE6200TTTD, Graphics: Gigabyte NVIDIA GeForce RTX 2080 Ti Rev. A 11GB, Audio: NVIDIA TU102 HD Audio, Monitor: BenQ EL2870U, Network: Intel I211 + Realtek Device 8125 + Intel Wi-Fi 6 AX200

OS: Fedora 30, Kernel: 5.2.7-200.fc30.x86_64 (x86_64), Desktop: GNOME Shell 3.32.2, Display Server: X Server, Compiler: GCC 9.1.1 20190503, File-System: ext4, Screen Resolution: 3840x2160

Compiler Notes: --build=x86_64-redhat-linux --disable-libunwind-exceptions --enable-__cxa_atexit --enable-bootstrap --enable-cet --enable-checking=release --enable-gnu-indirect-function --enable-gnu-unique-object --enable-initfini-array --enable-languages=c,c++,fortran,objc,obj-c++,ada,go,d,lto --enable-multilib --enable-offload-targets=nvptx-none --enable-plugin --enable-shared --enable-threads=posix --mandir=/usr/share/man --with-arch_32=i686 --with-gcc-major-version-only --with-isl --with-linker-hash-style=gnu --with-tune=generic --without-cuda-driver
Disk Notes: NONE / relatime,rw,seclabel,stripe=256 / raid0 nvme1n1p3[0] nvme0n1p2[1]
Processor Notes: Scaling Governor: acpi-cpufreq ondemand
Python Notes: Python 2.7.16 + Python 3.7.4
Security Notes: SELinux + usercopy/swapgs barriers and __user pointer sanitization + Full AMD retpoline IBPB: conditional STIBP: always-on RSB filling + SSB disabled via prctl and seccomp

Extra EPYC Rome BenchmarksProcessorMotherboardChipsetMemoryDiskGraphicsMonitorNetworkAudioOSKernelDesktopDisplay ServerDisplay DriverOpenGLCompilerFile-SystemScreen Resolution2 x EPYC 75022 x EPYC 77422 x Xeon Platinum 8280Ryzen 3900X2 x AMD EPYC 7502 32-Core @ 2.50GHz (64 Cores / 128 Threads)AMD DAYTONA_X (RDY1001C BIOS)AMD Device 1480516096MB280GB INTEL SSDPE21D280GA + 280GB INTEL SSDPED1D280GA + 5 x 3841GB Micron_9300_MTFDHAL3T8TDP + 256GB Micron_1100_MTFDllvmpipe 504GBVE2282 x Mellanox MT27710Ubuntu 19.045.2.0-050200rc7-generic (x86_64) 20190630GNOME Shell 3.32.1X Server 1.20.4modesetting 1.20.43.3 Mesa 19.0.2 (LLVM 8.0 128 bits)GCC 8.3.0ext41920x10802 x AMD EPYC 7742 64-Core @ 2.25GHz (128 Cores / 256 Threads)ASPEED5.3.0-rc1-power (x86_64) 201907282 x Intel Xeon Platinum 8280 @ 4.00GHz (56 Cores / 112 Threads)GIGABYTE MD61-SC2-00 v01000100 (T15 BIOS)Intel Sky Lake-E DMI3 Registers386048MB280GB INTEL SSDPED1D280GA + 32GB Flash Diskllvmpipe 377GB2 x Intel X722 for 1GbE + 2 x QLogic FastLinQ QL41000 10/25/40/50GbE5.2.0-050200rc7-generic (x86_64) 20190630GNOME Shell 3.32.23.3 Mesa 19.0.2 (LLVM 8.0 256 bits)AMD Ryzen 9 3900X 12-Core @ 3.80GHz (12 Cores / 24 Threads)Gigabyte X570 AORUS MASTER (F5k BIOS)AMD Starship/Matisse16384MB2 x 2000GB GIGABYTE GP-ASM2NE6200TTTDGigabyte NVIDIA GeForce RTX 2080 Ti Rev. A 11GBNVIDIA TU102 HD AudioBenQ EL2870UIntel I211 + Realtek Device 8125 + Intel Wi-Fi 6 AX200Fedora 305.2.7-200.fc30.x86_64 (x86_64)X ServerGCC 9.1.1 201905033840x2160OpenBenchmarking.orgCompiler Details- 2 x EPYC 7502: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++ --enable-libmpx --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none --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-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib --with-tune=generic --without-cuda-driver -v - 2 x EPYC 7742: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++ --enable-libmpx --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none --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-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib --with-tune=generic --without-cuda-driver -v - 2 x Xeon Platinum 8280: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++ --enable-libmpx --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none --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-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib --with-tune=generic --without-cuda-driver -v - Ryzen 3900X: --build=x86_64-redhat-linux --disable-libunwind-exceptions --enable-__cxa_atexit --enable-bootstrap --enable-cet --enable-checking=release --enable-gnu-indirect-function --enable-gnu-unique-object --enable-initfini-array --enable-languages=c,c++,fortran,objc,obj-c++,ada,go,d,lto --enable-multilib --enable-offload-targets=nvptx-none --enable-plugin --enable-shared --enable-threads=posix --mandir=/usr/share/man --with-arch_32=i686 --with-gcc-major-version-only --with-isl --with-linker-hash-style=gnu --with-tune=generic --without-cuda-driver Disk Details- 2 x EPYC 7502: NONE / errors=remount-ro,relatime,rw- 2 x EPYC 7742: NONE / errors=remount-ro,relatime,rw- 2 x Xeon Platinum 8280: NONE / errors=remount-ro,relatime,rw- Ryzen 3900X: NONE / relatime,rw,seclabel,stripe=256 / raid0 nvme1n1p3[0] nvme0n1p2[1]Processor Details- 2 x EPYC 7502: Scaling Governor: acpi-cpufreq ondemand- 2 x EPYC 7742: Scaling Governor: acpi-cpufreq ondemand- 2 x Xeon Platinum 8280: Scaling Governor: intel_pstate powersave- Ryzen 3900X: Scaling Governor: acpi-cpufreq ondemandJava Details- 2 x EPYC 7502: OpenJDK Runtime Environment (build 11.0.3+7-Ubuntu-1ubuntu219.04.1)- 2 x EPYC 7742: OpenJDK Runtime Environment (build 11.0.3+7-Ubuntu-1ubuntu219.04.1)- 2 x Xeon Platinum 8280: OpenJDK Runtime Environment (build 11.0.4+11-post-Ubuntu-1ubuntu219.04)Python Details- 2 x EPYC 7502: Python 2.7.16 + Python 3.7.3- 2 x EPYC 7742: Python 2.7.16 + Python 3.7.3- 2 x Xeon Platinum 8280: Python 2.7.16 + Python 3.7.3- Ryzen 3900X: Python 2.7.16 + Python 3.7.4Security Details- 2 x EPYC 7502: l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional IBRS_FW STIBP: conditional RSB filling - 2 x EPYC 7742: l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional IBRS_FW STIBP: conditional RSB filling - 2 x Xeon Platinum 8280: l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling- Ryzen 3900X: SELinux + usercopy/swapgs barriers and __user pointer sanitization + Full AMD retpoline IBPB: conditional STIBP: always-on RSB filling + SSB disabled via prctl and seccomp

2 x EPYC 75022 x EPYC 77422 x Xeon Platinum 8280Ryzen 3900XLogarithmic Result OverviewPhoronix Test SuiteNAMDCP2K Molecular DynamicsTimed MAFFT AlignmentRodiniaMBWBLAKE2StreamCLOMPParboilBork File EncrypterRAMspeed SMPBlogBenchRenaissanceTinymembenchGNU GMP GMPbenchFFTWJava SciMarkSockperf

Extra EPYC Rome Benchmarksmkl-dnn: IP Batch All - f32stress-ng: Socket Activitymkl-dnn: IP Batch 1D - f32mysqlslap: 16hpcc: Rand Ring Bandwidthhpcc: G-Rand Accesshpcc: G-Fftehpcc: G-Fftemkl-dnn: Deconvolution Batch deconv_all - u8s8u8s32john-the-ripper: Blowfishcp2k: Fayalite-FIST Datamkl-dnn: IP Batch 1D - u8s8f32s32mkl-dnn: Convolution Batch conv_alexnet - f32mkl-dnn: IP Batch 1D - u8s8u8s32mkl-dnn: Convolution Batch conv_all - f32mkl-dnn: Convolution Batch conv_googlenet_v3 - f32mkl-dnn: Convolution Batch conv_3d - u8s8u8s32mkl-dnn: Convolution Batch conv_3d - u8s8f32s32mkl-dnn: Deconvolution Batch deconv_3d - u8s8f32s32stress-ng: Matrix Mathstress-ng: Glibc Qsort Data Sortingstress-ng: CPU Stressstress-ng: Bsearchstress-ng: Semaphoresopenssl: RSA 4096-bit Performancestress-ng: Cryptomkl-dnn: Deconvolution Batch deconv_3d - u8s8u8s32mbw: Memory Copy - 8192 MiBmbw: Memory Copy, Fixed Block Size - 8192 MiBmkl-dnn: Deconvolution Batch deconv_3d - f32stress-ng: Hsearchstress-ng: Lsearchblender: Classroom - CPU-Onlymbw: Memory Copy - 4096 MiBmbw: Memory Copy - 1024 MiBstress-ng: Vector Mathc-ray: Total Time - 4K, 16 Rays Per Pixelm-queens: Time To Solvembw: Memory Copy - 512 MiBmbw: Memory Copy - 128 MiBstress-ng: Tsearchblender: Pabellon Barcelona - CPU-Onlycoremark: CoreMark Size 666 - Iterations Per Secondprimesieve: 1e12 Prime Number Generationasmfish: 1024 Hash Memory, 26 Depthstockfish: Total Timecompress-7zip: Compress Speed Testpts-self-test: Total Timeblender: Barbershop - CPU-Onlymysqlslap: 32povray: Trace Timeblender: BMW27 - CPU-Onlyrenaissance: In-Memory Database Shootoutpgbench: Buffer Test - Normal Load - Read Onlypts-self-test: List Suiteshpcc: EP-DGEMMmkl-dnn: Deconvolution Batch deconv_1d - f32hpcc: Rand Ring Latencyblender: Fishy Cat - CPU-Onlyhimeno: Poisson Pressure Solverappleseed: Disney Materialstress-ng: System V Message Passingparboil: OpenMP CUTCPpts-self-test: Info Commandbuild-llvm: Time To Compilemcperf: Replacey-cruncher: Calculating 500M Pi Digitspgbench: Buffer Test - Single Thread - Read Writehpcc: Max Ping Pong Bandwidthosbench: Create Processesblake2: compress-zstd: Compressing ubuntu-16.04.3-server-i386.img, Compression Level 19mcperf: Getosbench: Launch Programsmbw: Memory Copy, Fixed Block Size - 1024 MiBmcperf: Appendmbw: Memory Copy, Fixed Block Size - 512 MiBmbw: Memory Copy, Fixed Block Size - 4096 MiBmkl-dnn: Convolution Batch conv_googlenet_v3 - u8s8u8s32mkl-dnn: Convolution Batch conv_alexnet - u8s8u8s32mkl-dnn: Convolution Batch conv_googlenet_v3 - u8s8f32s32pts-self-test: Merge Resultsmkl-dnn: Convolution Batch conv_all - u8s8u8s32mkl-dnn: Convolution Batch conv_all - u8s8f32s32mkl-dnn: Deconvolution Batch deconv_1d - u8s8f32s32mkl-dnn: Deconvolution Batch deconv_1d - u8s8u8s32parboil: OpenMP MRI Griddingmkl-dnn: Convolution Batch conv_alexnet - u8s8f32s32mysqlslap: 4hpcc: G-HPLmkl-dnn: IP Batch All - u8s8f32s32mkl-dnn: IP Batch All - u8s8u8s32stream: Scalemysqlslap: 8pts-self-test: Render Testfftw: Float + SSE - 2D FFT Size 512mbw: Memory Copy, Fixed Block Size - 128 MiBfftw: Float + SSE - 2D FFT Size 1024ramspeed: Copy - Integerramspeed: Add - Floating Pointrodinia: OpenMP CFD Solverramspeed: Add - Integerfftw: Float + SSE - 2D FFT Size 256fftw: Float + SSE - 2D FFT Size 32appleseed: Material Testerapache-siege: 1java-scimark2: Monte Carloappleseed: Emilypgbench: Buffer Test - Single Thread - Read Onlymysqlslap: 128ramspeed: Scale - Floating Pointblogbench: Readbork: File Encryption Timefftw: Float + SSE - 2D FFT Size 128ramspeed: Triad - Floating Pointdav1d: Summer Nature 4Kdcraw: RAW To PPM Image Conversionblogbench: Writeramspeed: Average - Floating Pointdav1d: Summer Nature 1080pramspeed: Scale - Integerstress-ng: Forkingmysqlslap: 64dacapobench: Tradebeansramspeed: Average - Integerbotan: AES-256 - Encryptsvt-av1: 1080p 8-bit YUV To AV1 Video Encodebotan: AES-256 - Decryptttsiod-renderer: Phong Rendering With Soft-Shadow Mappingvpxenc: vpxenc VP9 1080p Video Encoderamspeed: Copy - Floating Pointfftw: Float + SSE - 1D FFT Size 512fftw: Float + SSE - 1D FFT Size 1024renaissance: Scala Dottyminion: Solitairephp: Zend micro_benchtinymembench: Standard Memsetfftw: Float + SSE - 2D FFT Size 64ramspeed: Triad - Integerjava-scimark2: Dense LU Matrix Factorizationhpcc: EP-STREAM Triadnpb: FT.Bfftw: Float + SSE - 1D FFT Size 128fftw: Float + SSE - 1D FFT Size 2048osbench: Memory Allocationsfftw: Stock - 2D FFT Size 512fftw: Stock - 2D FFT Size 4096phpbench: PHP Benchmark Suiteapache-siege: 250minion: Quasigroupnpb: FT.Afftw: Float + SSE - 1D FFT Size 4096tinymembench: Standard Memcpytjbench: Decompression Throughputsvt-vp9: 1080p 8-bit YUV To VP9 Video Encodenumpy: botan: CAST-256 - Decryptbotan: CAST-256 - Encryptbotan: KASUMI - Encryptfftw: Stock - 1D FFT Size 1024encode-mp3: WAV To MP3fftw: Stock - 2D FFT Size 256apache-siege: 100mencoder: AVI To LAVCbotan: KASUMI - Decryptphp: Zend benchfftw: Float + SSE - 1D FFT Size 32apache-siege: 200java-scimark2: Compositeencode-flac: WAV To FLACfftw: Stock - 2D FFT Size 64build-gcc: Time To Compileapache-siege: 50sockperf: Latency Ping Pongfftw: Stock - 1D FFT Size 128tscp: AI Chess Performancex265: H.265 1080p Video Encodingbotan: Twofish - Encryptfftw: Stock - 2D FFT Size 128dacapobench: Tradesoapfftw: Stock - 1D FFT Size 512fftw: Stock - 1D FFT Size 2048gmpbench: Total Timesvt-hevc: 1080p 8-bit YUV To HEVC Video Encodebotan: Twofish - Decryptapache-siege: 10botan: Blowfish - Decryptfftw: Stock - 2D FFT Size 32fftw: Stock - 1D FFT Size 4096fftw: Stock - 2D FFT Size 1024botan: Blowfish - Encryptfftw: Stock - 2D FFT Size 2048lzbench: XZ 0 - Compressionindigobench: Bedroomindigobench: Supercardacapobench: H2fftw: Float + SSE - 2D FFT Size 4096fftw: Stock - 1D FFT Size 256fftw: Float + SSE - 2D FFT Size 2048cython-bench: x264: H.264 Video Encodingjava-scimark2: Sparse Matrix Multiplyosbench: Create Filesstress-ng: Memory Copyingdarktable: Boat - CPU-onlylzbench: Libdeflate 1 - Decompressionredis: SADDcompress-xz: Compressing ubuntu-16.04.3-server-i386.img, Compression Level 9spec-jbb2015: SPECjbb2015-Composite critical-jOPSdacapobench: Jythonlzbench: Brotli 0 - Decompressionfftw: Stock - 1D FFT Size 64java-scimark2: Jacobi Successive Over-Relaxationspec-jbb2015: SPECjbb2015-Composite max-jOPSdarktable: Masskrug - CPU-onlylzbench: Libdeflate 1 - Compressionblender: Pabellon Barcelona - CPU-Onlyctx-clock: Context Switch Timeencode-ogg: WAV To Oggblender: Classroom - CPU-Onlybuild-php: Time To Compileblender: Fishy Cat - CPU-Onlylzbench: XZ 0 - Decompressiongimp: rotatenpb: EP.Cbuild-imagemagick: Time To Compilegimp: unsharp-maskblender: BMW27 - CPU-Onlygimp: auto-levelsgimp: resizetesseract-ocr: Time To OCR 7 Imageslzbench: Zstd 1 - Compressionv-ray: CPUnpb: LU.Cnpb: SP.Alzbench: Zstd 1 - Decompressionlzbench: Brotli 0 - Compressionblender: Barbershop - CPU-Onlyminion: Gracefulpts-self-test: Peak Memory Usagenpb: BT.Affte: N=256, 1D Complex FFT Routineosbench: Create Threadssunflow: Global Illumination + Image Synthesismcperf: Prependmcperf: Deletemcperf: Setmcperf: Addstress-ng: Glibc C String Functionsstress-ng: Context Switchingredis: SETredis: GETredis: LPUSHredis: LPOPdarktable: Server Room - CPU-onlydarktable: Server Rack - CPU-onlypgbench: Buffer Test - Normal Load - Read Writemysqlslap: 256build-linux-kernel: Time To Compilemkl-dnn: Deconvolution Batch deconv_all - f32mkl-dnn: Convolution Batch conv_3d - f32renaissance: Akka Unbalanced Cobwebbed Treerenaissance: Apache Spark PageRankrenaissance: Savina Reactors.IOrenaissance: Apache Spark Bayesrenaissance: Apache Spark ALSjava-scimark2: Fast Fourier Transformmafft: Multiple Sequence Alignmentfftw: Float + SSE - 1D FFT Size 256fftw: Float + SSE - 1D FFT Size 64fftw: Stock - 1D FFT Size 32namd: ATPase Simulation - 327,506 Atomsclomp: Static OMP Speeduphpcg: rodinia: OpenMP Streamclusterrodinia: OpenMP LavaMDparboil: OpenMP Stencilparboil: OpenMP LBMhpcc: G-Ptransnpb: LU.Asockperf: Latency Under Loadsockperf: Throughputstream: Addstream: Triadstream: Copy2 x EPYC 75022 x EPYC 77422 x Xeon Platinum 8280Ryzen 3900X81.93238119.4737131.383850.2023333.3048333.304831530710936358068.2880.9368.4762134.6624382447137737552593119073331731398972143881430615071567590423.8638839670880.60154721571113050510.9012.10156718986175210221012994.29136257996135223710255492253221175710.3245.20136838583920.7422.389708.111.7797666.52100459.80122160451.043.0896.43283238.8926891078163.317.4611.205365078.17909334850912690171752.47947.221764.584.842080020778795.68881.33120.92941.697634169.92267778.49772.47155497639769.673387989843325142739485478.874755233586345722844559.8613341732038834240026127205610.77345493224713.2451.4343807442755.693846543021835741942023458594.994640693145.29302753772243078732682.072.6615240358974011052702.802073660.37209604491596.596633511649491030037137.893332.60416189148169197.26721657712012077.84712933.1766152665520.2075.240.541184828569238910.097300746216924.856766103607843.4629967885601705268584519340.94299199082288824670960992315815284.787.7269111700471272610928.68148.94221115.2398233.30841180054626.97602185466441808215061066454.24190100.131506.81122.6449.1470.168716.752283.0321.3930.6047.3224.6211.6734.903546229733940.81396.26932383236.5352.78503834.437639.4533.790.742827940803335803315749089426207536014632461897043137585321703593.030.112193627019.7442293.6730304247262840435953877816243.42279501644787410.413490.370.4613.576.115.6425.001.3688231119.4622.4248436316410516611915078861115556905115436156021546490660.806.60915391889032182.831770293417890533389541105485309.10482533428435028135636196116406710.733500346521382264096037700774342272355263892243869727616031370114302353642105945337671752144303891317207674411985242374005.3068996948712470064589898268146146599328822617865721726661224686752644468203152919188360961383244123111541263753916223.41278601603288680.265890.230.3510.434.334.0235.9722.16454195194680200966185818100.803754011.5663470.837300.1754016.9060316.9060337128357816993.5548.573.6438221.533576353818903808767871795535116209893913753105032041474751351.1130130866594.294752477512095714.4016.3048194975109111718906494.82137371669125947505232626215223290511.5747.45173389128060.7125.173530.951.1335672.54279467.2465517141.292.40107.93711859.1226561727445.923.879.0111193054.51491670478491050535.7714.236.233.91172217560.230.23513.3118.9513278146.4503320.4820.421494901042259.802228949182235132285284007.313277924853271281852606.48984155190165442582219952397.10286132843218.2541.7753516290609.20290966899012861178031645382361.533828101094.82321105598364732693084.342.0610920336483186447933.002692556.20203436390887.576355464064673834608111.212378.80582226933188271.63526221613713789.63755529.9763863324217.6987.030.45143173343221829.567181892279153.987024125748932.7334774707404779373925098260.31351260122758898727463402825090366.149.8887731609972002119822.97120.04215613.77112634.051016212272722.5050770462951974041585939954.77213112.001366.12135.0248.5976.829518.252486.0519.6728.1851.1522.7912.4832.793756547632373.02379.98967391240.9453.28503834.3020.510.8670628113029620216151044836532655021317528602299349165646424752141.920.102370343821.4510553.3024373236912605832834916113925.16367041698471790.359230.430.9411.796.962.6022.4315.0181730471.2035.39448019156319154125168490791.93418364.099465.250310.032495.409715.4097120213920460321176.212393.04177.05185041021.921061181061235936477603206444083425724009353135205905344326038.1497237178320.2618783185993350742.2246.75184391852147537057944615.4740037020400495217758569770791431.55138.06573530230120.171.07655197.670.596171941278164.4446454262.071.20245.504639521.9211502502327.638.3018.967872437.8599634655498961017011526.1027767.6711596.5074.2829463529314835857.1735756.5034.1627648.1095312.794802276.912265.861517098635.2942671939341799234242695713.1427016442034819516117162703308558923339129895812.08485122745922.2631.1532787273147.4923852638658042733359105896656233945680658387489156.551.8013889491462959769442.07594290166060668.01901265566982542462698.29582179630233164164106.29972624.3886832445014.92101.820.40159552490629247.539601668288894.84899613769623968988928889985922391204002971146687137909299657320418898121767267417.069153216921825.328701171615244.6819.9550014.681.10525877772258184469881262789572185617927912538538173843427272081765620647.7931340101.701200827352122058382488612802.083997822302116911.444422.4523.7120.0816.90168.102.4977118.51373777162841630514931OpenBenchmarking.org

