new xeon

Tests for a future article. Intel Xeon Gold 6421N testing with a Quanta Cloud S6Q-MB-MPS (3A10.uh BIOS) and ASPEED on Ubuntu 22.04 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 2307315-NE-NEWXEON9432
Jump To Table - Results

View

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

Statistics

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

Graph Settings

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

Multi-Way Comparison

Condense Multi-Option Tests Into Single Result Graphs

Table

Show Detailed System Result Table

Run Management

Highlight
Result
Toggle/Hide
Result
Result
Identifier
View Logs
Performance Per
Dollar
Date
Run
  Test
  Duration
a
July 30 2023
  5 Hours, 55 Minutes
b
July 31 2023
  5 Hours, 22 Minutes
Invert Behavior (Only Show Selected Data)
  5 Hours, 38 Minutes
Only show results matching title/arguments (delimit multiple options with a comma):
Do not show results matching title/arguments (delimit multiple options with a comma):


new xeonOpenBenchmarking.orgPhoronix Test SuiteIntel Xeon Gold 6421N @ 3.60GHz (32 Cores / 64 Threads)Quanta Cloud S6Q-MB-MPS (3A10.uh BIOS)Intel Device 1bce512GB3 x 3841GB Micron_9300_MTFDHAL3T8TDPASPEEDVGA HDMI4 x Intel E810-C for QSFPUbuntu 22.045.15.0-47-generic (x86_64)GNOME Shell 42.4X Server 1.21.1.31.2.204GCC 11.2.0ext41600x1200ProcessorMotherboardChipsetMemoryDiskGraphicsMonitorNetworkOSKernelDesktopDisplay ServerVulkanCompilerFile-SystemScreen ResolutionNew Xeon BenchmarksSystem Logs- Transparent Huge Pages: madvise- --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,brig,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-link-serialization=2 --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-targets=nvptx-none=/build/gcc-11-gBFGDP/gcc-11-11.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-11-gBFGDP/gcc-11-11.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -v - Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x2b0000c0 - OpenJDK Runtime Environment (build 11.0.16+8-post-Ubuntu-0ubuntu122.04)- Python 3.10.6- itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl and seccomp + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling + srbds: Not affected + tsx_async_abort: Not affected

a vs. b ComparisonPhoronix Test SuiteBaseline+9.5%+9.5%+19%+19%+28.5%+28.5%37.8%22.7%7.1%6.5%6.2%6.2%5.8%4.8%4.4%4.4%4%3.9%3.2%2.8%2.8%2.7%2.7%2.7%2.3%2%100 - 100 - 200100 - 100 - 20026%CPU Cache25615.9%200 - 100 - 200100 - 100 - 500500 - 1 - 5006.2%c2c - Stock - double - 128B.L.N.Q.A - A.M.SRedis - 100 - 1:106.2%200 - 100 - 2005.9%B.L.N.Q.A - A.M.S100 - 100 - 5005.4%500 - 1 - 500Cloning4.4%N.T.C.B.b.u.S - A.M.SN.T.C.B.b.u.S - A.M.S500 - 1 - 200r2c - Stock - float - 256500 - 1 - 2003.6%c2c - FFTW - double - 1283.4%Futex3.3%P.P.B.T.TPiper2c - FFTW - float - 256100 - 1 - 200200 - 1 - 200SENDFILERedis - 100 - 1:52.6%Matrix Math2.5%500 - 100 - 5002.5%200 - 100 - 5002.4%200 - 1 - 5002.4%200 - 100 - 50016 - 256 - 512Apache IoTDBApache IoTDBStress-NGlibxsmmApache IoTDBApache IoTDBApache IoTDBHeFFTe - Highly Efficient FFT for ExascaleNeural Magic DeepSparseRedis 7.0.12 + memtier_benchmarkApache IoTDBNeural Magic DeepSparseApache IoTDBApache IoTDBStress-NGNeural Magic DeepSparseNeural Magic DeepSparseApache IoTDBHeFFTe - Highly Efficient FFT for ExascaleApache IoTDBHeFFTe - Highly Efficient FFT for ExascaleStress-NGsrsRAN ProjectStress-NGHeFFTe - Highly Efficient FFT for ExascaleApache IoTDBApache IoTDBStress-NGRedis 7.0.12 + memtier_benchmarkStress-NGApache IoTDBApache IoTDBApache IoTDBApache IoTDBLiquid-DSPab