MKL-DNN

This is a test of the Intel MKL-DNN as the Intel Math Kernel Library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN 2019-04-16Harness: IP Batch All - Data Type: f322 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X2004006008001000SE +/- 0.98, N = 3SE +/- 1.72, N = 3SE +/- 0.86, N = 381.93100.80791.93-O2 - MIN: 786.541. (CXX) g++ options: -std=c++11 -march=native -mtune=native -fPIC -fopenmp -pie -lmklml_intel -ldl

Stress-NG

Stress-NG is a Linux stress tool developed by Colin King of Canonical. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.07.26Test: Socket Activity2 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X8K16K24K32K40KSE +/- 497.69, N = 3SE +/- 58.06, N = 3SE +/- 44.07, N = 3375402381141831. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc

MKL-DNN

This is a test of the Intel MKL-DNN as the Intel Math Kernel Library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN 2019-04-16Harness: IP Batch 1D - Data Type: f322 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X1428425670SE +/- 0.09, N = 3SE +/- 0.15, N = 3SE +/- 0.11, N = 39.4711.5664.09-O3 - MIN: 4.89-O3 - MIN: 6.18-O2 - MIN: 63.621. (CXX) g++ options: -std=c++11 -march=native -mtune=native -fPIC -fopenmp -pie -lmklml_intel -ldl

MariaDB

This is a MariaDB MySQL database server benchmark making use of mysqlslap. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 10.3.8Clients: 162 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X14002800420056007000SE +/- 73.23, N = 6SE +/- 43.08, N = 6SE +/- 3.79, N = 363473713946-llzma -lbz2 -lnuma -lpcre-llzma -lbz2 -lnuma -lpcre1. (CXX) g++ options: -pie -fPIC -fstack-protector -fno-rtti -O2 -lpthread -laio -lz -lm -lcrypt -lssl -lcrypto -ldl

HPC Challenge

HPC Challenge (HPCC) is a cluster-focused benchmark consisting of the HPL Linpack TPP benchmark, DGEMM, STREAM, PTRANS, RandomAccess, FFT, and communication bandwidth and latency. This HPC Challenge test profile attempts to ship with standard yet versatile configuration/input files though they can be modified. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGB/s, More Is BetterHPC Challenge 1.5.0Test / Class: Random Ring BandwidthRyzen 3900X2 x EPYC 75022 x Xeon Platinum 82801.18132.36263.54394.72525.9065SE +/- 0.08933, N = 3SE +/- 0.00593, N = 3SE +/- 0.00164, N = 35.250311.383850.837301. (CC) gcc options: -lblas -lm -pthread -lmpi -fomit-frame-pointer -funroll-loops2. Ryzen 3900X: ATLAS + Open MPI 3.1.43. 2 x EPYC 7502: ATLAS + Open MPI 3.1.34. 2 x Xeon Platinum 8280: ATLAS + Open MPI 3.1.3

OpenBenchmarking.orgGUP/s, More Is BetterHPC Challenge 1.5.0Test / Class: G-Random Access2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X0.04550.0910.13650.1820.2275SE +/- 0.00214, N = 3SE +/- 0.00165, N = 3SE +/- 0.00049, N = 30.202330.175400.032491. (CC) gcc options: -lblas -lm -pthread -lmpi -fomit-frame-pointer -funroll-loops2. 2 x EPYC 7502: ATLAS + Open MPI 3.1.33. 2 x Xeon Platinum 8280: ATLAS + Open MPI 3.1.34. Ryzen 3900X: ATLAS + Open MPI 3.1.4

OpenBenchmarking.orgGFLOP/s, More Is BetterHPC Challenge 1.5.0Test / Class: G-Ffte2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X816243240SE +/- 0.15951, N = 3SE +/- 0.05498, N = 3SE +/- 0.03972, N = 333.3048316.906035.409711. (CC) gcc options: -lblas -lm -pthread -lmpi -fomit-frame-pointer -funroll-loops2. 2 x EPYC 7502: ATLAS + Open MPI 3.1.33. 2 x Xeon Platinum 8280: ATLAS + Open MPI 3.1.34. Ryzen 3900X: ATLAS + Open MPI 3.1.4

OpenBenchmarking.orgGFLOPS, More Is BetterHPC Challenge 1.5.0Test / Class: G-Ffte2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X816243240SE +/- 0.15951, N = 3SE +/- 0.05498, N = 3SE +/- 0.03972, N = 333.3048316.906035.409711. (CC) gcc options: -lblas -lm -pthread -lmpi -fomit-frame-pointer -funroll-loops2. 2 x EPYC 7502: ATLAS + Open MPI 3.1.33. 2 x Xeon Platinum 8280: ATLAS + Open MPI 3.1.34. Ryzen 3900X: ATLAS + Open MPI 3.1.4

MKL-DNN

This is a test of the Intel MKL-DNN as the Intel Math Kernel Library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN 2019-04-16Harness: Deconvolution Batch deconv_all - Data Type: u8s8u8s322 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X40K80K120K160K200KSE +/- 45.90, N = 9SE +/- 81.72, N = 3SE +/- 43.36, N = 3371215307202139-O2 - MIN: 2019281. (CXX) g++ options: -std=c++11 -march=native -mtune=native -fPIC -fopenmp -pie -lmklml_intel -ldl

John The Ripper

This is a benchmark of John The Ripper, which is a password cracker. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgReal C/S, More Is BetterJohn The Ripper 1.9.0-jumbo-1Test: Blowfish2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X20K40K60K80K100KSE +/- 1300.44, N = 3SE +/- 567.99, N = 3SE +/- 37.66, N = 310936383578204601. (CC) gcc options: -m64 -lssl -lcrypto -fopenmp -lgmp -pthread -lm -lz -ldl -lcrypt -lbz2

CP2K Molecular Dynamics

CP2K is an open-source molecular dynamics software package focused on quantum chemistry and solid-state physics. This test profile currently makes use of the OpenMP implementation and using the Fayalite-FIST molecular dynamics run and measures the total time to complete. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterCP2K Molecular Dynamics 6.1Fayalite-FIST DataRyzen 3900X2 x EPYC 75022 x EPYC 77422 x Xeon Platinum 82804008001200160020003215806111699

MKL-DNN

This is a test of the Intel MKL-DNN as the Intel Math Kernel Library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN 2019-04-16Harness: IP Batch 1D - Data Type: u8s8f32s322 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X4080120160200SE +/- 0.04, N = 8SE +/- 1.00, N = 4SE +/- 0.39, N = 33.5568.28176.21-O3 - MIN: 48-O2 - MIN: 171.621. (CXX) g++ options: -std=c++11 -march=native -mtune=native -fPIC -fopenmp -pie -lmklml_intel -ldl

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN 2019-04-16Harness: Convolution Batch conv_alexnet - Data Type: f322 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X5001000150020002500SE +/- 0.18, N = 3SE +/- 0.89, N = 3SE +/- 0.24, N = 348.5780.932393.04-O2 - MIN: 2389.741. (CXX) g++ options: -std=c++11 -march=native -mtune=native -fPIC -fopenmp -pie -lmklml_intel -ldl

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN 2019-04-16Harness: IP Batch 1D - Data Type: u8s8u8s322 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X4080120160200SE +/- 0.03, N = 3SE +/- 0.66, N = 15SE +/- 0.25, N = 33.6468.47177.05-O3 - MIN: 47.15-O2 - MIN: 172.21. (CXX) g++ options: -std=c++11 -march=native -mtune=native -fPIC -fopenmp -pie -lmklml_intel -ldl

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN 2019-04-16Harness: Convolution Batch conv_all - Data Type: f322 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X4K8K12K16K20KSE +/- 0.44, N = 3SE +/- 8.02, N = 9SE +/- 0.66, N = 338262118504-O2 - MIN: 18471.71. (CXX) g++ options: -std=c++11 -march=native -mtune=native -fPIC -fopenmp -pie -lmklml_intel -ldl

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN 2019-04-16Harness: Convolution Batch conv_googlenet_v3 - Data Type: f322 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X2004006008001000SE +/- 0.20, N = 3SE +/- 0.43, N = 12SE +/- 0.10, N = 321.5334.661021.92-O2 - MIN: 1017.721. (CXX) g++ options: -std=c++11 -march=native -mtune=native -fPIC -fopenmp -pie -lmklml_intel -ldl

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN 2019-04-16Harness: Convolution Batch conv_3d - Data Type: u8s8u8s322 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X20K40K60K80K100KSE +/- 3.79, N = 3SE +/- 35.12, N = 3SE +/- 29.28, N = 324383576106118-O2 - MIN: 1060441. (CXX) g++ options: -std=c++11 -march=native -mtune=native -fPIC -fopenmp -pie -lmklml_intel -ldl

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN 2019-04-16Harness: Convolution Batch conv_3d - Data Type: u8s8f32s322 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X20K40K60K80K100KSE +/- 2.13, N = 3SE +/- 2.31, N = 3SE +/- 50.84, N = 324473538106123-O2 - MIN: 1059971. (CXX) g++ options: -std=c++11 -march=native -mtune=native -fPIC -fopenmp -pie -lmklml_intel -ldl

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN 2019-04-16Harness: Deconvolution Batch deconv_3d - Data Type: u8s8f32s322 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X13K26K39K52K65KSE +/- 2.32, N = 3SE +/- 1.15, N = 3SE +/- 70.25, N = 31377189059364-O2 - MIN: 58877.31. (CXX) g++ options: -std=c++11 -march=native -mtune=native -fPIC -fopenmp -pie -lmklml_intel -ldl

Stress-NG

Stress-NG is a Linux stress tool developed by Colin King of Canonical. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.07.26Test: Matrix Math2 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X80K160K240K320K400KSE +/- 1051.08, N = 3SE +/- 512.79, N = 3SE +/- 112.22, N = 3380876375525776031. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.07.26Test: Glibc Qsort Data Sorting2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X2004006008001000SE +/- 6.92, N = 3SE +/- 1.75, N = 3SE +/- 1.95, N = 39317872061. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.07.26Test: CPU Stress2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X4K8K12K16K20KSE +/- 14.09, N = 3SE +/- 121.24, N = 3SE +/- 4.96, N = 3190731795544401. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.07.26Test: Bsearch2 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X8K16K24K32K40KSE +/- 71.90, N = 3SE +/- 238.75, N = 3SE +/- 6.88, N = 3351163317383421. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.07.26Test: SemaphoresRyzen 3900X2 x Xeon Platinum 82802 x EPYC 75021.2M2.4M3.6M4.8M6MSE +/- 15164.55, N = 3SE +/- 23663.90, N = 6SE +/- 7978.41, N = 35724009209893913989721. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc

OpenSSL

OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. This test measures the RSA 4096-bit performance of OpenSSL. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSigns Per Second, More Is BetterOpenSSL 1.1.1RSA 4096-bit Performance2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X3K6K9K12K15KSE +/- 40.05, N = 3SE +/- 64.31, N = 3SE +/- 4.61, N = 3143881375335311. (CC) gcc options: -pthread -m64 -O3 -lssl -lcrypto -ldl

Stress-NG

Stress-NG is a Linux stress tool developed by Colin King of Canonical. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.07.26Test: Crypto2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X3K6K9K12K15KSE +/- 28.56, N = 3SE +/- 24.84, N = 3SE +/- 6.14, N = 3143061050335201. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc

MKL-DNN

This is a test of the Intel MKL-DNN as the Intel Math Kernel Library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN 2019-04-16Harness: Deconvolution Batch deconv_3d - Data Type: u8s8u8s322 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X13K26K39K52K65KSE +/- 6.93, N = 3SE +/- 2.38, N = 3SE +/- 163.13, N = 31507204159053-O2 - MIN: 58346.11. (CXX) g++ options: -std=c++11 -march=native -mtune=native -fPIC -fopenmp -pie -lmklml_intel -ldl

MBW

This is a basic/simple memory (RAM) bandwidth benchmark for memory copy operations. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterMBW 2018-09-08Test: Memory Copy - Array Size: 8192 MiB2 x EPYC 75022 x EPYC 77422 x Xeon Platinum 8280Ryzen 3900X3K6K9K12K15KSE +/- 51.91, N = 3SE +/- 38.55, N = 3SE +/- 0.90, N = 3SE +/- 3.95, N = 3156751555647474431. (CC) gcc options: -O3 -march=native

OpenBenchmarking.orgMiB/s, More Is BetterMBW 2018-09-08Test: Memory Copy, Fixed Block Size - Array Size: 8192 MiB2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X2K4K6K8K10KSE +/- 16.13, N = 3SE +/- 19.23, N = 3SE +/- 9.79, N = 3SE +/- 0.76, N = 39051904251352601. (CC) gcc options: -O3 -march=native

MKL-DNN

This is a test of the Intel MKL-DNN as the Intel Math Kernel Library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN 2019-04-16Harness: Deconvolution Batch deconv_3d - Data Type: f322 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X918273645SE +/- 0.00, N = 3SE +/- 0.05, N = 3SE +/- 0.06, N = 31.113.8638.14-O2 - MIN: 37.771. (CXX) g++ options: -std=c++11 -march=native -mtune=native -fPIC -fopenmp -pie -lmklml_intel -ldl

Stress-NG

Stress-NG is a Linux stress tool developed by Colin King of Canonical. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.07.26Test: Hsearch2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X80K160K240K320K400KSE +/- 782.39, N = 3SE +/- 1743.97, N = 3SE +/- 522.67, N = 3388396301308972371. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.07.26Test: Lsearch2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X150300450600750SE +/- 1.35, N = 3SE +/- 1.81, N = 3SE +/- 0.00, N = 37086651781. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc

Blender

Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing via OpenCL or CUDA is supported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.79aBlend File: Classroom - Compute: CPU-Only2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X7014021028035080.6094.29320.26

MBW

This is a basic/simple memory (RAM) bandwidth benchmark for memory copy operations. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterMBW 2018-09-08Test: Memory Copy - Array Size: 4096 MiBRyzen 3900X2 x EPYC 75022 x EPYC 77422 x Xeon Platinum 82804K8K12K16K20KSE +/- 12.79, N = 3SE +/- 23.34, N = 3SE +/- 51.75, N = 3SE +/- 2.51, N = 318783154721543647521. (CC) gcc options: -O3 -march=native

OpenBenchmarking.orgMiB/s, More Is BetterMBW 2018-09-08Test: Memory Copy - Array Size: 1024 MiBRyzen 3900X2 x EPYC 75022 x EPYC 77422 x Xeon Platinum 82804K8K12K16K20KSE +/- 26.28, N = 3SE +/- 66.85, N = 3SE +/- 35.67, N = 3SE +/- 1.38, N = 318599157111560247751. (CC) gcc options: -O3 -march=native

Stress-NG

Stress-NG is a Linux stress tool developed by Colin King of Canonical. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.07.26Test: Vector Math2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X30K60K90K120K150KSE +/- 244.96, N = 3SE +/- 433.67, N = 3SE +/- 3.94, N = 3130505120957335071. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc

C-Ray

This is a test of C-Ray, a simple raytracer designed to test the floating-point CPU performance. This test is multi-threaded (16 threads per core), will shoot 8 rays per pixel for anti-aliasing, and will generate a 1600 x 1200 image. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterC-Ray 1.1Total Time - 4K, 16 Rays Per Pixel2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X1020304050SE +/- 0.03, N = 3SE +/- 0.10, N = 3SE +/- 0.02, N = 310.9014.4042.221. (CC) gcc options: -lm -lpthread -O3

m-queens

A solver for the N-queens problem with multi-threading support via the OpenMP library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is Betterm-queens 1.2Time To Solve2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X1122334455SE +/- 0.10, N = 3SE +/- 0.03, N = 3SE +/- 0.04, N = 312.1016.3046.751. (CXX) g++ options: -fopenmp -O2 -march=native

MBW

This is a basic/simple memory (RAM) bandwidth benchmark for memory copy operations. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterMBW 2018-09-08Test: Memory Copy - Array Size: 512 MiBRyzen 3900X2 x EPYC 75022 x EPYC 77422 x Xeon Platinum 82804K8K12K16K20KSE +/- 25.17, N = 3SE +/- 59.01, N = 3SE +/- 48.20, N = 3SE +/- 2.75, N = 318439156711546448191. (CC) gcc options: -O3 -march=native

OpenBenchmarking.orgMiB/s, More Is BetterMBW 2018-09-08Test: Memory Copy - Array Size: 128 MiBRyzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 82804K8K12K16K20KSE +/- 17.09, N = 3SE +/- 16.75, N = 3SE +/- 5.56, N = 3SE +/- 3.38, N = 3185219066898649751. (CC) gcc options: -O3 -march=native

Stress-NG

Stress-NG is a Linux stress tool developed by Colin King of Canonical. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.07.26Test: Tsearch2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X400800120016002000SE +/- 6.47, N = 3SE +/- 0.14, N = 3SE +/- 2.30, N = 3175210914751. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc

Blender

Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing via OpenCL or CUDA is supported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.79aBlend File: Pabellon Barcelona - Compute: CPU-Only2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X80160240320400102117370

Coremark

This is a test of EEMBC CoreMark processor benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgIterations/Sec, More Is BetterCoremark 1.0CoreMark Size 666 - Iterations Per Second2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X500K1000K1500K2000K2500KSE +/- 6421.46, N = 3SE +/- 20125.94, N = 3SE +/- 280.48, N = 3210129918906495794461. (CC) gcc options: -O2 -lrt" -lrt

Primesieve

Primesieve generates prime numbers using a highly optimized sieve of Eratosthenes implementation. Primesieve benchmarks the CPU's L1/L2 cache performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterPrimesieve 7.41e12 Prime Number Generation2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X48121620SE +/- 0.05, N = 3SE +/- 0.01, N = 3SE +/- 0.02, N = 34.294.8215.47-O3-O3-O21. (CXX) g++ options: -lpthread

asmFish

This is a test of asmFish, an advanced chess benchmark written in Assembly. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNodes/second, More Is BetterasmFish 2018-07-231024 Hash Memory, 26 Depth2 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X30M60M90M120M150MSE +/- 1715314.92, N = 3SE +/- 769910.95, N = 3SE +/- 231748.24, N = 313737166913625799640037020

Stockfish

This is a test of Stockfish, an advanced C++11 chess benchmark that can scale up to 128 CPU cores. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNodes Per Second, More Is BetterStockfish 9Total Time2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X30M60M90M120M150MSE +/- 810846.41, N = 3SE +/- 879556.81, N = 3SE +/- 247195.47, N = 3135223710125947505400495211. (CXX) g++ options: -m64 -lpthread -fno-exceptions -std=c++11 -pedantic -O3 -msse -msse3 -mpopcnt -flto

7-Zip Compression

This is a test of 7-Zip using p7zip with its integrated benchmark feature or upstream 7-Zip for the Windows x64 build. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMIPS, More Is Better7-Zip Compression 16.02Compress Speed Test2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X50K100K150K200K250KSE +/- 3053.66, N = 3SE +/- 924.31, N = 3SE +/- 207.54, N = 3255492232626775851. (CXX) g++ options: -pipe -lpthread

Phoronix Test Suite Self Test

This is a wrapper around the phoronix-test-suite debug-self-test sub-command for testing hot Phoronix Test Suite code paths and if wanting to compare different PHP implementations. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterPhoronix Test Suite Self Test 1.0Test: Total Time2 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X150300450600750SE +/- 1.33, N = 3SE +/- 0.68, N = 3SE +/- 0.44, N = 2215253697

Blender

Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing via OpenCL or CUDA is supported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.79aBlend File: Barbershop - Compute: CPU-Only2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X150300450600750221223707

MariaDB

This is a MariaDB MySQL database server benchmark making use of mysqlslap. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 10.3.8Clients: 322 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X6001200180024003000SE +/- 31.65, N = 3SE +/- 23.57, N = 4SE +/- 1.21, N = 329051757914-llzma -lbz2 -lnuma -lpcre-llzma -lbz2 -lnuma -lpcre1. (CXX) g++ options: -pie -fPIC -fstack-protector -fno-rtti -O2 -lpthread -laio -lz -lm -lcrypt -lssl -lcrypto -ldl

POV-Ray

This is a test of POV-Ray, the Persistence of Vision Raytracer. POV-Ray is used to create 3D graphics using ray-tracing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterPOV-Ray 3.7.0.7Trace Time2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X714212835SE +/- 0.04, N = 3SE +/- 0.12, N = 3SE +/- 0.03, N = 310.3211.5731.55-lSDL -lXpm -lSM -lICE -lIlmImf -lImath -lHalf -lIex -lIexMath -lIlmThread-lXpm -lSM -lICE -lIlmImf -lImath -lHalf -lIex -lIexMath -lIlmThread-lSDL1. (CXX) g++ options: -pipe -O3 -ffast-math -march=native -pthread -lX11 -lpthread -ltiff -ljpeg -lpng -lz -lrt -lm -lboost_thread -lboost_system

Blender

Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing via OpenCL or CUDA is supported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.79aBlend File: BMW27 - Compute: CPU-Only2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X30609012015045.2047.45138.06

Renaissance

Renaissance is a suite of benchmarks designed to test the Java JVM from Apache Spark to a Twitter-like service to Scala and other features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.9.0Test: In-Memory Database ShootoutRyzen 3900X2 x EPYC 75022 x Xeon Platinum 82804K8K12K16K20KSE +/- 35.03, N = 8SE +/- 120.36, N = 11SE +/- 158.15, N = 1057351368317338

PostgreSQL pgbench

This is a simple benchmark of PostgreSQL using pgbench. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 10.3Scaling: Buffer Test - Test: Normal Load - Mode: Read Only2 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X200K400K600K800K1000KSE +/- 3679.97, N = 3SE +/- 8471.64, N = 3SE +/- 689.91, N = 39128068583923023011. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lpthread -lrt -lcrypt -ldl -lm

Phoronix Test Suite Self Test

This is a wrapper around the phoronix-test-suite debug-self-test sub-command for testing hot Phoronix Test Suite code paths and if wanting to compare different PHP implementations. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterPhoronix Test Suite Self Test 1.0Test: List Suites2 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X510152025SE +/- 0.00, N = 2SE +/- 0.01, N = 2SE +/- 0.00, N = 20.710.7420.17

HPC Challenge

HPC Challenge (HPCC) is a cluster-focused benchmark consisting of the HPL Linpack TPP benchmark, DGEMM, STREAM, PTRANS, RandomAccess, FFT, and communication bandwidth and latency. This HPC Challenge test profile attempts to ship with standard yet versatile configuration/input files though they can be modified. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOPS, More Is BetterHPC Challenge 1.5.0Test / Class: EP-DGEMM2 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X612182430SE +/- 0.05258, N = 3SE +/- 0.11773, N = 3SE +/- 0.00010, N = 325.1735322.389701.076551. (CC) gcc options: -lblas -lm -pthread -lmpi -fomit-frame-pointer -funroll-loops2. 2 x Xeon Platinum 8280: ATLAS + Open MPI 3.1.33. 2 x EPYC 7502: ATLAS + Open MPI 3.1.34. Ryzen 3900X: ATLAS + Open MPI 3.1.4

MKL-DNN

This is a test of the Intel MKL-DNN as the Intel Math Kernel Library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN 2019-04-16Harness: Deconvolution Batch deconv_1d - Data Type: f322 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X4080120160200SE +/- 0.01, N = 3SE +/- 0.08, N = 3SE +/- 0.22, N = 30.958.11197.67-O2 - MIN: 196.861. (CXX) g++ options: -std=c++11 -march=native -mtune=native -fPIC -fopenmp -pie -lmklml_intel -ldl

HPC Challenge

HPC Challenge (HPCC) is a cluster-focused benchmark consisting of the HPL Linpack TPP benchmark, DGEMM, STREAM, PTRANS, RandomAccess, FFT, and communication bandwidth and latency. This HPC Challenge test profile attempts to ship with standard yet versatile configuration/input files though they can be modified. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgusecs, Fewer Is BetterHPC Challenge 1.5.0Test / Class: Random Ring LatencyRyzen 3900X2 x Xeon Platinum 82802 x EPYC 75020.40040.80081.20121.60162.002SE +/- 0.00148, N = 3SE +/- 0.00802, N = 3SE +/- 0.00742, N = 30.596171.133561.779761. (CC) gcc options: -lblas -lm -pthread -lmpi -fomit-frame-pointer -funroll-loops2. Ryzen 3900X: ATLAS + Open MPI 3.1.43. 2 x Xeon Platinum 8280: ATLAS + Open MPI 3.1.34. 2 x EPYC 7502: ATLAS + Open MPI 3.1.3

Blender

Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing via OpenCL or CUDA is supported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.79aBlend File: Fishy Cat - Compute: CPU-Only2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X408012016020066.5272.54194.00

Himeno Benchmark

The Himeno benchmark is a linear solver of pressure Poisson using a point-Jacobi method. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMFLOPS, More Is BetterHimeno Benchmark 3.0Poisson Pressure Solver2 x Xeon Platinum 8280Ryzen 3900X2 x EPYC 75026001200180024003000SE +/- 4.24, N = 3SE +/- 19.25, N = 3SE +/- 4.17, N = 32794127810041. (CC) gcc options: -O3 -mavx2

Appleseed

Appleseed is an open-source production renderer focused on physically-based global illumination rendering engine primarily designed for animation and visual effects. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterAppleseed 2.0 BetaScene: Disney Material2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X408012016020059.8067.24164.44

Stress-NG

Stress-NG is a Linux stress tool developed by Colin King of Canonical. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.07.26Test: System V Message Passing2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X3M6M9M12M15MSE +/- 99301.92, N = 15SE +/- 79946.02, N = 3SE +/- 23189.98, N = 312216045655171446454261. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc

Parboil

The Parboil Benchmarks from the IMPACT Research Group at University of Illinois are a set of throughput computing applications for looking at computing architecture and compilers. Parboil test-cases support OpenMP, OpenCL, and CUDA multi-processing environments. However, at this time the test profile is just making use of the OpenMP and OpenCL test workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterParboil 2.5Test: OpenMP CUTCP2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X0.46580.93161.39741.86322.329SE +/- 0.01, N = 3SE +/- 0.02, N = 4SE +/- 0.02, N = 12SE +/- 0.01, N = 30.801.041.292.071. (CXX) g++ options: -lm -lpthread -lgomp -O3 -ffast-math -fopenmp

Phoronix Test Suite Self Test

This is a wrapper around the phoronix-test-suite debug-self-test sub-command for testing hot Phoronix Test Suite code paths and if wanting to compare different PHP implementations. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterPhoronix Test Suite Self Test 1.0Test: Info CommandRyzen 3900X2 x Xeon Platinum 82802 x EPYC 75020.6931.3862.0792.7723.465SE +/- 0.01, N = 2SE +/- 0.00, N = 2SE +/- 0.05, N = 21.202.403.08

Timed LLVM Compilation

This test times how long it takes to build the LLVM compiler. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed LLVM Compilation 6.0.1Time To Compile2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X5010015020025096.43107.93245.50

Memcached mcperf

This is a test of twmperf/mcperf with memcached. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOperations Per Second, More Is BetterMemcached mcperf 1.5.10Method: Replace2 x Xeon Platinum 8280Ryzen 3900X2 x EPYC 750215K30K45K60K75KSE +/- 251.51, N = 3SE +/- 276.40, N = 3SE +/- 86.93, N = 37118546395283231. (CC) gcc options: -O2 -lm -rdynamic

Y-Cruncher

Y-Cruncher is a multi-threaded Pi benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterY-Cruncher 0.7.7Calculating 500M Pi Digits2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X510152025SE +/- 0.02, N = 3SE +/- 0.05, N = 3SE +/- 0.00, N = 38.899.1221.92

PostgreSQL pgbench

This is a simple benchmark of PostgreSQL using pgbench. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 10.3Scaling: Buffer Test - Test: Single Thread - Mode: Read Write2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X6001200180024003000SE +/- 14.28, N = 3SE +/- 39.04, N = 3SE +/- 5.82, N = 32689265611501. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lpthread -lrt -lcrypt -ldl -lm

HPC Challenge

HPC Challenge (HPCC) is a cluster-focused benchmark consisting of the HPL Linpack TPP benchmark, DGEMM, STREAM, PTRANS, RandomAccess, FFT, and communication bandwidth and latency. This HPC Challenge test profile attempts to ship with standard yet versatile configuration/input files though they can be modified. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterHPC Challenge 1.5.0Test / Class: Max Ping Pong BandwidthRyzen 3900X2 x Xeon Platinum 82802 x EPYC 75025K10K15K20K25KSE +/- 67.88, N = 3SE +/- 56.79, N = 3SE +/- 77.41, N = 32502317274107811. (CC) gcc options: -lblas -lm -pthread -lmpi -fomit-frame-pointer -funroll-loops2. Ryzen 3900X: ATLAS + Open MPI 3.1.43. 2 x Xeon Platinum 8280: ATLAS + Open MPI 3.1.34. 2 x EPYC 7502: ATLAS + Open MPI 3.1.3

OSBench

OSBench is a collection of micro-benchmarks for measuring operating system primitives like time to create threads/processes, launching programs, creating files, and memory allocation. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgus Per Event, Fewer Is BetterOSBenchTest: Create ProcessesRyzen 3900X2 x Xeon Platinum 82802 x EPYC 75021428425670SE +/- 0.25, N = 3SE +/- 0.79, N = 3SE +/- 0.76, N = 327.6345.9263.311. (CC) gcc options: -lm

BLAKE2

This is a benchmark of BLAKE2 using the blake2s binary. BLAKE2 is a high-performance crypto alternative to MD5 and SHA-2/3. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgCycles Per Byte, Fewer Is BetterBLAKE2 201703072 x Xeon Platinum 82802 x EPYC 77422 x EPYC 7502Ryzen 3900X246810SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.04, N = 33.876.607.468.301. (CC) gcc options: -O3 -march=native -lcrypto -lz

Zstd Compression

This test measures the time needed to compress a sample file (an Ubuntu file-system image) using Zstd compression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterZstd Compression 1.3.4Compressing ubuntu-16.04.3-server-i386.img, Compression Level 192 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X510152025SE +/- 0.10, N = 6SE +/- 0.05, N = 3SE +/- 0.07, N = 39.0111.2018.96-llzma-llzma1. (CC) gcc options: -O3 -pthread -lz

Memcached mcperf

This is a test of twmperf/mcperf with memcached. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOperations Per Second, More Is BetterMemcached mcperf 1.5.10Method: Get2 x Xeon Platinum 8280Ryzen 3900X2 x EPYC 750220K40K60K80K100KSE +/- 712.60, N = 3SE +/- 1204.09, N = 3SE +/- 464.12, N = 1511193078724536501. (CC) gcc options: -O2 -lm -rdynamic

OSBench

OSBench is a collection of micro-benchmarks for measuring operating system primitives like time to create threads/processes, launching programs, creating files, and memory allocation. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgus Per Event, Fewer Is BetterOSBenchTest: Launch ProgramsRyzen 3900X2 x Xeon Platinum 82802 x EPYC 750220406080100SE +/- 0.19, N = 3SE +/- 0.52, N = 9SE +/- 0.82, N = 337.8554.5178.171. (CC) gcc options: -lm

MBW

This is a basic/simple memory (RAM) bandwidth benchmark for memory copy operations. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterMBW 2018-09-08Test: Memory Copy, Fixed Block Size - Array Size: 1024 MiBRyzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 82802K4K6K8K10KSE +/- 12.73, N = 3SE +/- 3.77, N = 3SE +/- 4.01, N = 3SE +/- 12.12, N = 399639153909349161. (CC) gcc options: -O3 -march=native

Memcached mcperf

This is a test of twmperf/mcperf with memcached. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOperations Per Second, More Is BetterMemcached mcperf 1.5.10Method: Append2 x Xeon Platinum 8280Ryzen 3900X2 x EPYC 750215K30K45K60K75KSE +/- 150.54, N = 3SE +/- 54.84, N = 3SE +/- 247.66, N = 37047846554348501. (CC) gcc options: -O2 -lm -rdynamic

MBW

This is a basic/simple memory (RAM) bandwidth benchmark for memory copy operations. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterMBW 2018-09-08Test: Memory Copy, Fixed Block Size - Array Size: 512 MiBRyzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 82802K4K6K8K10KSE +/- 10.38, N = 3SE +/- 17.16, N = 3SE +/- 3.34, N = 3SE +/- 1.22, N = 398969188912649101. (CC) gcc options: -O3 -march=native

OpenBenchmarking.orgMiB/s, More Is BetterMBW 2018-09-08Test: Memory Copy, Fixed Block Size - Array Size: 4096 MiBRyzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 82802K4K6K8K10KSE +/- 13.37, N = 3SE +/- 7.99, N = 3SE +/- 16.61, N = 3SE +/- 22.38, N = 3101709032901750531. (CC) gcc options: -O3 -march=native

MKL-DNN

This is a test of the Intel MKL-DNN as the Intel Math Kernel Library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN 2019-04-16Harness: Convolution Batch conv_googlenet_v3 - Data Type: u8s8u8s322 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X2K4K6K8K10KSE +/- 0.01, N = 3SE +/- 14.43, N = 3SE +/- 101.14, N = 35.771752.4711526.10-O3 - MIN: 1493.25-O2 - MIN: 11291.81. (CXX) g++ options: -std=c++11 -march=native -mtune=native -fPIC -fopenmp -pie -lmklml_intel -ldl

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN 2019-04-16Harness: Convolution Batch conv_alexnet - Data Type: u8s8u8s322 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X6K12K18K24K30KSE +/- 0.03, N = 3SE +/- 0.75, N = 3SE +/- 213.03, N = 314.23947.2227767.67-O2 - MIN: 27421.81. (CXX) g++ options: -std=c++11 -march=native -mtune=native -fPIC -fopenmp -pie -lmklml_intel -ldl

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN 2019-04-16Harness: Convolution Batch conv_googlenet_v3 - Data Type: u8s8f32s322 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X2K4K6K8K10KSE +/- 0.00, N = 3SE +/- 14.62, N = 3SE +/- 34.01, N = 36.231764.5811596.50-O3 - MIN: 1518.91-O2 - MIN: 11477.41. (CXX) g++ options: -std=c++11 -march=native -mtune=native -fPIC -fopenmp -pie -lmklml_intel -ldl

Phoronix Test Suite Self Test

This is a wrapper around the phoronix-test-suite debug-self-test sub-command for testing hot Phoronix Test Suite code paths and if wanting to compare different PHP implementations. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterPhoronix Test Suite Self Test 1.0Test: Merge Results2 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X1632486480SE +/- 0.00, N = 2SE +/- 0.03, N = 2SE +/- 0.00, N = 23.914.8474.28

MKL-DNN

This is a test of the Intel MKL-DNN as the Intel Math Kernel Library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN 2019-04-16Harness: Convolution Batch conv_all - Data Type: u8s8u8s322 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X60K120K180K240K300KSE +/- 7.41, N = 3SE +/- 105.95, N = 3SE +/- 332.06, N = 3172220800294635-O2 - MIN: 2932281. (CXX) g++ options: -std=c++11 -march=native -mtune=native -fPIC -fopenmp -pie -lmklml_intel -ldl