new xeonapache-iotdb: 100 - 100 - 200apache-iotdb: 100 - 100 - 200stress-ng: CPU Cachelibxsmm: 256apache-iotdb: 200 - 100 - 200apache-iotdb: 100 - 100 - 500apache-iotdb: 500 - 1 - 500heffte: c2c - Stock - double - 128deepsparse: BERT-Large, NLP Question Answering - Asynchronous Multi-Streammemtier-benchmark: Redis - 100 - 1:10apache-iotdb: 200 - 100 - 200deepsparse: BERT-Large, NLP Question Answering - Asynchronous Multi-Streamapache-iotdb: 100 - 100 - 500apache-iotdb: 500 - 1 - 500stress-ng: Cloningdeepsparse: NLP Text Classification, BERT base uncased SST2 - Asynchronous Multi-Streamdeepsparse: NLP Text Classification, BERT base uncased SST2 - Asynchronous Multi-Streamapache-iotdb: 500 - 1 - 200heffte: r2c - Stock - float - 256apache-iotdb: 500 - 1 - 200heffte: c2c - FFTW - double - 128stress-ng: Futexsrsran: PUSCH Processor Benchmark, Throughput Totalstress-ng: Pipeheffte: r2c - FFTW - float - 256apache-iotdb: 100 - 1 - 200apache-iotdb: 200 - 1 - 200stress-ng: SENDFILEmemtier-benchmark: Redis - 100 - 1:5stress-ng: Matrix Mathapache-iotdb: 500 - 100 - 500apache-iotdb: 200 - 100 - 500apache-iotdb: 200 - 1 - 500apache-iotdb: 200 - 100 - 500liquid-dsp: 16 - 256 - 512apache-iotdb: 100 - 1 - 200deepsparse: NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Streamdeepsparse: NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Streamsrsran: PUSCH Processor Benchmark, Throughput Threadstress-ng: IO_uringliquid-dsp: 16 - 256 - 57heffte: r2c - Stock - double - 128deepsparse: NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Streamdeepsparse: NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Streamapache-iotdb: 500 - 100 - 200stress-ng: Socket Activityapache-iotdb: 200 - 1 - 500liquid-dsp: 32 - 256 - 512vvenc: Bosphorus 4K - Fastbuild-llvm: Unix Makefilesheffte: r2c - Stock - float - 128deepsparse: CV Detection, YOLOv5s COCO, Sparse INT8 - Asynchronous Multi-Streamdeepsparse: CV Detection, YOLOv5s COCO, Sparse INT8 - Asynchronous Multi-Streamlibxsmm: 128heffte: c2c - FFTW - double - 256libxsmm: 32memtier-benchmark: Redis - 50 - 1:10heffte: c2c - FFTW - float - 256vvenc: Bosphorus 1080p - Faststress-ng: Atomicstress-ng: Semaphoresheffte: c2c - Stock - double - 256libxsmm: 64srsran: Downlink Processor Benchmarkapache-iotdb: 100 - 1 - 500stress-ng: MMAPheffte: r2c - FFTW - double - 128palabos: 400apache-iotdb: 100 - 1 - 500heffte: c2c - FFTW - float - 128heffte: r2c - FFTW - float - 128laghos: Triple Point Problemapache-iotdb: 500 - 100 - 500deepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Asynchronous Multi-Streamdeepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Asynchronous Multi-Streamstress-ng: Fused Multiply-Addhpcg: 160 160 160 - 60stress-ng: Function Callapache-iotdb: 500 - 100 - 200heffte: r2c - FFTW - double - 512liquid-dsp: 32 - 256 - 57stress-ng: NUMAstress-ng: Mutexheffte: c2c - Stock - float - 128deepsparse: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Asynchronous Multi-Streamdeepsparse: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Asynchronous Multi-Streamstress-ng: Wide Vector Mathapache-iotdb: 200 - 1 - 200liquid-dsp: 64 - 256 - 