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN 2019-04-16Harness: Convolution Batch conv_all - Data Type: u8s8f32s322 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X60K120K180K240K300KSE +/- 6.52, N = 3SE +/- 183.52, N = 3SE +/- 402.11, N = 3175620778293148-O2 - MIN: 2920731. (CXX) g++ options: -std=c++11 -march=native -mtune=native -fPIC -fopenmp -pie -lmklml_intel -ldl

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN 2019-04-16Harness: Deconvolution Batch deconv_1d - Data Type: u8s8f32s322 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X8K16K24K32K40KSE +/- 0.00, N = 4SE +/- 0.82, N = 3SE +/- 116.40, N = 30.23795.6835857.17-O2 - MIN: 35696.31. (CXX) g++ options: -std=c++11 -march=native -mtune=native -fPIC -fopenmp -pie -lmklml_intel -ldl

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN 2019-04-16Harness: Deconvolution Batch deconv_1d - Data Type: u8s8u8s322 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X8K16K24K32K40KSE +/- 0.00, N = 3SE +/- 1.86, N = 3SE +/- 20.50, N = 30.23881.3335756.50-O2 - MIN: 35695.51. (CXX) g++ options: -std=c++11 -march=native -mtune=native -fPIC -fopenmp -pie -lmklml_intel -ldl

Parboil

The Parboil Benchmarks from the IMPACT Research Group at University of Illinois are a set of throughput computing applications for looking at computing architecture and compilers. Parboil test-cases support OpenMP, OpenCL, and CUDA multi-processing environments. However, at this time the test profile is just making use of the OpenMP and OpenCL test workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterParboil 2.5Test: OpenMP MRI GriddingRyzen 3900X2 x EPYC 75022 x EPYC 77422 x Xeon Platinum 8280110220330440550SE +/- 0.01, N = 3SE +/- 0.54, N = 3SE +/- 2.09, N = 6SE +/- 7.08, N = 934.16120.92182.83513.311. (CXX) g++ options: -lm -lpthread -lgomp -O3 -ffast-math -fopenmp

MKL-DNN

This is a test of the Intel MKL-DNN as the Intel Math Kernel Library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN 2019-04-16Harness: Convolution Batch conv_alexnet - Data Type: u8s8f32s322 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X6K12K18K24K30KSE +/- 0.05, N = 3SE +/- 2.17, N = 3SE +/- 101.98, N = 318.95941.6927648.10-O2 - MIN: 27449.31. (CXX) g++ options: -std=c++11 -march=native -mtune=native -fPIC -fopenmp -pie -lmklml_intel -ldl

MariaDB

This is a MariaDB MySQL database server benchmark making use of mysqlslap. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 10.3.8Clients: 42 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X3K6K9K12K15KSE +/- 154.49, N = 3SE +/- 33.65, N = 3SE +/- 11.81, N = 3132787634953-llzma -lbz2 -lnuma -lpcre-llzma -lbz2 -lnuma -lpcre1. (CXX) g++ options: -pie -fPIC -fstack-protector -fno-rtti -O2 -lpthread -laio -lz -lm -lcrypt -lssl -lcrypto -ldl

HPC Challenge

HPC Challenge (HPCC) is a cluster-focused benchmark consisting of the HPL Linpack TPP benchmark, DGEMM, STREAM, PTRANS, RandomAccess, FFT, and communication bandwidth and latency. This HPC Challenge test profile attempts to ship with standard yet versatile configuration/input files though they can be modified. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOPS, More Is BetterHPC Challenge 1.5.0Test / Class: G-HPL2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X4080120160200SE +/- 0.06, N = 3SE +/- 0.03, N = 3SE +/- 0.00, N = 3169.92146.4512.791. (CC) gcc options: -lblas -lm -pthread -lmpi -fomit-frame-pointer -funroll-loops2. 2 x EPYC 7502: ATLAS + Open MPI 3.1.33. 2 x Xeon Platinum 8280: ATLAS + Open MPI 3.1.34. Ryzen 3900X: ATLAS + Open MPI 3.1.4

MKL-DNN

This is a test of the Intel MKL-DNN as the Intel Math Kernel Library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN 2019-04-16Harness: IP Batch All - Data Type: u8s8f32s322 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X5001000150020002500SE +/- 0.27, N = 3SE +/- 5.40, N = 3SE +/- 9.52, N = 320.48778.492276.91-O3 - MIN: 664.03-O2 - MIN: 2248.791. (CXX) g++ options: -std=c++11 -march=native -mtune=native -fPIC -fopenmp -pie -lmklml_intel -ldl

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN 2019-04-16Harness: IP Batch All - Data Type: u8s8u8s322 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X5001000150020002500SE +/- 0.05, N = 3SE +/- 13.00, N = 3SE +/- 11.34, N = 320.42772.472265.86-O3 - MIN: 607.02-O2 - MIN: 2243.091. (CXX) g++ options: -std=c++11 -march=native -mtune=native -fPIC -fopenmp -pie -lmklml_intel -ldl

Stream

This benchmark tests the system memory (RAM) performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterStream 2013-01-17Type: Scale2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X40K80K120K160K200KSE +/- 986.94, N = 5SE +/- 1282.73, N = 5SE +/- 682.69, N = 5SE +/- 22.49, N = 5177029155497149490151701. (CC) gcc options: -O3 -march=native -fopenmp

MariaDB

This is a MariaDB MySQL database server benchmark making use of mysqlslap. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 10.3.8Clients: 82 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X2K4K6K8K10KSE +/- 140.23, N = 4SE +/- 49.47, N = 3SE +/- 1.81, N = 3104226397986-llzma -lbz2 -lnuma -lpcre-llzma -lbz2 -lnuma -lpcre1. (CXX) g++ options: -pie -fPIC -fstack-protector -fno-rtti -O2 -lpthread -laio -lz -lm -lcrypt -lssl -lcrypto -ldl

Phoronix Test Suite Self Test

This is a wrapper around the phoronix-test-suite debug-self-test sub-command for testing hot Phoronix Test Suite code paths and if wanting to compare different PHP implementations. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterPhoronix Test Suite Self Test 1.0Test: Render TestRyzen 3900X2 x Xeon Platinum 82802 x EPYC 75021632486480SE +/- 0.12, N = 2SE +/- 0.37, N = 2SE +/- 0.25, N = 235.2959.8069.67

FFTW

FFTW is a C subroutine library for computing the discrete Fourier transform (DFT) in one or more dimensions. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Float + SSE - Size: 2D FFT Size 512Ryzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 82809K18K27K36K45KSE +/- 311.93, N = 3SE +/- 85.48, N = 3SE +/- 128.07, N = 3SE +/- 15.95, N = 3426713417833879222891. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

MBW

This is a basic/simple memory (RAM) bandwidth benchmark for memory copy operations. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterMBW 2018-09-08Test: Memory Copy, Fixed Block Size - Array Size: 128 MiBRyzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 82802K4K6K8K10KSE +/- 16.17, N = 3SE +/- 9.56, N = 3SE +/- 30.78, N = 3SE +/- 10.85, N = 393939053898449181. (CC) gcc options: -O3 -march=native

FFTW

FFTW is a C subroutine library for computing the discrete Fourier transform (DFT) in one or more dimensions. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Float + SSE - Size: 2D FFT Size 1024Ryzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 82809K18K27K36K45KSE +/- 255.01, N = 3SE +/- 264.46, N = 3SE +/- 141.25, N = 3SE +/- 40.70, N = 3417993389533251223511. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

RAMspeed SMP

This benchmark tests the system memory (RAM) performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterRAMspeed SMP 3.5.0Type: Copy - Benchmark: Integer2 x EPYC 75022 x EPYC 77422 x Xeon Platinum 8280Ryzen 3900X9K18K27K36K45K427394110532285234241. (CC) gcc options: -O3 -march=native

OpenBenchmarking.orgMB/s, More Is BetterRAMspeed SMP 3.5.0Type: Add - Benchmark: Floating Point2 x EPYC 75022 x EPYC 77422 x Xeon Platinum 8280Ryzen 3900X10K20K30K40K50K485474853028400269571. (CC) gcc options: -O3 -march=native

Rodinia

Rodinia is a suite focused upon accelerating compute-intensive applications with accelerators. CUDA, OpenMP, and OpenCL parallel models are supported by the included applications. This profile utilizes the OpenCL and OpenMP test binaries at the moment. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 2.4Test: OpenMP CFD Solver2 x Xeon Platinum 82802 x EPYC 75022 x EPYC 7742Ryzen 3900X3691215SE +/- 0.10, N = 15SE +/- 0.13, N = 15SE +/- 0.13, N = 3SE +/- 0.01, N = 37.318.879.1013.141. (CXX) g++ options: -O2 -lOpenCL

RAMspeed SMP

This benchmark tests the system memory (RAM) performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterRAMspeed SMP 3.5.0Type: Add - Benchmark: Integer2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X10K20K30K40K50K482534755232779270161. (CC) gcc options: -O3 -march=native

FFTW

FFTW is a C subroutine library for computing the discrete Fourier transform (DFT) in one or more dimensions. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Float + SSE - Size: 2D FFT Size 256Ryzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 82809K18K27K36K45KSE +/- 217.10, N = 3SE +/- 194.96, N = 3SE +/- 93.28, N = 3SE +/- 413.74, N = 3442033428433586248531. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Float + SSE - Size: 2D FFT Size 32Ryzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 828010K20K30K40K50KSE +/- 306.88, N = 3SE +/- 76.70, N = 3SE +/- 35.51, N = 3SE +/- 223.03, N = 3481953502834572271281. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

Appleseed

Appleseed is an open-source production renderer focused on physically-based global illumination rendering engine primarily designed for animation and visual effects. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterAppleseed 2.0 BetaScene: Material TesterRyzen 3900X2 x Xeon Platinum 82802 x EPYC 750260120180240300161185284

Apache Siege

This is a test of the Apache web server performance being facilitated by the Siege web serverb enchmark program. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTransactions Per Second, More Is BetterApache Siege 2.4.29Concurrent Users: 12 x EPYC 75022 x Xeon Platinum 828010002000300040005000SE +/- 36.83, N = 3SE +/- 8.15, N = 34559.862606.481. (CC) gcc options: -O2 -lpthread -ldl -lssl -lcrypto

Java SciMark

This test runs the Java version of SciMark 2.0, which is a benchmark for scientific and numerical computing developed by programmers at the National Institute of Standards and Technology. This benchmark is made up of Fast Foruier Transform, Jacobi Successive Over-relaxation, Monte Carlo, Sparse Matrix Multiply, and dense LU matrix factorization benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterJava SciMark 2.0Computational Test: Monte CarloRyzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 8280400800120016002000SE +/- 9.95, N = 4SE +/- 0.37, N = 4SE +/- 2.07, N = 4SE +/- 0.22, N = 4171613561334984

Appleseed

Appleseed is an open-source production renderer focused on physically-based global illumination rendering engine primarily designed for animation and visual effects. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterAppleseed 2.0 BetaScene: Emily2 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X60120180240300155173270

PostgreSQL pgbench

This is a simple benchmark of PostgreSQL using pgbench. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 10.3Scaling: Buffer Test - Test: Single Thread - Mode: Read OnlyRyzen 3900X2 x EPYC 75022 x Xeon Platinum 82807K14K21K28K35KSE +/- 270.22, N = 3SE +/- 102.32, N = 3SE +/- 48.37, N = 33308520388190161. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lpthread -lrt -lcrypt -ldl -lm

MariaDB

This is a MariaDB MySQL database server benchmark making use of mysqlslap. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 10.3.8Clients: 128Ryzen 3900X2 x Xeon Platinum 82802 x EPYC 7502130260390520650SE +/- 0.83, N = 3SE +/- 0.98, N = 3SE +/- 3.87, N = 3589544342-llzma -lbz2 -lnuma -lpcre-llzma -lbz2 -lnuma -lpcre1. (CXX) g++ options: -pie -fPIC -fstack-protector -fno-rtti -O2 -lpthread -laio -lz -lm -lcrypt -lssl -lcrypto -ldl

RAMspeed SMP

This benchmark tests the system memory (RAM) performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterRAMspeed SMP 3.5.0Type: Scale - Benchmark: Floating Point2 x EPYC 75022 x EPYC 77422 x Xeon Platinum 8280Ryzen 3900X9K18K27K36K45K400263619625822233391. (CC) gcc options: -O3 -march=native

BlogBench

BlogBench is designed to replicate the load of a real-world busy file server by stressing the file-system with multiple threads of random reads, writes, and rewrites. The behavior is mimicked of that of a blog by creating blogs with content and pictures, modifying blog posts, adding comments to these blogs, and then reading the content of the blogs. All of these blogs generated are created locally with fake content and pictures. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFinal Score, More Is BetterBlogBench 1.1Test: Read2 x Xeon Platinum 8280Ryzen 3900X2 x EPYC 75022 x EPYC 7742400K800K1200K1600K2000KSE +/- 20800.00, N = 3SE +/- 3745.14, N = 3SE +/- 4626.16, N = 3SE +/- 14055.95, N = 319952391298958127205611640671. (CC) gcc options: -O2 -pthread

Bork File Encrypter

Bork is a small, cross-platform file encryption utility. It is written in Java and designed to be included along with the files it encrypts for long-term storage. This test measures the amount of time it takes to encrypt a sample file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBork File Encrypter 1.4File Encryption Time2 x Xeon Platinum 82802 x EPYC 77422 x EPYC 7502Ryzen 3900X3691215SE +/- 0.02, N = 3SE +/- 0.05, N = 3SE +/- 0.05, N = 3SE +/- 0.22, N = 37.1010.7310.7712.08

FFTW

FFTW is a C subroutine library for computing the discrete Fourier transform (DFT) in one or more dimensions. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Float + SSE - Size: 2D FFT Size 128Ryzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 828010K20K30K40K50KSE +/- 891.44, N = 3SE +/- 152.49, N = 3SE +/- 72.57, N = 3SE +/- 356.69, N = 3485123500334549286131. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

RAMspeed SMP

This benchmark tests the system memory (RAM) performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterRAMspeed SMP 3.5.0Type: Triad - Benchmark: Floating Point2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X10K20K30K40K50K465213224728432274591. (CC) gcc options: -O3 -march=native

dav1d

Dav1d is an open-source, speedy AV1 video decoder. This test profile times how long it takes to decode some sample AV1 video content. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is Betterdav1d 0.3Video Input: Summer Nature 4K2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X510152025SE +/- 0.07, N = 3SE +/- 0.04, N = 3SE +/- 0.05, N = 313.2418.2522.261. (CC) gcc options: -pthread

dcraw

This test times how long it takes to convert several high-resolution RAW NEF image files to PPM image format using dcraw. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterdcrawRAW To PPM Image ConversionRyzen 3900X2 x Xeon Platinum 82802 x EPYC 75021224364860SE +/- 0.06, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 331.1541.7751.431. (CC) gcc options: -lm

BlogBench

BlogBench is designed to replicate the load of a real-world busy file server by stressing the file-system with multiple threads of random reads, writes, and rewrites. The behavior is mimicked of that of a blog by creating blogs with content and pictures, modifying blog posts, adding comments to these blogs, and then reading the content of the blogs. All of these blogs generated are created locally with fake content and pictures. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFinal Score, More Is BetterBlogBench 1.1Test: Write2 x Xeon Platinum 82802 x EPYC 75022 x EPYC 7742Ryzen 3900X11K22K33K44K55KSE +/- 449.95, N = 3SE +/- 882.24, N = 3SE +/- 7.00, N = 2SE +/- 498.08, N = 3535164380738226327871. (CC) gcc options: -O2 -pthread

RAMspeed SMP

This benchmark tests the system memory (RAM) performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterRAMspeed SMP 3.5.0Type: Average - Benchmark: Floating Point2 x EPYC 75022 x EPYC 77422 x Xeon Platinum 8280Ryzen 3900X9K18K27K36K45K442754096029060273141. (CC) gcc options: -O3 -march=native

dav1d

Dav1d is an open-source, speedy AV1 video decoder. This test profile times how long it takes to decode some sample AV1 video content. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is Betterdav1d 0.3Video Input: Summer Nature 1080p2 x EPYC 7502Ryzen 3900X2 x Xeon Platinum 82803691215SE +/- 0.02, N = 3SE +/- 0.03, N = 3SE +/- 0.02, N = 35.697.499.201. (CC) gcc options: -pthread

RAMspeed SMP

This benchmark tests the system memory (RAM) performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterRAMspeed SMP 3.5.0Type: Scale - Benchmark: Integer2 x EPYC 75022 x EPYC 77422 x Xeon Platinum 8280Ryzen 3900X8K16K24K32K40K384653770029096238521. (CC) gcc options: -O3 -march=native

Stress-NG

Stress-NG is a Linux stress tool developed by Colin King of Canonical. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.07.26Test: Forking2 x Xeon Platinum 8280Ryzen 3900X2 x EPYC 750215K30K45K60K75KSE +/- 870.64, N = 4SE +/- 349.82, N = 3SE +/- 120.17, N = 36899063865430211. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc

MariaDB

This is a MariaDB MySQL database server benchmark making use of mysqlslap. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 10.3.8Clients: 642 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X30060090012001500SE +/- 18.28, N = 3SE +/- 0.93, N = 3SE +/- 0.99, N = 31286835804-llzma -lbz2 -lnuma -lpcre-llzma -lbz2 -lnuma -lpcre1. (CXX) g++ options: -pie -fPIC -fstack-protector -fno-rtti -O2 -lpthread -laio -lz -lm -lcrypt -lssl -lcrypto -ldl

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 9.12-MR1Java Test: Tradebeans2 x EPYC 75022 x EPYC 77422 x Xeon Platinum 82803K6K9K12K15KSE +/- 59.83, N = 4SE +/- 56.27, N = 4SE +/- 98.03, N = 47419774311780

RAMspeed SMP

This benchmark tests the system memory (RAM) performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterRAMspeed SMP 3.5.0Type: Average - Benchmark: Integer2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X9K18K27K36K45K422724202331645273331. (CC) gcc options: -O3 -march=native

Botan

Botan is a cross-platform open-source C++ crypto library that supports most all publicly known cryptographic algorithms. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.8.0Test: AES-256 - EncryptRyzen 3900X2 x EPYC 75022 x Xeon Platinum 828013002600390052006500SE +/- 4.49, N = 3SE +/- 1.53, N = 3SE +/- 1.15, N = 35910458538231. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt

SVT-AV1

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-AV1 CPU-based multi-threaded video encoder for the AV1 video format with a sample 1080p YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-AV1 0.51080p 8-bit YUV To AV1 Video Encode2 x EPYC 75022 x Xeon Platinum 828020406080100SE +/- 0.46, N = 3SE +/- 0.65, N = 394.9961.531. (CXX) g++ options: -O3 -pie -lpthread -lm

Botan

Botan is a cross-platform open-source C++ crypto library that supports most all publicly known cryptographic algorithms. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.8.0Test: AES-256 - DecryptRyzen 3900X2 x EPYC 75022 x Xeon Platinum 828013002600390052006500SE +/- 11.04, N = 3SE +/- 4.72, N = 3SE +/- 1.30, N = 35896464038281. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt

TTSIOD 3D Renderer

A portable GPL 3D software renderer that supports OpenMP and Intel Threading Building Blocks with many different rendering modes. This version does not use OpenGL but is entirely CPU/software based. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFPS, More Is BetterTTSIOD 3D Renderer 2.3bPhong Rendering With Soft-Shadow Mapping2 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X2004006008001000SE +/- 12.66, N = 3SE +/- 9.93, N = 15SE +/- 1.32, N = 31010693656-lpthread1. (CXX) g++ options: -O3 -fomit-frame-pointer -ffast-math -mtune=native -flto -msse -mrecip -mfpmath=sse -msse2 -mssse3 -lSDL -fopenmp -fwhole-program -lstdc++

VP9 libvpx Encoding

This is a standard video encoding performance test of Google's libvpx library and the vpxenc command for the VP9/WebM format using a sample 1080p video. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterVP9 libvpx Encoding 1.8.0vpxenc VP9 1080p Video Encode2 x EPYC 75022 x Xeon Platinum 8280306090120150SE +/- 0.79, N = 3SE +/- 1.15, N = 3145.2994.821. (CXX) g++ options: -m64 -lm -lpthread -O3 -fPIC -U_FORTIFY_SOURCE -std=c++11

RAMspeed SMP

This benchmark tests the system memory (RAM) performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterRAMspeed SMP 3.5.0Type: Copy - Benchmark: Floating Point2 x EPYC 77422 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X8K16K24K32K40K355263211030275233941. (CC) gcc options: -O3 -march=native

FFTW

FFTW is a C subroutine library for computing the discrete Fourier transform (DFT) in one or more dimensions. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Float + SSE - Size: 1D FFT Size 512Ryzen 3900X2 x Xeon Platinum 82802 x EPYC 77422 x EPYC 750212K24K36K48K60KSE +/- 388.57, N = 3SE +/- 886.61, N = 3SE +/- 336.45, N = 3SE +/- 56.26, N = 3568065598338922377221. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Float + SSE - Size: 1D FFT Size 10242 x Xeon Platinum 8280Ryzen 3900X2 x EPYC 77422 x EPYC 750214K28K42K56K70KSE +/- 1102.91, N = 3SE +/- 855.91, N = 3SE +/- 58.50, N = 3SE +/- 362.20, N = 3647325838743869430781. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

Renaissance

Renaissance is a suite of benchmarks designed to test the Java JVM from Apache Spark to a Twitter-like service to Scala and other features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.9.0Test: Scala DottyRyzen 3900X2 x Xeon Platinum 82802 x EPYC 77422 x EPYC 750216003200480064008000SE +/- 24.44, N = 8SE +/- 55.49, N = 8SE +/- 65.40, N = 40SE +/- 43.69, N = 404891693072767326

Minion

Minion is an open-source constraint solver that is designed to be very scalable. This test profile uses Minion's integrated benchmarking problems to solve. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterMinion 1.8Benchmark: SolitaireRyzen 3900X2 x EPYC 75022 x Xeon Platinum 828020406080100SE +/- 0.30, N = 3SE +/- 0.17, N = 3SE +/- 0.48, N = 356.5582.0784.341. (CXX) g++ options: -std=gnu++11 -O3 -fomit-frame-pointer -rdynamic

PHP Micro Benchmarks

Various small PHP micro-benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterPHP Micro BenchmarksTest: Zend micro_benchRyzen 3900X2 x Xeon Platinum 82802 x EPYC 75020.59851.1971.79552.3942.9925SE +/- 0.03, N = 6SE +/- 0.01, N = 3SE +/- 0.04, N = 31.802.062.66

Tinymembench

This benchmark tests the system memory (RAM) performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterTinymembench 2018-05-28Standard Memset2 x EPYC 77422 x EPYC 7502Ryzen 3900X2 x Xeon Platinum 82803K6K9K12K15KSE +/- 17.98, N = 3SE +/- 16.26, N = 3SE +/- 28.57, N = 3SE +/- 67.15, N = 3160311524013889109201. (CC) gcc options: -O2 -lm

FFTW

FFTW is a C subroutine library for computing the discrete Fourier transform (DFT) in one or more dimensions. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Float + SSE - Size: 2D FFT Size 64Ryzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 828011K22K33K44K55KSE +/- 450.07, N = 3SE +/- 22.10, N = 3SE +/- 327.91, N = 3SE +/- 124.46, N = 3491463701135897336481. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

RAMspeed SMP

This benchmark tests the system memory (RAM) performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterRAMspeed SMP 3.5.0Type: Triad - Benchmark: Integer2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X9K18K27K36K45K430234011031864295971. (CC) gcc options: -O3 -march=native

Java SciMark

This test runs the Java version of SciMark 2.0, which is a benchmark for scientific and numerical computing developed by programmers at the National Institute of Standards and Technology. This benchmark is made up of Fast Foruier Transform, Jacobi Successive Over-relaxation, Monte Carlo, Sparse Matrix Multiply, and dense LU matrix factorization benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterJava SciMark 2.0Computational Test: Dense LU Matrix FactorizationRyzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 828015003000450060007500SE +/- 34.41, N = 4SE +/- 9.32, N = 4SE +/- 9.37, N = 4SE +/- 15.06, N = 46944536452704793

HPC Challenge

HPC Challenge (HPCC) is a cluster-focused benchmark consisting of the HPL Linpack TPP benchmark, DGEMM, STREAM, PTRANS, RandomAccess, FFT, and communication bandwidth and latency. This HPC Challenge test profile attempts to ship with standard yet versatile configuration/input files though they can be modified. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGB/s, More Is BetterHPC Challenge 1.5.0Test / Class: EP-STREAM Triad2 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X0.67561.35122.02682.70243.378SE +/- 0.00024, N = 3SE +/- 0.00633, N = 3SE +/- 0.00135, N = 33.002692.802072.075941. (CC) gcc options: -lblas -lm -pthread -lmpi -fomit-frame-pointer -funroll-loops2. 2 x Xeon Platinum 8280: ATLAS + Open MPI 3.1.33. 2 x EPYC 7502: ATLAS + Open MPI 3.1.34. Ryzen 3900X: ATLAS + Open MPI 3.1.4

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.3.1Test / Class: FT.B2 x EPYC 75022 x Xeon Platinum 82808001600240032004000SE +/- 3.88, N = 3SE +/- 4.18, N = 33660.372556.201. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi

FFTW

FFTW is a C subroutine library for computing the discrete Fourier transform (DFT) in one or more dimensions. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Float + SSE - Size: 1D FFT Size 128Ryzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 82806K12K18K24K30KSE +/- 315.54, N = 12SE +/- 232.04, N = 7SE +/- 71.04, N = 3SE +/- 316.85, N = 3290162105920960203431. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Float + SSE - Size: 1D FFT Size 20482 x Xeon Platinum 8280Ryzen 3900X2 x EPYC 77422 x EPYC 750214K28K42K56K70KSE +/- 650.59, N = 3SE +/- 462.69, N = 3SE +/- 332.33, N = 3SE +/- 549.65, N = 3639086060645337449151. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

OSBench

OSBench is a collection of micro-benchmarks for measuring operating system primitives like time to create threads/processes, launching programs, creating files, and memory allocation. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNs Per Event, Fewer Is BetterOSBenchTest: Memory AllocationsRyzen 3900X2 x Xeon Platinum 82802 x EPYC 750220406080100SE +/- 0.09, N = 3SE +/- 0.15, N = 3SE +/- 0.23, N = 368.0187.5796.591. (CC) gcc options: -lm

FFTW

FFTW is a C subroutine library for computing the discrete Fourier transform (DFT) in one or more dimensions. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Stock - Size: 2D FFT Size 512Ryzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 82802K4K6K8K10KSE +/- 34.09, N = 3SE +/- 9.03, N = 3SE +/- 3.11, N = 3SE +/- 31.31, N = 390126717663363551. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Stock - Size: 2D FFT Size 4096Ryzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 828014002800420056007000SE +/- 18.60, N = 3SE +/- 10.42, N = 3SE +/- 2.50, N = 3SE +/- 31.33, N = 365565214511646401. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

PHPBench

PHPBench is a benchmark suite for PHP. It performs a large number of simple tests in order to bench various aspects of the PHP interpreter. PHPBench can be used to compare hardware, operating systems, PHP versions, PHP accelerators and caches, compiler options, etc. The number of iterations used is 1,000,000. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgScore, More Is BetterPHPBench 0.8.1PHP Benchmark SuiteRyzen 3900X2 x Xeon Platinum 82802 x EPYC 7502150K300K450K600K750KSE +/- 11875.52, N = 3SE +/- 240.33, N = 3SE +/- 1963.83, N = 3698254646738494910

Apache Siege

This is a test of the Apache web server performance being facilitated by the Siege web serverb enchmark program. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTransactions Per Second, More Is BetterApache Siege 2.4.29Concurrent Users: 2502 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X7K14K21K28K35KSE +/- 126.68, N = 3SE +/- 98.88, N = 3SE +/- 203.16, N = 33460830037246261. (CC) gcc options: -O2 -lpthread -ldl -lssl -lcrypto

Minion

Minion is an open-source constraint solver that is designed to be very scalable. This test profile uses Minion's integrated benchmarking problems to solve. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterMinion 1.8Benchmark: QuasigroupRyzen 3900X2 x Xeon Platinum 82802 x EPYC 7502306090120150SE +/- 0.32, N = 3SE +/- 0.23, N = 3SE +/- 0.20, N = 398.29111.21137.891. (CXX) g++ options: -std=gnu++11 -O3 -fomit-frame-pointer -rdynamic

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.3.1Test / Class: FT.A2 x EPYC 75022 x Xeon Platinum 82807001400210028003500SE +/- 5.92, N = 3SE +/- 2.56, N = 33332.602378.801. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi

FFTW

FFTW is a C subroutine library for computing the discrete Fourier transform (DFT) in one or more dimensions. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Float + SSE - Size: 1D FFT Size 40962 x Xeon Platinum 8280Ryzen 3900X2 x EPYC 77422 x EPYC 750212K24K36K48K60KSE +/- 504.24, N = 3SE +/- 914.81, N = 3SE +/- 395.93, N = 3SE +/- 553.18, N = 3582225821743038416181. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

Tinymembench

This benchmark tests the system memory (RAM) performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterTinymembench 2018-05-28Standard MemcpyRyzen 3900X2 x EPYC 75022 x EPYC 77422 x Xeon Platinum 82802K4K6K8K10KSE +/- 5.75, N = 3SE +/- 4.52, N = 3SE +/- 4.70, N = 3SE +/- 51.26, N = 396309148913169331. (CC) gcc options: -O2 -lm

libjpeg-turbo tjbench

tjbench is a JPEG decompression/compression benchmark part of libjpeg-turbo. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMegapixels/sec, More Is Betterlibjpeg-turbo tjbench 2.0.2Test: Decompression ThroughputRyzen 3900X2 x Xeon Platinum 82802 x EPYC 750250100150200250SE +/- 0.42, N = 3SE +/- 0.48, N = 3SE +/- 0.07, N = 32331881691. (CC) gcc options: -O3 -rdynamic

SVT-VP9

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-VP9 CPU-based multi-threaded video encoder for the VP9 video format with a sample 1080p YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-VP9 2019-02-171080p 8-bit YUV To VP9 Video Encode2 x Xeon Platinum 82802 x EPYC 750260120180240300SE +/- 3.53, N = 4SE +/- 1.34, N = 3271.63197.261. (CC) gcc options: -fPIE -fPIC -O2 -flto -fvisibility=hidden -mavx -pie -rdynamic -lpthread -lrt -lm

Numpy Benchmark

This is a test to obtain the general Numpy performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNanoseconds, Fewer Is BetterNumpy Benchmark2 x Xeon Platinum 82802 x EPYC 75021.5M3M4.5M6M7.5M52622167216577

Botan

Botan is a cross-platform open-source C++ crypto library that supports most all publicly known cryptographic algorithms. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.8.0Test: CAST-256 - DecryptRyzen 3900X2 x Xeon Platinum 82802 x EPYC 75024080120160200SE +/- 0.59, N = 3SE +/- 0.01, N = 3SE +/- 0.03, N = 31641371201. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt

OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.8.0Test: CAST-256 - EncryptRyzen 3900X2 x Xeon Platinum 82802 x EPYC 75024080120160200SE +/- 0.55, N = 3SE +/- 0.57, N = 3SE +/- 0.07, N = 31641371201. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt

OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.8.0Test: KASUMI - EncryptRyzen 3900X2 x Xeon Platinum 82802 x EPYC 750220406080100SE +/- 0.36, N = 3SE +/- 0.06, N = 3SE +/- 0.04, N = 3106.2989.6377.841. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt

FFTW

FFTW is a C subroutine library for computing the discrete Fourier transform (DFT) in one or more dimensions. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Stock - Size: 1D FFT Size 1024Ryzen 3900X2 x Xeon Platinum 82802 x EPYC 77422 x EPYC 75022K4K6K8K10KSE +/- 10.71, N = 3SE +/- 81.37, N = 3SE +/- 22.72, N = 3SE +/- 2.19, N = 397267555720771291. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

LAME MP3 Encoding

LAME is an MP3 encoder licensed under the LGPL. This test measures the time required to encode a WAV file to MP3 format. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterLAME MP3 Encoding 3.100WAV To MP3Ryzen 3900X2 x Xeon Platinum 82802 x EPYC 7502816243240SE +/- 0.24, N = 3SE +/- 0.01, N = 3SE +/- 0.01, N = 324.3829.9733.171. (CC) gcc options: -lncurses -lm

FFTW

FFTW is a C subroutine library for computing the discrete Fourier transform (DFT) in one or more dimensions. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Stock - Size: 2D FFT Size 256Ryzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 82802K4K6K8K10KSE +/- 16.61, N = 3SE +/- 18.60, N = 3SE +/- 13.59, N = 3SE +/- 7.72, N = 386836744661563861. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

Apache Siege

This is a test of the Apache web server performance being facilitated by the Siege web serverb enchmark program. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTransactions Per Second, More Is BetterApache Siege 2.4.29Concurrent Users: 1002 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X7K14K21K28K35KSE +/- 119.32, N = 3SE +/- 186.18, N = 3SE +/- 69.47, N = 33324226655244501. (CC) gcc options: -O2 -lpthread -ldl -lssl -lcrypto

Mencoder

This test uses mplayer's mencoder utility and the libavcodec family for testing the system's audio/video encoding performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterMencoder 1.3.0AVI To LAVCRyzen 3900X2 x Xeon Platinum 82802 x EPYC 7502510152025SE +/- 0.25, N = 3SE +/- 0.03, N = 3SE +/- 0.01, N = 314.9217.6920.20-lfreetype -lXext -lX11 -lXxf86vm -lSDL -lGL-lmad -lvorbisenc -lvorbis -logg1. (CC) gcc options: -ffast-math -fpie -pie -lncurses -lrt -lpng -lz -ljpeg -lasound -ldl -lpthread -lbz2 -rdynamic -lm

Botan

Botan is a cross-platform open-source C++ crypto library that supports most all publicly known cryptographic algorithms. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.8.0Test: KASUMI - DecryptRyzen 3900X2 x Xeon Platinum 82802 x EPYC 750220406080100SE +/- 0.35, N = 3SE +/- 0.02, N = 3SE +/- 0.03, N = 3101.8287.0375.241. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt

PHP Micro Benchmarks

Various small PHP micro-benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterPHP Micro BenchmarksTest: Zend benchRyzen 3900X2 x Xeon Platinum 82802 x EPYC 75020.12150.2430.36450.4860.6075SE +/- 0.00, N = 3SE +/- 0.00, N = 3SE +/- 0.00, N = 30.400.450.54

FFTW

FFTW is a C subroutine library for computing the discrete Fourier transform (DFT) in one or more dimensions. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Float + SSE - Size: 1D FFT Size 32Ryzen 3900X2 x Xeon Platinum 82802 x EPYC 77422 x EPYC 75023K6K9K12K15KSE +/- 16.23, N = 3SE +/- 31.05, N = 3SE +/- 31.83, N = 3SE +/- 10.11, N = 3159551431711985118481. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

Apache Siege

This is a test of the Apache web server performance being facilitated by the Siege web serverb enchmark program. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTransactions Per Second, More Is BetterApache Siege 2.4.29Concurrent Users: 2002 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X7K14K21K28K35KSE +/- 38.45, N = 3SE +/- 61.21, N = 3SE +/- 262.82, N = 33343228569249061. (CC) gcc options: -O2 -lpthread -ldl -lssl -lcrypto

Java SciMark

This test runs the Java version of SciMark 2.0, which is a benchmark for scientific and numerical computing developed by programmers at the National Institute of Standards and Technology. This benchmark is made up of Fast Foruier Transform, Jacobi Successive Over-relaxation, Monte Carlo, Sparse Matrix Multiply, and dense LU matrix factorization benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterJava SciMark 2.0Computational Test: CompositeRyzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 82806001200180024003000SE +/- 28.76, N = 16SE +/- 1.16, N = 4SE +/- 2.65, N = 4SE +/- 4.49, N = 42924242323892182

FLAC Audio Encoding

This test times how long it takes to encode a sample WAV file to FLAC format five times. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterFLAC Audio Encoding 1.3.2WAV To FLACRyzen 3900X2 x Xeon Platinum 82802 x EPYC 75023691215SE +/- 0.01, N = 5SE +/- 0.01, N = 5SE +/- 0.02, N = 57.539.5610.09-logg-logg1. (CXX) g++ options: -O2 -fvisibility=hidden -lm

FFTW

FFTW is a C subroutine library for computing the discrete Fourier transform (DFT) in one or more dimensions. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Stock - Size: 2D FFT Size 64Ryzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 82802K4K6K8K10KSE +/- 30.93, N = 3SE +/- 5.63, N = 3SE +/- 2.64, N = 3SE +/- 55.12, N = 396017400730071811. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

Timed GCC Compilation

This test times how long it takes to build the GNU Compiler Collection (GCC). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed GCC Compilation 8.2Time To CompileRyzen 3900X2 x EPYC 75022 x Xeon Platinum 82802004006008001000SE +/- 0.57, N = 3SE +/- 0.64, N = 3SE +/- 0.37, N = 3668746892

Apache Siege

This is a test of the Apache web server performance being facilitated by the Siege web serverb enchmark program. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTransactions Per Second, More Is BetterApache Siege 2.4.29Concurrent Users: 50Ryzen 3900X2 x Xeon Platinum 82802 x EPYC 75026K12K18K24K30KSE +/- 397.01, N = 6SE +/- 426.87, N = 3SE +/- 253.10, N = 32888927915216921. (CC) gcc options: -O2 -lpthread -ldl -lssl -lcrypto

Sockperf

This is a network socket API performance benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgusec, Fewer Is BetterSockperf 3.4Test: Latency Ping Pong2 x Xeon Platinum 8280Ryzen 3900X2 x EPYC 75022 x EPYC 77421.19252.3853.57754.775.9625SE +/- 0.03, N = 5SE +/- 0.02, N = 5SE +/- 0.02, N = 5SE +/- 0.02, N = 53.984.844.855.301. (CXX) g++ options: --param -O3 -rdynamic -ldl -lpthread

FFTW

FFTW is a C subroutine library for computing the discrete Fourier transform (DFT) in one or more dimensions. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Stock - Size: 1D FFT Size 128Ryzen 3900X2 x Xeon Platinum 82802 x EPYC 77422 x EPYC 75022K4K6K8K10KSE +/- 108.32, N = 3SE +/- 99.21, N = 3SE +/- 55.49, N = 3SE +/- 2.34, N = 389967024689967661. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