57deepsparse: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Asynchronous Multi-Streamdeepsparse: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Asynchronous Multi-Streamvvenc: Bosphorus 4K - Fastermemtier-benchmark: Redis - 50 - 1:5build-gdb: Time To Compilestress-ng: Glibc C String Functionshpcg: 104 104 104 - 60heffte: c2c - Stock - float - 256openfoam: drivaerFastback, Small Mesh Size - Execution Timedeepsparse: ResNet-50, Sparse INT8 - Asynchronous Multi-Streamdeepsparse: ResNet-50, Sparse INT8 - Asynchronous Multi-Streamheffte: c2c - Stock - double - 512palabos: 500heffte: r2c - Stock - double - 256heffte: c2c - FFTW - float - 512openfoam: drivaerFastback, Medium Mesh Size - Mesh Timeheffte: r2c - FFTW - float - 512deepsparse: ResNet-50, Baseline - Asynchronous Multi-Streamlaghos: Sedov Blast Wave, ube_922_hex.meshdeepsparse: ResNet-50, Baseline - Asynchronous Multi-Streamblender: BMW27 - CPU-Onlyheffte: r2c - Stock - float - 512stress-ng: AVL Treepalabos: 100stress-ng: Floating Pointliquid-dsp: 16 - 256 - 32heffte: r2c - FFTW - double - 256stress-ng: Mallocstress-ng: Hashdeepsparse: CV Detection, YOLOv5s COCO - Asynchronous Multi-Streamhpcg: 144 144 144 - 60deepsparse: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Asynchronous Multi-Streamdeepsparse: BERT-Large, NLP Question Answering, Sparse INT8 - Asynchronous Multi-Streambuild-llvm: Ninjastress-ng: Pthreadblender: Fishy Cat - CPU-Onlyheffte: c2c - FFTW - double - 512openfoam: drivaerFastback, Medium Mesh Size - Execution Timedeepsparse: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Asynchronous Multi-Streamdeepsparse: BERT-Large, NLP Question Answering, Sparse INT8 - Asynchronous Multi-Streambuild-php: Time To Compilestress-ng: MEMFDliquid-dsp: 32 - 256 - 32stress-ng: Context Switchingdeepsparse: NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Streamdeepsparse: CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Streamdeepsparse: CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Streamstress-ng: Pollvvenc: Bosphorus 1080p - Fasterstress-ng: Memory Copyingopenfoam: drivaerFastback, Small Mesh Size - Mesh Timestress-ng: Matrix 3D Mathstress-ng: Forkingdeepsparse: CV Detection, YOLOv5s COCO - Asynchronous Multi-Streamstress-ng: Glibc Qsort Data Sortingstress-ng: Zlibstress-ng: System V Message Passingblender: Barbershop - CPU-Onlybuild-linux-kernel: defconfigheffte: c2c - Stock - float - 512stress-ng: Vector Mathliquid-dsp: 64 - 256 - 32liquid-dsp: 64 - 256 - 512stress-ng: Vector Floating Pointblender: Classroom - CPU-Onlystress-ng: CPU Stressheffte: r2c - Stock - double - 512stress-ng: Cryptodeepsparse: NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Streamstress-ng: x86_64 RdRandstress-ng: Vector Shufflebuild-linux-kernel: allmodconfigbrl-cad: VGR Performance Metriccassandra: Writesblender: Pabellon Barcelona - CPU-Onlymemtier-benchmark: Redis - 500 - 