TSCP

This is a performance test of TSCP, Tom Kerrigan's Simple Chess Program, which has a built-in performance benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgNodes Per Second, More Is BetterTSCP 1.81AI Chess PerformanceRyzen 3900X2 x Xeon Platinum 82802 x EPYC 7502300K600K900K1200K1500KSE +/- 2434.15, N = 5SE +/- 1283.95, N = 5SE +/- 778.60, N = 51376962125748910360781. (CC) gcc options: -O3 -march=native

x265

This is a simple test of the x265 encoder run on the CPU with a sample 1080p video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betterx265 3.0H.265 1080p Video Encoding2 x EPYC 75022 x Xeon Platinum 82801020304050SE +/- 0.11, N = 3SE +/- 0.10, N = 343.4632.731. (CXX) g++ options: -O3 -rdynamic -lpthread -lrt -ldl -lnuma

Botan

Botan is a cross-platform open-source C++ crypto library that supports most all publicly known cryptographic algorithms. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.8.0Test: Twofish - EncryptRyzen 3900X2 x Xeon Platinum 82802 x EPYC 750290180270360450SE +/- 0.63, N = 3SE +/- 0.20, N = 3SE +/- 0.14, N = 33963472991. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt

FFTW

FFTW is a C subroutine library for computing the discrete Fourier transform (DFT) in one or more dimensions. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Stock - Size: 2D FFT Size 128Ryzen 3900X2 x Xeon Platinum 82802 x EPYC 77422 x EPYC 75022K4K6K8K10KSE +/- 26.29, N = 3SE +/- 108.79, N = 3SE +/- 15.81, N = 3SE +/- 40.50, N = 389887470694867881. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 9.12-MR1Java Test: Tradesoap2 x EPYC 75022 x Xeon Platinum 828016003200480064008000SE +/- 41.94, N = 20SE +/- 75.42, N = 2056017404

FFTW

FFTW is a C subroutine library for computing the discrete Fourier transform (DFT) in one or more dimensions. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Stock - Size: 1D FFT Size 512Ryzen 3900X2 x Xeon Platinum 82802 x EPYC 77422 x EPYC 75022K4K6K8K10KSE +/- 9.39, N = 3SE +/- 9.12, N = 3SE +/- 15.02, N = 3SE +/- 10.19, N = 392887793712470521. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Stock - Size: 1D FFT Size 2048Ryzen 3900X2 x Xeon Platinum 82802 x EPYC 77422 x EPYC 75022K4K6K8K10KSE +/- 100.27, N = 3SE +/- 56.60, N = 3SE +/- 0.67, N = 3SE +/- 15.82, N = 389987392700668581. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

GNU GMP GMPbench

GMPbench is a test of the GMP 6.1.2 math library. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGMPbench Score, More Is BetterGNU GMP GMPbench 6.1.2Total TimeRyzen 3900X2 x Xeon Platinum 82802 x EPYC 77422 x EPYC 75021300260039005200650059225098458945191. (CC) gcc options: -O3 -fomit-frame-pointer -lm

SVT-HEVC

This is a test of the Intel Open Visual Cloud Scalable Video Technology SVT-HEVC CPU-based multi-threaded video encoder for the HEVC / H.265 video format with a sample 1080p YUV video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterSVT-HEVC 2019-02-031080p 8-bit YUV To HEVC Video Encode2 x EPYC 75022 x Xeon Platinum 828070140210280350SE +/- 2.17, N = 3SE +/- 3.63, N = 4340.94260.311. (CC) gcc options: -fPIE -fPIC -O2 -flto -fvisibility=hidden -march=native -pie -rdynamic -lpthread -lrt

Botan

Botan is a cross-platform open-source C++ crypto library that supports most all publicly known cryptographic algorithms. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.8.0Test: Twofish - DecryptRyzen 3900X2 x Xeon Platinum 82802 x EPYC 750280160240320400SE +/- 0.64, N = 3SE +/- 0.19, N = 3SE +/- 0.10, N = 33913512991. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt

Apache Siege

This is a test of the Apache web server performance being facilitated by the Siege web serverb enchmark program. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTransactions Per Second, More Is BetterApache Siege 2.4.29Concurrent Users: 102 x Xeon Platinum 8280Ryzen 3900X2 x EPYC 75026K12K18K24K30KSE +/- 326.78, N = 15SE +/- 238.51, N = 3SE +/- 86.97, N = 32601220400199081. (CC) gcc options: -O2 -lpthread -ldl -lssl -lcrypto

Botan

Botan is a cross-platform open-source C++ crypto library that supports most all publicly known cryptographic algorithms. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.8.0Test: Blowfish - DecryptRyzen 3900X2 x Xeon Platinum 82802 x EPYC 750260120180240300SE +/- 0.05, N = 3SE +/- 0.07, N = 3SE +/- 0.07, N = 32972752281. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt

FFTW

FFTW is a C subroutine library for computing the discrete Fourier transform (DFT) in one or more dimensions. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Stock - Size: 2D FFT Size 32Ryzen 3900X2 x EPYC 77422 x Xeon Platinum 82802 x EPYC 75022K4K6K8K10KSE +/- 11.50, N = 3SE +/- 16.76, N = 3SE +/- 69.98, N = 3SE +/- 13.25, N = 3114668982889888241. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Stock - Size: 1D FFT Size 4096Ryzen 3900X2 x Xeon Platinum 82802 x EPYC 77422 x EPYC 75022K4K6K8K10KSE +/- 85.21, N = 3SE +/- 20.22, N = 3SE +/- 13.44, N = 3SE +/- 9.28, N = 387137274681467091. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Stock - Size: 2D FFT Size 1024Ryzen 3900X2 x Xeon Platinum 82802 x EPYC 77422 x EPYC 75022K4K6K8K10KSE +/- 21.10, N = 3SE +/- 33.68, N = 3SE +/- 6.26, N = 3SE +/- 60.11, N = 379096340614660991. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

Botan

Botan is a cross-platform open-source C++ crypto library that supports most all publicly known cryptographic algorithms. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMiB/s, More Is BetterBotan 2.8.0Test: Blowfish - EncryptRyzen 3900X2 x Xeon Platinum 82802 x EPYC 750270140210280350SE +/- 0.00, N = 3SE +/- 0.03, N = 3SE +/- 0.16, N = 32992822311. (CXX) g++ options: -fstack-protector -m64 -pthread -lbotan-2 -ldl -lrt

FFTW

FFTW is a C subroutine library for computing the discrete Fourier transform (DFT) in one or more dimensions. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Stock - Size: 2D FFT Size 2048Ryzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 828014002800420056007000SE +/- 4.41, N = 3SE +/- 9.32, N = 3SE +/- 71.82, N = 5SE +/- 75.14, N = 465735993581550901. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

lzbench

lzbench is an in-memory benchmark of various compressors. The file used for compression is a Linux kernel source tree tarball. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is Betterlzbench 2017-08-08Test: XZ 0 - Process: Compression2 x Xeon Platinum 82802 x EPYC 77422 x EPYC 75028162432403628281. (CXX) g++ options: -lrt -static -lpthread -fomit-frame-pointer -fstrict-aliasing -ffast-math -O3

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 4.0.64Scene: Bedroom2 x Xeon Platinum 82802 x EPYC 7502246810SE +/- 0.01, N = 3SE +/- 0.01, N = 36.144.78

OpenBenchmarking.orgM samples/s, More Is BetterIndigoBench 4.0.64Scene: Supercar2 x Xeon Platinum 82802 x EPYC 75023691215SE +/- 0.03, N = 3SE +/- 0.01, N = 39.887.72

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 9.12-MR1Java Test: H22 x EPYC 75022 x EPYC 77422 x Xeon Platinum 82802K4K6K8K10KSE +/- 92.11, N = 4SE +/- 35.70, N = 4SE +/- 38.38, N = 4691182268773

FFTW

FFTW is a C subroutine library for computing the discrete Fourier transform (DFT) in one or more dimensions. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Float + SSE - Size: 2D FFT Size 4096Ryzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 82804K8K12K16K20KSE +/- 251.60, N = 3SE +/- 63.98, N = 3SE +/- 214.67, N = 3SE +/- 82.58, N = 3204181786517004160991. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Stock - Size: 1D FFT Size 256Ryzen 3900X2 x EPYC 77422 x Xeon Platinum 82802 x EPYC 75022K4K6K8K10KSE +/- 88.12, N = 3SE +/- 14.81, N = 3SE +/- 56.78, N = 3SE +/- 1.71, N = 389817217720071271. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Float + SSE - Size: 2D FFT Size 20482 x EPYC 77422 x EPYC 7502Ryzen 3900X2 x Xeon Platinum 82806K12K18K24K30KSE +/- 51.68, N = 3SE +/- 195.72, N = 3SE +/- 99.90, N = 3SE +/- 101.12, N = 3266612610921767211981. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

Cython benchmark

Stress benchmark tests to measure time consumed by cython code Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterCython benchmark 0.272 x Xeon Platinum 82802 x EPYC 7502714212835SE +/- 0.05, N = 3SE +/- 0.07, N = 322.9728.68

x264

This is a simple test of the x264 encoder run on the CPU (OpenCL support disabled) with a sample video file. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is Betterx264 2018-09-25H.264 Video Encoding2 x EPYC 75022 x Xeon Platinum 8280306090120150SE +/- 0.99, N = 3SE +/- 1.20, N = 15148.94120.041. (CC) gcc options: -ldl -lavformat -lavcodec -lavutil -lswscale -m64 -lm -lpthread -O3 -ffast-math -std=gnu99 -fPIC -fomit-frame-pointer -fno-tree-vectorize

Java SciMark

This test runs the Java version of SciMark 2.0, which is a benchmark for scientific and numerical computing developed by programmers at the National Institute of Standards and Technology. This benchmark is made up of Fast Foruier Transform, Jacobi Successive Over-relaxation, Monte Carlo, Sparse Matrix Multiply, and dense LU matrix factorization benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterJava SciMark 2.0Computational Test: Sparse Matrix MultiplyRyzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 82806001200180024003000SE +/- 15.03, N = 4SE +/- 0.56, N = 4SE +/- 1.10, N = 4SE +/- 0.56, N = 42674224622112156

OSBench

OSBench is a collection of micro-benchmarks for measuring operating system primitives like time to create threads/processes, launching programs, creating files, and memory allocation. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgus Per Event, Fewer Is BetterOSBenchTest: Create Files2 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X48121620SE +/- 0.04, N = 3SE +/- 0.03, N = 3SE +/- 0.05, N = 313.7715.2317.061. (CC) gcc options: -lm

Stress-NG

Stress-NG is a Linux stress tool developed by Colin King of Canonical. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.07.26Test: Memory Copying2 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X2K4K6K8K10KSE +/- 108.22, N = 3SE +/- 100.69, N = 3SE +/- 80.33, N = 311263982391531. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc

Darktable

Darktable is an open-source photography / workflow application this will use any system-installed Darktable program or on Windows will automatically download the pre-built binary from the project. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.6.0Test: Boat - Acceleration: CPU-only2 x EPYC 75022 x Xeon Platinum 82800.91131.82262.73393.64524.5565SE +/- 0.05, N = 3SE +/- 0.04, N = 33.304.05

lzbench

lzbench is an in-memory benchmark of various compressors. The file used for compression is a Linux kernel source tree tarball. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is Betterlzbench 2017-08-08Test: Libdeflate 1 - Process: Decompression2 x Xeon Platinum 82802 x EPYC 77422 x EPYC 75022004006008001000SE +/- 8.00, N = 310168678411. (CXX) g++ options: -lrt -static -lpthread -fomit-frame-pointer -fstrict-aliasing -ffast-math -O3

Redis

Redis is an open-source data structure server. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRequests Per Second, More Is BetterRedis 4.0.8Test: SADDRyzen 3900X2 x Xeon Platinum 82802 x EPYC 7502500K1000K1500K2000K2500KSE +/- 4715.75, N = 3SE +/- 20580.51, N = 15SE +/- 26432.00, N = 152169218212272718005461. (CC) gcc options: -ggdb -rdynamic -lm -ldl -pthread

XZ Compression

This test measures the time needed to compress a sample file (an Ubuntu file-system image) using XZ compression. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterXZ Compression 5.2.4Compressing ubuntu-16.04.3-server-i386.img, Compression Level 92 x Xeon Platinum 8280Ryzen 3900X2 x EPYC 7502612182430SE +/- 0.19, N = 3SE +/- 0.04, N = 3SE +/- 0.13, N = 322.5025.3226.971. (CC) gcc options: -pthread -fvisibility=hidden -O2

SPECjbb 2015

This is a benchmark of SPECjbb 2015. For this test profile to work, you must have a valid license/copy of the SPECjbb 2015 ISO (SPECjbb2015-1.02.iso) in your Phoronix Test Suite download cache. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgjOPS, More Is BetterSPECjbb 2015SPECjbb2015-Composite critical-jOPS2 x EPYC 75022 x Xeon Platinum 828013K26K39K52K65K6021850770

DaCapo Benchmark

This test runs the DaCapo Benchmarks written in Java and intended to test system/CPU performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgmsec, Fewer Is BetterDaCapo Benchmark 9.12-MR1Java Test: Jython2 x Xeon Platinum 82802 x EPYC 77422 x EPYC 750212002400360048006000SE +/- 36.59, N = 4SE +/- 26.33, N = 4SE +/- 67.79, N = 4462952645466

lzbench

lzbench is an in-memory benchmark of various compressors. The file used for compression is a Linux kernel source tree tarball. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is Betterlzbench 2017-08-08Test: Brotli 0 - Process: Decompression2 x Xeon Platinum 82802 x EPYC 77422 x EPYC 7502110220330440550SE +/- 0.88, N = 3SE +/- 0.33, N = 35194464411. (CXX) g++ options: -lrt -static -lpthread -fomit-frame-pointer -fstrict-aliasing -ffast-math -O3

FFTW

FFTW is a C subroutine library for computing the discrete Fourier transform (DFT) in one or more dimensions. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Stock - Size: 1D FFT Size 64Ryzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 82802K4K6K8K10KSE +/- 61.23, N = 3SE +/- 3.05, N = 3SE +/- 0.58, N = 3SE +/- 5.52, N = 387018203808274041. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

Java SciMark

This test runs the Java version of SciMark 2.0, which is a benchmark for scientific and numerical computing developed by programmers at the National Institute of Standards and Technology. This benchmark is made up of Fast Foruier Transform, Jacobi Successive Over-relaxation, Monte Carlo, Sparse Matrix Multiply, and dense LU matrix factorization benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterJava SciMark 2.0Computational Test: Jacobi Successive Over-RelaxationRyzen 3900X2 x Xeon Platinum 82802 x EPYC 77422 x EPYC 7502400800120016002000SE +/- 10.90, N = 4SE +/- 6.63, N = 4SE +/- 0.15, N = 4SE +/- 0.00, N = 41716158515291506

SPECjbb 2015

This is a benchmark of SPECjbb 2015. For this test profile to work, you must have a valid license/copy of the SPECjbb 2015 ISO (SPECjbb2015-1.02.iso) in your Phoronix Test Suite download cache. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgjOPS, More Is BetterSPECjbb 2015SPECjbb2015-Composite max-jOPS2 x EPYC 75022 x Xeon Platinum 828020K40K60K80K100K10664593995

Darktable

Darktable is an open-source photography / workflow application this will use any system-installed Darktable program or on Windows will automatically download the pre-built binary from the project. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.6.0Test: Masskrug - Acceleration: CPU-only2 x EPYC 75022 x Xeon Platinum 82801.07332.14663.21994.29325.3665SE +/- 0.02, N = 3SE +/- 0.02, N = 34.244.77

lzbench

lzbench is an in-memory benchmark of various compressors. The file used for compression is a Linux kernel source tree tarball. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is Betterlzbench 2017-08-08Test: Libdeflate 1 - Process: Compression2 x Xeon Platinum 82802 x EPYC 77422 x EPYC 750250100150200250SE +/- 0.33, N = 3SE +/- 1.00, N = 32131911901. (CXX) g++ options: -lrt -static -lpthread -fomit-frame-pointer -fstrict-aliasing -ffast-math -O3

Blender

Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing is supported. This system/blender test profile makes use of the system-supplied Blender. Use pts/blender if wishing to stick to a fixed version of Blender. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.79Blend File: Pabellon Barcelona - Compute: CPU-Only2 x EPYC 75022 x Xeon Platinum 8280306090120150SE +/- 0.37, N = 3SE +/- 0.09, N = 3100.13112.00

ctx_clock

Ctx_clock is a simple test program to measure the context switch time in clock cycles. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgClocks, Fewer Is Betterctx_clockContext Switch Time2 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X306090120150136150152

Ogg Encoding

This test times how long it takes to encode a sample WAV file to Ogg format using vorbis-tools, libvorbis, and libogg. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOgg Encoding 1.3.3WAV To Ogg2 x Xeon Platinum 82802 x EPYC 7502246810SE +/- 0.00, N = 3SE +/- 0.02, N = 36.126.811. (CC) gcc options: -O2 -ffast-math -fsigned-char -logg

Blender

Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing is supported. This system/blender test profile makes use of the system-supplied Blender. Use pts/blender if wishing to stick to a fixed version of Blender. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.79Blend File: Classroom - Compute: CPU-Only2 x EPYC 75022 x Xeon Platinum 8280306090120150SE +/- 0.99, N = 3SE +/- 0.34, N = 3122.64135.02

Timed PHP Compilation

This test times how long it takes to build PHP 5 with the Zend engine. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed PHP Compilation 7.1.9Time To CompileRyzen 3900X2 x Xeon Platinum 82802 x EPYC 75021122334455SE +/- 0.10, N = 3SE +/- 0.17, N = 3SE +/- 0.02, N = 344.6848.5949.141. (CC) gcc options: -O2 -pedantic -ldl -lz -lm

Blender

Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing is supported. This system/blender test profile makes use of the system-supplied Blender. Use pts/blender if wishing to stick to a fixed version of Blender. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.79Blend File: Fishy Cat - Compute: CPU-Only2 x EPYC 75022 x Xeon Platinum 828020406080100SE +/- 0.10, N = 3SE +/- 0.04, N = 370.1676.82

lzbench

lzbench is an in-memory benchmark of various compressors. The file used for compression is a Linux kernel source tree tarball. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is Betterlzbench 2017-08-08Test: XZ 0 - Process: Decompression2 x Xeon Platinum 82802 x EPYC 77422 x EPYC 7502204060801009588871. (CXX) g++ options: -lrt -static -lpthread -fomit-frame-pointer -fstrict-aliasing -ffast-math -O3

GIMP

GIMP is an open-source image manipulaton program. This test profile will use the system-provided GIMP program otherwise on Windows relys upon a pre-packaged Windows binary from upstream GIMP.org. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterGIMP 2.10.8Test: rotate2 x EPYC 75022 x Xeon Platinum 828048121620SE +/- 0.02, N = 3SE +/- 0.03, N = 316.7518.25

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.3.1Test / Class: EP.C2 x Xeon Platinum 82802 x EPYC 75025001000150020002500SE +/- 22.95, N = 3SE +/- 1.59, N = 32486.052283.031. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi

Timed ImageMagick Compilation

This test times how long it takes to build ImageMagick. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed ImageMagick Compilation 6.9.0Time To Compile2 x Xeon Platinum 8280Ryzen 3900X2 x EPYC 7502510152025SE +/- 0.21, N = 3SE +/- 0.07, N = 3SE +/- 0.36, N = 319.6719.9521.39

GIMP

GIMP is an open-source image manipulaton program. This test profile will use the system-provided GIMP program otherwise on Windows relys upon a pre-packaged Windows binary from upstream GIMP.org. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterGIMP 2.10.8Test: unsharp-mask2 x Xeon Platinum 82802 x EPYC 7502714212835SE +/- 0.38, N = 3SE +/- 0.11, N = 328.1830.60

Blender

Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing is supported. This system/blender test profile makes use of the system-supplied Blender. Use pts/blender if wishing to stick to a fixed version of Blender. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.79Blend File: BMW27 - Compute: CPU-Only2 x EPYC 75022 x Xeon Platinum 82801224364860SE +/- 0.21, N = 3SE +/- 0.05, N = 347.3251.15

GIMP

GIMP is an open-source image manipulaton program. This test profile will use the system-provided GIMP program otherwise on Windows relys upon a pre-packaged Windows binary from upstream GIMP.org. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterGIMP 2.10.8Test: auto-levels2 x Xeon Platinum 82802 x EPYC 7502612182430SE +/- 0.08, N = 3SE +/- 0.29, N = 622.7924.62

OpenBenchmarking.orgSeconds, Fewer Is BetterGIMP 2.10.8Test: resize2 x EPYC 75022 x Xeon Platinum 82803691215SE +/- 0.03, N = 3SE +/- 0.03, N = 311.6712.48

Tesseract OCR

Tesseract-OCR is the open-source optical character recognition (OCR) engine for the conversion of text within images to raw text output. This test profile relies upon a system-supplied Tesseract installation. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTesseract OCR 4.0.0Time To OCR 7 Images2 x Xeon Platinum 82802 x EPYC 7502816243240SE +/- 0.03, N = 3SE +/- 0.20, N = 332.7934.90

lzbench

lzbench is an in-memory benchmark of various compressors. The file used for compression is a Linux kernel source tree tarball. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is Betterlzbench 2017-08-08Test: Zstd 1 - Process: Compression2 x Xeon Platinum 82802 x EPYC 77422 x EPYC 7502801602403204003753603541. (CXX) g++ options: -lrt -static -lpthread -fomit-frame-pointer -fstrict-aliasing -ffast-math -O3

Chaos Group V-RAY

This is a test of Chaos Group's V-RAY benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgKsamples, More Is BetterChaos Group V-RAY 4.10.03Mode: CPU2 x Xeon Platinum 82802 x EPYC 750214K28K42K56K70KSE +/- 89.03, N = 3SE +/- 755.94, N = 36547662297

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.3.1Test / Class: LU.C2 x EPYC 75022 x Xeon Platinum 82807K14K21K28K35KSE +/- 81.55, N = 3SE +/- 34.81, N = 333940.8132373.021. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.3.1Test / Class: SP.A2 x EPYC 75022 x Xeon Platinum 828090180270360450SE +/- 0.46, N = 3SE +/- 0.55, N = 3396.26379.981. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi

lzbench

lzbench is an in-memory benchmark of various compressors. The file used for compression is a Linux kernel source tree tarball. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is Betterlzbench 2017-08-08Test: Zstd 1 - Process: Decompression2 x Xeon Platinum 82802 x EPYC 77422 x EPYC 75022004006008001000SE +/- 7.88, N = 3SE +/- 0.33, N = 39679619321. (CXX) g++ options: -lrt -static -lpthread -fomit-frame-pointer -fstrict-aliasing -ffast-math -O3

OpenBenchmarking.orgMB/s, More Is Betterlzbench 2017-08-08Test: Brotli 0 - Process: Compression2 x Xeon Platinum 82802 x EPYC 77422 x EPYC 750280160240320400SE +/- 1.00, N = 3SE +/- 0.67, N = 33913833831. (CXX) g++ options: -lrt -static -lpthread -fomit-frame-pointer -fstrict-aliasing -ffast-math -O3

Blender

Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing is supported. This system/blender test profile makes use of the system-supplied Blender. Use pts/blender if wishing to stick to a fixed version of Blender. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 2.79Blend File: Barbershop - Compute: CPU-Only2 x EPYC 75022 x Xeon Platinum 828050100150200250SE +/- 0.50, N = 3SE +/- 0.08, N = 3236.53240.94

Minion

Minion is an open-source constraint solver that is designed to be very scalable. This test profile uses Minion's integrated benchmarking problems to solve. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterMinion 1.8Benchmark: Graceful2 x EPYC 75022 x Xeon Platinum 82801224364860SE +/- 0.13, N = 3SE +/- 0.06, N = 352.7853.281. (CXX) g++ options: -std=gnu++11 -O3 -fomit-frame-pointer -rdynamic

Phoronix Test Suite Self Test

This is a wrapper around the phoronix-test-suite debug-self-test sub-command for testing hot Phoronix Test Suite code paths and if wanting to compare different PHP implementations. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB, Fewer Is BetterPhoronix Test Suite Self Test 1.0Test: Peak Memory UsageRyzen 3900X2 x EPYC 75022 x Xeon Platinum 8280110220330440550SE +/- 0.00, N = 2SE +/- 0.00, N = 2SE +/- 0.00, N = 2500503503

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.3.1Test / Class: BT.A2 x EPYC 75022 x Xeon Platinum 82802004006008001000SE +/- 0.50, N = 3SE +/- 1.67, N = 3834.43834.301. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi

FFTE

FFTE is a package by Daisuke Takahashi to compute Discrete Fourier Transforms of 1-, 2- and 3- dimensional sequences of length (2^p)*(3^q)*(5^r). Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMFLOPS, More Is BetterFFTE 6.0Test: N=256, 1D Complex FFT Routine2 x EPYC 750216003200480064008000SE +/- 1.52, N = 37639.451. (F9X) gfortran options: -O3 -fomit-frame-pointer -fopenmp -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi

OSBench

OSBench is a collection of micro-benchmarks for measuring operating system primitives like time to create threads/processes, launching programs, creating files, and memory allocation. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgus Per Event, Fewer Is BetterOSBenchTest: Create ThreadsRyzen 3900X2 x Xeon Platinum 82802 x EPYC 7502816243240SE +/- 0.35, N = 12SE +/- 0.06, N = 3SE +/- 0.37, N = 314.6820.5133.791. (CC) gcc options: -lm

Sunflow Rendering System

This test runs benchmarks of the Sunflow Rendering System. The Sunflow Rendering System is an open-source render engine for photo-realistic image synthesis with a ray-tracing core. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterSunflow Rendering System 0.07.2Global Illumination + Image Synthesis2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X0.24750.4950.74250.991.2375SE +/- 0.01, N = 15SE +/- 0.01, N = 3SE +/- 0.02, N = 120.740.861.10

Memcached mcperf

This is a test of twmperf/mcperf with memcached. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOperations Per Second, More Is BetterMemcached mcperf 1.5.10Method: Prepend2 x Xeon Platinum 8280Ryzen 3900X2 x EPYC 750215K30K45K60K75KSE +/- 200.54, N = 3SE +/- 3176.72, N = 12SE +/- 158.97, N = 37062852587282791. (CC) gcc options: -O2 -lm -rdynamic

OpenBenchmarking.orgOperations Per Second, More Is BetterMemcached mcperf 1.5.10Method: Delete2 x Xeon Platinum 8280Ryzen 3900X2 x EPYC 750220K40K60K80K100KSE +/- 309.68, N = 3SE +/- 355.57, N = 3SE +/- 799.81, N = 1511302977722408031. (CC) gcc options: -O2 -lm -rdynamic

OpenBenchmarking.orgOperations Per Second, More Is BetterMemcached mcperf 1.5.10Method: Set2 x Xeon Platinum 8280Ryzen 3900X2 x EPYC 750213K26K39K52K65KSE +/- 106.71, N = 3SE +/- 4427.74, N = 12SE +/- 423.60, N = 56202158184335801. (CC) gcc options: -O2 -lm -rdynamic

OpenBenchmarking.orgOperations Per Second, More Is BetterMemcached mcperf 1.5.10Method: Add2 x Xeon Platinum 8280Ryzen 3900X2 x EPYC 750213K26K39K52K65KSE +/- 153.08, N = 3SE +/- 1038.10, N = 12SE +/- 537.65, N = 36151046988331571. (CC) gcc options: -O2 -lm -rdynamic

Stress-NG

Stress-NG is a Linux stress tool developed by Colin King of Canonical. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.07.26Test: Glibc C String Functions2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X1.1M2.2M3.3M4.4M5.5MSE +/- 54834.48, N = 6SE +/- 25939.97, N = 3SE +/- 28284.32, N = 124908942448365312627891. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.07.26Test: Context Switching2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X13M26M39M52M65MSE +/- 2793972.74, N = 13SE +/- 1626045.16, N = 15SE +/- 41870.47, N = 3620753602655021357218561. (CC) gcc options: -O2 -std=gnu99 -lm -lz -lcrypt -lrt -lpthread -laio -lc

Redis

Redis is an open-source data structure server. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgRequests Per Second, More Is BetterRedis 4.0.8Test: SETRyzen 3900X2 x Xeon Platinum 82802 x EPYC 7502400K800K1200K1600K2000KSE +/- 29404.03, N = 12SE +/- 21457.62, N = 3SE +/- 28243.18, N = 151792791175286014632461. (CC) gcc options: -ggdb -rdynamic -lm -ldl -pthread

OpenBenchmarking.orgRequests Per Second, More Is BetterRedis 4.0.8Test: GETRyzen 3900X2 x Xeon Platinum 82802 x EPYC 7502500K1000K1500K2000K2500KSE +/- 24305.19, N = 3SE +/- 30997.57, N = 15SE +/- 43798.91, N = 152538538229934918970431. (CC) gcc options: -ggdb -rdynamic -lm -ldl -pthread

OpenBenchmarking.orgRequests Per Second, More Is BetterRedis 4.0.8Test: LPUSHRyzen 3900X2 x Xeon Platinum 82802 x EPYC 7502400K800K1200K1600K2000KSE +/- 25015.51, N = 12SE +/- 18930.27, N = 15SE +/- 26034.52, N = 151738434165646413758531. (CC) gcc options: -ggdb -rdynamic -lm -ldl -pthread

OpenBenchmarking.orgRequests Per Second, More Is BetterRedis 4.0.8Test: LPOPRyzen 3900X2 x Xeon Platinum 82802 x EPYC 7502600K1200K1800K2400K3000KSE +/- 64125.23, N = 9SE +/- 23924.88, N = 15SE +/- 80051.04, N = 152727208247521421703591. (CC) gcc options: -ggdb -rdynamic -lm -ldl -pthread

Darktable

Darktable is an open-source photography / workflow application this will use any system-installed Darktable program or on Windows will automatically download the pre-built binary from the project. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.6.0Test: Server Room - Acceleration: CPU-only2 x Xeon Platinum 82802 x EPYC 75020.68181.36362.04542.72723.409SE +/- 0.02, N = 15SE +/- 0.17, N = 151.923.03

OpenBenchmarking.orgSeconds, Fewer Is BetterDarktable 2.6.0Test: Server Rack - Acceleration: CPU-only2 x Xeon Platinum 82802 x EPYC 75020.02480.04960.07440.09920.124SE +/- 0.01, N = 12SE +/- 0.01, N = 120.100.11

PostgreSQL pgbench

This is a simple benchmark of PostgreSQL using pgbench. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTPS, More Is BetterPostgreSQL pgbench 10.3Scaling: Buffer Test - Test: Normal Load - Mode: Read Write2 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X5K10K15K20K25KSE +/- 98.31, N = 3SE +/- 151.66, N = 3SE +/- 1091.07, N = 122370321936176561. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lpthread -lrt -lcrypt -ldl -lm

MariaDB

This is a MariaDB MySQL database server benchmark making use of mysqlslap. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgQueries Per Second, More Is BetterMariaDB 10.3.8Clients: 2562 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X90180270360450SE +/- 0.73, N = 3SE +/- 1.96, N = 3SE +/- 5.52, N = 9438270206-llzma -lbz2 -lnuma -lpcre-llzma -lbz2 -lnuma -lpcre1. (CXX) g++ options: -pie -fPIC -fstack-protector -fno-rtti -O2 -lpthread -laio -lz -lm -lcrypt -lssl -lcrypto -ldl

Timed Linux Kernel Compilation

This test times how long it takes to build the Linux kernel in a default configuration. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 4.18Time To Compile2 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X1122334455SE +/- 0.25, N = 15SE +/- 0.35, N = 14SE +/- 0.55, N = 319.7421.4547.79

MKL-DNN

This is a test of the Intel MKL-DNN as the Intel Math Kernel Library for Deep Neural Networks and making use of its built-in benchdnn functionality. The result is the total perf time reported. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN 2019-04-16Harness: Deconvolution Batch deconv_all - Data Type: f322 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X7K14K21K28K35KSE +/- 6.16, N = 3SE +/- 89.09, N = 9SE +/- 58.11, N = 31055422931340-O3 - MIN: 3301.92-O2 - MIN: 30877.11. (CXX) g++ options: -std=c++11 -march=native -mtune=native -fPIC -fopenmp -pie -lmklml_intel -ldl

OpenBenchmarking.orgms, Fewer Is BetterMKL-DNN 2019-04-16Harness: Convolution Batch conv_3d - Data Type: f322 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X20406080100SE +/- 0.01, N = 3SE +/- 0.10, N = 15SE +/- 0.34, N = 33.303.67101.70-O2 - MIN: 100.151. (CXX) g++ options: -std=c++11 -march=native -mtune=native -fPIC -fopenmp -pie -lmklml_intel -ldl

Renaissance

Renaissance is a suite of benchmarks designed to test the Java JVM from Apache Spark to a Twitter-like service to Scala and other features. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.9.0Test: Akka Unbalanced Cobwebbed TreeRyzen 3900X2 x Xeon Platinum 82802 x EPYC 75026K12K18K24K30KSE +/- 71.27, N = 8SE +/- 482.12, N = 32SE +/- 268.51, N = 32120082437330304

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.9.0Test: Apache Spark PageRank2 x Xeon Platinum 82802 x EPYC 77422 x EPYC 7502Ryzen 3900X6K12K18K24K30KSE +/- 200.36, N = 8SE +/- 224.85, N = 40SE +/- 294.87, N = 32SE +/- 234.52, N = 823691244122472627352

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.9.0Test: Savina Reactors.IORyzen 3900X2 x Xeon Platinum 82802 x EPYC 75022 x EPYC 77427K14K21K28K35KSE +/- 144.45, N = 8SE +/- 297.31, N = 40SE +/- 733.25, N = 40SE +/- 831.26, N = 4012205260582840431115

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.9.0Test: Apache Spark Bayes2 x Xeon Platinum 82802 x EPYC 75022 x EPYC 7742Ryzen 3900X2K4K6K8K10KSE +/- 49.08, N = 32SE +/- 42.13, N = 40SE +/- 69.40, N = 40SE +/- 95.93, N = 83283359541268382

OpenBenchmarking.orgms, Fewer Is BetterRenaissance 0.9.0Test: Apache Spark ALSRyzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 828011K22K33K44K55KSE +/- 32.09, N = 32SE +/- 1216.19, N = 40SE +/- 351.03, N = 40SE +/- 325.62, N = 404886375393877849161

Java SciMark

This test runs the Java version of SciMark 2.0, which is a benchmark for scientific and numerical computing developed by programmers at the National Institute of Standards and Technology. This benchmark is made up of Fast Foruier Transform, Jacobi Successive Over-relaxation, Monte Carlo, Sparse Matrix Multiply, and dense LU matrix factorization benchmarks. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterJava SciMark 2.0Computational Test: Fast Fourier Transform2 x EPYC 75022 x EPYC 77422 x Xeon Platinum 8280Ryzen 3900X30060090012001500SE +/- 7.36, N = 4SE +/- 4.68, N = 4SE +/- 5.82, N = 4SE +/- 227.53, N = 41624162213921280

Timed MAFFT Alignment

This test performs an alignment of 100 pyruvate decarboxylase sequences. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed MAFFT Alignment 7.392Multiple Sequence AlignmentRyzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 82801.1612.3223.4834.6445.805SE +/- 0.02, N = 2SE +/- 0.10, N = 15SE +/- 0.03, N = 15SE +/- 0.07, N = 32.083.413.425.161. (CC) gcc options: -std=c99 -O3 -lm -lpthread

FFTW

FFTW is a C subroutine library for computing the discrete Fourier transform (DFT) in one or more dimensions. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Float + SSE - Size: 1D FFT Size 256Ryzen 3900X2 x Xeon Platinum 82802 x EPYC 75022 x EPYC 77429K18K27K36K45KSE +/- 83.95, N = 3SE +/- 722.12, N = 15SE +/- 470.01, N = 3SE +/- 247.84, N = 3399783670427950278601. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Float + SSE - Size: 1D FFT Size 64Ryzen 3900X2 x Xeon Platinum 82802 x EPYC 75022 x EPYC 77425K10K15K20K25KSE +/- 12.55, N = 3SE +/- 219.94, N = 13SE +/- 26.96, N = 3SE +/- 401.12, N = 15223021698416447160321. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

OpenBenchmarking.orgMflops, More Is BetterFFTW 3.3.6Build: Stock - Size: 1D FFT Size 32Ryzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 82803K6K9K12K15KSE +/- 9.02, N = 3SE +/- 0.23, N = 3SE +/- 2.81, N = 3SE +/- 358.52, N = 15116918868874171791. (CC) gcc options: -pthread -O3 -fomit-frame-pointer -mtune=native -malign-double -fstrict-aliasing -fno-schedule-insns -ffast-math -lm

NAMD

NAMD is a parallel molecular dynamics code designed for high-performance simulation of large biomolecular systems. NAMD was developed by the Theoretical and Computational Biophysics Group in the Beckman Institute for Advanced Science and Technology at the University of Illinois at Urbana-Champaign. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgdays/ns, Fewer Is BetterNAMD 2.13b1ATPase Simulation - 327,506 Atoms2 x EPYC 77422 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X0.3250.650.9751.31.625SE +/- 0.00162, N = 8SE +/- 0.00011, N = 15SE +/- 0.02081, N = 15SE +/- 0.00006, N = 30.265890.359230.413491.44442

CLOMP

CLOMP is the C version of the Livermore OpenMP benchmark developed to measure OpenMP overheads and other performance impacts due to threading in order to influence future system designs. This particular test profile configuration is currently set to look at the OpenMP static schedule speed-up across all available CPU cores using the recommended test configuration. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSpeedup, More Is BetterCLOMP 3.3Static OMP SpeedupRyzen 3900X2 x Xeon Platinum 82802 x EPYC 75022 x EPYC 77420.55131.10261.65392.20522.7565SE +/- 0.02, N = 5SE +/- 0.00, N = 12SE +/- 0.01, N = 25SE +/- 0.00, N = 52.450.430.370.231. (CC) gcc options: --openmp -O3 -lm

High Performance Conjugate Gradient

HPCG is the High Performance Conjugate Gradient and is a new scientific benchmark from Sandia National Lans focused for super-computer testing with modern real-world workloads compared to HPCC. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHigh Performance Conjugate Gradient 3.02 x Xeon Platinum 82802 x EPYC 75022 x EPYC 77420.21150.4230.63450.8461.0575SE +/- 0.01, N = 3SE +/- 0.01, N = 12SE +/- 0.00, N = 30.940.460.35

Rodinia

Rodinia is a suite focused upon accelerating compute-intensive applications with accelerators. CUDA, OpenMP, and OpenCL parallel models are supported by the included applications. This profile utilizes the OpenCL and OpenMP test binaries at the moment. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 2.4Test: OpenMP Streamcluster2 x EPYC 77422 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X612182430SE +/- 0.06, N = 3SE +/- 0.01, N = 3SE +/- 0.38, N = 15SE +/- 0.01, N = 310.4311.7913.5723.711. (CXX) g++ options: -O2 -lOpenCL

OpenBenchmarking.orgSeconds, Fewer Is BetterRodinia 2.4Test: OpenMP LavaMD2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X510152025SE +/- 0.06, N = 4SE +/- 0.07, N = 15SE +/- 0.12, N = 15SE +/- 0.11, N = 34.336.116.9620.081. (CXX) g++ options: -O2 -lOpenCL

Parboil

The Parboil Benchmarks from the IMPACT Research Group at University of Illinois are a set of throughput computing applications for looking at computing architecture and compilers. Parboil test-cases support OpenMP, OpenCL, and CUDA multi-processing environments. However, at this time the test profile is just making use of the OpenMP and OpenCL test workloads. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterParboil 2.5Test: OpenMP Stencil2 x Xeon Platinum 82802 x EPYC 77422 x EPYC 7502Ryzen 3900X48121620SE +/- 0.06, N = 15SE +/- 0.01, N = 3SE +/- 0.05, N = 3SE +/- 0.01, N = 32.604.025.6416.901. (CXX) g++ options: -lm -lpthread -lgomp -O3 -ffast-math -fopenmp

OpenBenchmarking.orgSeconds, Fewer Is BetterParboil 2.5Test: OpenMP LBM2 x Xeon Platinum 82802 x EPYC 75022 x EPYC 7742Ryzen 3900X4080120160200SE +/- 0.52, N = 15SE +/- 0.29, N = 15SE +/- 0.30, N = 3SE +/- 0.04, N = 322.4325.0035.97168.101. (CXX) g++ options: -lm -lpthread -lgomp -O3 -ffast-math -fopenmp

HPC Challenge

HPC Challenge (HPCC) is a cluster-focused benchmark consisting of the HPL Linpack TPP benchmark, DGEMM, STREAM, PTRANS, RandomAccess, FFT, and communication bandwidth and latency. This HPC Challenge test profile attempts to ship with standard yet versatile configuration/input files though they can be modified. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGB/s, More Is BetterHPC Challenge 1.5.0Test / Class: G-Ptrans2 x Xeon Platinum 8280Ryzen 3900X2 x EPYC 750248121620SE +/- 0.01412, N = 3SE +/- 0.00171, N = 3SE +/- 0.21006, N = 315.018172.497711.368821. (CC) gcc options: -lblas -lm -pthread -lmpi -fomit-frame-pointer -funroll-loops2. 2 x Xeon Platinum 8280: ATLAS + Open MPI 3.1.33. Ryzen 3900X: ATLAS + Open MPI 3.1.44. 2 x EPYC 7502: ATLAS + Open MPI 3.1.3

NAS Parallel Benchmarks

NPB, NAS Parallel Benchmarks, is a benchmark developed by NASA for high-end computer systems. This test profile currently uses the MPI version of NPB. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgTotal Mop/s, More Is BetterNAS Parallel Benchmarks 3.3.1Test / Class: LU.A2 x EPYC 75022 x Xeon Platinum 82807K14K21K28K35KSE +/- 966.94, N = 12SE +/- 136.93, N = 331119.4630471.201. (F9X) gfortran options: -O3 -march=native -pthread -lmpi_usempif08 -lmpi_mpifh -lmpi

Sockperf

This is a network socket API performance benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgusec, Fewer Is BetterSockperf 3.4Test: Latency Under LoadRyzen 3900X2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 8280816243240SE +/- 1.02, N = 15SE +/- 0.96, N = 25SE +/- 0.38, N = 25SE +/- 1.51, N = 2518.5122.1622.4235.391. (CXX) g++ options: --param -O3 -rdynamic -ldl -lpthread

OpenBenchmarking.orgMessages Per Second, More Is BetterSockperf 3.4Test: Throughput2 x EPYC 75022 x EPYC 77422 x Xeon Platinum 8280Ryzen 3900X100K200K300K400K500KSE +/- 4514.21, N = 5SE +/- 7584.07, N = 25SE +/- 4430.66, N = 9SE +/- 4526.82, N = 54843634541954480193737771. (CXX) g++ options: --param -O3 -rdynamic -ldl -lpthread

Stream

This benchmark tests the system memory (RAM) performance. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMB/s, More Is BetterStream 2013-01-17Type: Add2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X40K80K120K160K200KSE +/- 713.30, N = 5SE +/- 1387.96, N = 5SE +/- 4249.65, N = 5SE +/- 31.89, N = 5194680164105156319162841. (CC) gcc options: -O3 -march=native -fopenmp

OpenBenchmarking.orgMB/s, More Is BetterStream 2013-01-17Type: Triad2 x EPYC 77422 x EPYC 75022 x Xeon Platinum 8280Ryzen 3900X40K80K120K160K200KSE +/- 624.77, N = 5SE +/- 2801.94, N = 5SE +/- 5754.74, N = 5SE +/- 2.30, N = 5200966166119154125163051. (CC) gcc options: -O3 -march=native -fopenmp

OpenBenchmarking.orgMB/s, More Is BetterStream 2013-01-17Type: Copy2 x EPYC 77422 x Xeon Platinum 82802 x EPYC 7502Ryzen 3900X40K80K120K160K200KSE +/- 645.18, N = 5SE +/- 4580.61, N = 25SE +/- 1580.30, N = 8SE +/- 21.78, N = 5185818168490150788149311. (CC) gcc options: -O3 -march=native -fopenmp

285 Results Shown

MKL-DNN
Stress-NG
MKL-DNN
MariaDB
HPC Challenge:
  Rand Ring Bandwidth
  G-Rand Access
  G-Ffte
  G-Ffte
MKL-DNN
John The Ripper
CP2K Molecular Dynamics
MKL-DNN:
  IP Batch 1D - u8s8f32s32
  Convolution Batch conv_alexnet - f32
  IP Batch 1D - u8s8u8s32
  Convolution Batch conv_all - f32
  Convolution Batch conv_googlenet_v3 - f32
  Convolution Batch conv_3d - u8s8u8s32
  Convolution Batch conv_3d - u8s8f32s32
  Deconvolution Batch deconv_3d - u8s8f32s32
Stress-NG:
  Matrix Math
  Glibc Qsort Data Sorting
  CPU Stress
  Bsearch
  Semaphores
OpenSSL
Stress-NG
MKL-DNN
MBW:
  Memory Copy - 8192 MiB
  Memory Copy, Fixed Block Size - 8192 MiB
MKL-DNN
Stress-NG:
  Hsearch
  Lsearch
Blender
MBW:
  Memory Copy - 4096 MiB
  Memory Copy - 1024 MiB
Stress-NG
C-Ray
m-queens
MBW:
  Memory Copy - 512 MiB
  Memory Copy - 128 MiB
Stress-NG
Blender
Coremark
Primesieve
asmFish
Stockfish
7-Zip Compression
Phoronix Test Suite Self Test
Blender
MariaDB
POV-Ray
Blender
Renaissance
PostgreSQL pgbench
Phoronix Test Suite Self Test
HPC Challenge
MKL-DNN
HPC Challenge
Blender
Himeno Benchmark
Appleseed
Stress-NG
Parboil
Phoronix Test Suite Self Test
Timed LLVM Compilation
Memcached mcperf
Y-Cruncher
PostgreSQL pgbench
HPC Challenge
OSBench
BLAKE2
Zstd Compression
Memcached mcperf
OSBench
MBW
Memcached mcperf
MBW:
  Memory Copy, Fixed Block Size - 512 MiB
  Memory Copy, Fixed Block Size - 4096 MiB
MKL-DNN:
  Convolution Batch conv_googlenet_v3 - u8s8u8s32
  Convolution Batch conv_alexnet - u8s8u8s32
  Convolution Batch conv_googlenet_v3 - u8s8f32s32
Phoronix Test Suite Self Test
MKL-DNN:
  Convolution Batch conv_all - u8s8u8s32
  Convolution Batch conv_all - u8s8f32s32
  Deconvolution Batch deconv_1d - u8s8f32s32
  Deconvolution Batch deconv_1d - u8s8u8s32
Parboil
MKL-DNN
MariaDB
HPC Challenge
MKL-DNN:
  IP Batch All - u8s8f32s32
  IP Batch All - u8s8u8s32
Stream
MariaDB
Phoronix Test Suite Self Test
FFTW
MBW
FFTW
RAMspeed SMP:
  Copy - Integer
  Add - Floating Point
Rodinia
RAMspeed SMP
FFTW:
  Float + SSE - 2D FFT Size 256
  Float + SSE - 2D FFT Size 32
Appleseed
Apache Siege
Java SciMark
Appleseed
PostgreSQL pgbench
MariaDB
RAMspeed SMP
BlogBench
Bork File Encrypter
FFTW
RAMspeed SMP
dav1d
dcraw
BlogBench
RAMspeed SMP
dav1d
RAMspeed SMP
Stress-NG
MariaDB
DaCapo Benchmark
RAMspeed SMP
Botan
SVT-AV1
Botan
TTSIOD 3D Renderer
VP9 libvpx Encoding
RAMspeed SMP
FFTW:
  Float + SSE - 1D FFT Size 512
  Float + SSE - 1D FFT Size 1024
Renaissance
Minion
PHP Micro Benchmarks
Tinymembench
FFTW
RAMspeed SMP
Java SciMark
HPC Challenge
NAS Parallel Benchmarks
FFTW:
  Float + SSE - 1D FFT Size 128
  Float + SSE - 1D FFT Size 2048
OSBench
FFTW:
  Stock - 2D FFT Size 512
  Stock - 2D FFT Size 4096
PHPBench
Apache Siege
Minion
NAS Parallel Benchmarks
FFTW
Tinymembench
libjpeg-turbo tjbench
SVT-VP9
Numpy Benchmark
Botan:
  CAST-256 - Decrypt
  CAST-256 - Encrypt
  KASUMI - Encrypt
FFTW
LAME MP3 Encoding
FFTW
Apache Siege
Mencoder
Botan
PHP Micro Benchmarks
FFTW
Apache Siege
Java SciMark
FLAC Audio Encoding
FFTW
Timed GCC Compilation
Apache Siege
Sockperf
FFTW
TSCP
x265
Botan
FFTW
DaCapo Benchmark
FFTW:
  Stock - 1D FFT Size 512
  Stock - 1D FFT Size 2048
GNU GMP GMPbench
SVT-HEVC
Botan
Apache Siege
Botan
FFTW:
  Stock - 2D FFT Size 32
  Stock - 1D FFT Size 4096
  Stock - 2D FFT Size 1024
Botan
FFTW
lzbench
IndigoBench:
  Bedroom
  Supercar
DaCapo Benchmark
FFTW:
  Float + SSE - 2D FFT Size 4096
  Stock - 1D FFT Size 256
  Float + SSE - 2D FFT Size 2048
Cython benchmark
x264
Java SciMark
OSBench
Stress-NG
Darktable
lzbench
Redis
XZ Compression
SPECjbb 2015
DaCapo Benchmark
lzbench
FFTW
Java SciMark
SPECjbb 2015
Darktable
lzbench
Blender
ctx_clock
Ogg Encoding
Blender
Timed PHP Compilation
Blender
lzbench
GIMP
NAS Parallel Benchmarks
Timed ImageMagick Compilation
GIMP
Blender
GIMP:
  auto-levels
  resize
Tesseract OCR
lzbench
Chaos Group V-RAY
NAS Parallel Benchmarks:
  LU.C
  SP.A
lzbench:
  Zstd 1 - Decompression
  Brotli 0 - Compression
Blender
Minion
Phoronix Test Suite Self Test
NAS Parallel Benchmarks
FFTE
OSBench
Sunflow Rendering System
Memcached mcperf:
  Prepend
  Delete
  Set
  Add
Stress-NG:
  Glibc C String Functions
  Context Switching
Redis:
  SET
  GET
  LPUSH
  LPOP
Darktable:
  Server Room - CPU-only
  Server Rack - CPU-only
PostgreSQL pgbench
MariaDB
Timed Linux Kernel Compilation
MKL-DNN:
  Deconvolution Batch deconv_all - f32
  Convolution Batch conv_3d - f32
Renaissance:
  Akka Unbalanced Cobwebbed Tree
  Apache Spark PageRank
  Savina Reactors.IO
  Apache Spark Bayes
  Apache Spark ALS
Java SciMark
Timed MAFFT Alignment
FFTW:
  Float + SSE - 1D FFT Size 256
  Float + SSE - 1D FFT Size 64
  Stock - 1D FFT Size 32
NAMD
CLOMP
High Performance Conjugate Gradient
Rodinia:
  OpenMP Streamcluster
  OpenMP LavaMD
Parboil:
  OpenMP Stencil
  OpenMP LBM
HPC Challenge
NAS Parallel Benchmarks
Sockperf:
  Latency Under Load
  Throughput
Stream:
  Add
  Triad
  Copy

2 x EPYC 7502

Processor: 2 x AMD EPYC 7502 32-Core @ 2.50GHz (64 Cores / 128 Threads), Motherboard: AMD DAYTONA_X (RDY1001C BIOS), Chipset: AMD Device 1480, Memory: 516096MB, Disk: 280GB INTEL SSDPE21D280GA + 280GB INTEL SSDPED1D280GA + 5 x 3841GB Micron_9300_MTFDHAL3T8TDP + 256GB Micron_1100_MTFD, Graphics: llvmpipe 504GB, Monitor: VE228, Network: 2 x Mellanox MT27710

OS: Ubuntu 19.04, Kernel: 5.2.0-050200rc7-generic (x86_64) 20190630, Desktop: GNOME Shell 3.32.1, Display Server: X Server 1.20.4, Display Driver: modesetting 1.20.4, OpenGL: 3.3 Mesa 19.0.2 (LLVM 8.0 128 bits), Compiler: GCC 8.3.0, File-System: ext4, Screen Resolution: 1920x1080

Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++ --enable-libmpx --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none --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-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib --with-tune=generic --without-cuda-driver -v
Disk Notes: NONE / errors=remount-ro,relatime,rw
Processor Notes: Scaling Governor: acpi-cpufreq ondemand
Java Notes: OpenJDK Runtime Environment (build 11.0.3+7-Ubuntu-1ubuntu219.04.1)
Python Notes: Python 2.7.16 + Python 3.7.3
Security Notes: l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional IBRS_FW STIBP: conditional RSB filling

Testing initiated at 25 July 2019 20:33 by user phoronix.

2 x EPYC 7742

Processor: 2 x AMD EPYC 7742 64-Core @ 2.25GHz (128 Cores / 256 Threads), Motherboard: AMD DAYTONA_X (RDY1001C BIOS), Chipset: AMD Device 1480, Memory: 516096MB, Disk: 280GB INTEL SSDPE21D280GA + 280GB INTEL SSDPED1D280GA + 5 x 3841GB Micron_9300_MTFDHAL3T8TDP + 256GB Micron_1100_MTFD, Graphics: ASPEED, Monitor: VE228, Network: 2 x Mellanox MT27710

OS: Ubuntu 19.04, Kernel: 5.3.0-rc1-power (x86_64) 20190728, Desktop: GNOME Shell 3.32.1, Display Server: X Server 1.20.4, Display Driver: modesetting 1.20.4, Compiler: GCC 8.3.0, File-System: ext4, Screen Resolution: 1920x1080

Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++ --enable-libmpx --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none --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-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib --with-tune=generic --without-cuda-driver -v
Disk Notes: NONE / errors=remount-ro,relatime,rw
Processor Notes: Scaling Governor: acpi-cpufreq ondemand
Java Notes: OpenJDK Runtime Environment (build 11.0.3+7-Ubuntu-1ubuntu219.04.1)
Python Notes: Python 2.7.16 + Python 3.7.3
Security Notes: l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of __user pointer sanitization + spectre_v2: Mitigation of Full AMD retpoline IBPB: conditional IBRS_FW STIBP: conditional RSB filling

Testing initiated at 28 July 2019 13:01 by user phoronix.

2 x Xeon Platinum 8280

Processor: 2 x Intel Xeon Platinum 8280 @ 4.00GHz (56 Cores / 112 Threads), Motherboard: GIGABYTE MD61-SC2-00 v01000100 (T15 BIOS), Chipset: Intel Sky Lake-E DMI3 Registers, Memory: 386048MB, Disk: 280GB INTEL SSDPED1D280GA + 32GB Flash Disk, Graphics: llvmpipe 377GB, Monitor: VE228, Network: 2 x Intel X722 for 1GbE + 2 x QLogic FastLinQ QL41000 10/25/40/50GbE

OS: Ubuntu 19.04, Kernel: 5.2.0-050200rc7-generic (x86_64) 20190630, Desktop: GNOME Shell 3.32.2, Display Server: X Server 1.20.4, Display Driver: modesetting 1.20.4, OpenGL: 3.3 Mesa 19.0.2 (LLVM 8.0 256 bits), Compiler: GCC 8.3.0, File-System: ext4, Screen Resolution: 1920x1080

Compiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++ --enable-libmpx --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none --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-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib --with-tune=generic --without-cuda-driver -v
Disk Notes: NONE / errors=remount-ro,relatime,rw
Processor Notes: Scaling Governor: intel_pstate powersave
Java Notes: OpenJDK Runtime Environment (build 11.0.4+11-post-Ubuntu-1ubuntu219.04)
Python Notes: Python 2.7.16 + Python 3.7.3
Security Notes: l1tf: Not affected + mds: Not affected + meltdown: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling

Testing initiated at 11 August 2019 08:29 by user phoronix.

Ryzen 3900X

Processor: AMD Ryzen 9 3900X 12-Core @ 3.80GHz (12 Cores / 24 Threads), Motherboard: Gigabyte X570 AORUS MASTER (F5k BIOS), Chipset: AMD Starship/Matisse, Memory: 16384MB, Disk: 2 x 2000GB GIGABYTE GP-ASM2NE6200TTTD, Graphics: Gigabyte NVIDIA GeForce RTX 2080 Ti Rev. A 11GB, Audio: NVIDIA TU102 HD Audio, Monitor: BenQ EL2870U, Network: Intel I211 + Realtek Device 8125 + Intel Wi-Fi 6 AX200

OS: Fedora 30, Kernel: 5.2.7-200.fc30.x86_64 (x86_64), Desktop: GNOME Shell 3.32.2, Display Server: X Server, Compiler: GCC 9.1.1 20190503, File-System: ext4, Screen Resolution: 3840x2160

Compiler Notes: --build=x86_64-redhat-linux --disable-libunwind-exceptions --enable-__cxa_atexit --enable-bootstrap --enable-cet --enable-checking=release --enable-gnu-indirect-function --enable-gnu-unique-object --enable-initfini-array --enable-languages=c,c++,fortran,objc,obj-c++,ada,go,d,lto --enable-multilib --enable-offload-targets=nvptx-none --enable-plugin --enable-shared --enable-threads=posix --mandir=/usr/share/man --with-arch_32=i686 --with-gcc-major-version-only --with-isl --with-linker-hash-style=gnu --with-tune=generic --without-cuda-driver
Disk Notes: NONE / relatime,rw,seclabel,stripe=256 / raid0 nvme1n1p3[0] nvme0n1p2[1]
Processor Notes: Scaling Governor: acpi-cpufreq ondemand
Python Notes: Python 2.7.16 + Python 3.7.4
Security Notes: SELinux + usercopy/swapgs barriers and __user pointer sanitization + Full AMD retpoline IBPB: conditional STIBP: always-on RSB filling + SSB disabled via prctl and seccomp

Testing initiated at 14 August 2019 10:04 by user w.