1:5ab31.8343074031.841537111.20879.629.5469.0822.9746.635735.15292447092.0154224351.1453.480259041436.641916642.99740.57137.3780116.37619.49157.8671576432.2564.42631541676.365372.935837711.85149.82514.5811.86582724.632285996.17160653.4467607191.64101.251505080.3445677447.24243940000710382.4433.9358468.8046240.41529665.9884843500092.3973295.829554.067456894390.6124947.1426.293835550005.842323.856149.935208.847176.58071211.838.9304440.02316281.2676.029916.100133.8362126446.2138.9613833.8705.828.27861.28121.794287.2681191500.88131.656207.244177.7868.34345.149146.330734197705.6327.508622028.0331.5874.47341328100000390.8715147444.5185.7398131.4497121.68931745029.271045806.811728850000390.907640.910911.0202211638.6541.90526067360.6027.780875.089267.7073313227.09544.941640.7438300.27676.904278.8291144.69646141.41479.7876216.8633.327847.15137.536294.26235.18610587.4855794500072.289399373474.315577252.3276.559727.42131074.8218504.6114263.154136846.0164.0743.9665615.9907414.860031.675042.351549.948470850002572801.7534.5311478.910833.38943669281.6930.9467176.1927.9652149599.9389918.21208.8975696.652647.815852281.71493.4540.43872.5609151386.31157730000051313500058243.38127.7864111.1176.611050240.09460.7818331416.52167204.21445.385466686155626159.9443.8634191814.861885833.11758.931.6373.5621.6349.523037.32532304730.1951199962.11428.669556018457.872009050.469326.09143.4387111.49769.87164.0471521587.462.29741492979.465543.736852791.12154.05314.9812.18598173.562227152.02156668.4365935725.6798.871469808.8946726912.46248820000697217.5534.5447460.6707236.31503623.7986219500090.9851299.927753.329156137174.725282.3126.643786500005.917319.852151.803211.227075.72181225.038.5182444.62293467.6275.300116.249132.6161651485.4338.6757839.9710.928.45856.14122.460285.7611185338.02130.982206.217176.9268.01343.517046.549134050669.2327.397822106.4931.6974.71481323900000392.0815192892.5985.4850131.0664122.03671750003.431042859.031733700000391.912540.806110.9922217192.1242.00626125214.8427.840574.928667.5631633233.95884.931240.6648300.85577.034578.9605144.93674141.193480.5223217.1933.278147.22137.740294.66234.87410601.1055865500072.198199251227.285583978.1476.468427.38901075.9571505.1309262.884136709.8164.0144.0064615.4601814.847331.648842.382549.558476750002571092.6934.5539479.224133.36803671617.9730.9277180.4327.9487179605.3089966.29208.9908696.922648.815854201.78493.6140.45172.5391151431.15157685000051304000058232.70127.7664118.8776.604150243.48460.7588331423.04167202.07445.380OpenBenchmarking.org

Apache IoTDB

OpenBenchmarking.orgAverage Latency, More Is BetterApache IoTDB 1.1.2Device Count: 100 - Batch Size Per Write: 100 - Sensor Count: 200ba102030405043.8631.83MAX: 2550.76MAX: 790.74

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.1.2Device Count: 100 - Batch Size Per Write: 100 - Sensor Count: 200ba9M18M27M36M45M34191814.8643074031.84

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: CPU Cacheba400K800K1200K1600K2000KSE +/- 234949.06, N = 2SE +/- 31294.95, N = 21885833.111537111.201. (CXX) g++ options: -O2 -std=gnu99 -lc

libxsmm

Libxsmm is an open-source library for specialized dense and sparse matrix operations and deep learning primitives. Libxsmm supports making use of Intel AMX, AVX-512, and other modern CPU instruction set capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOPS/s, More Is Betterlibxsmm 2-1.17-3645M N K: 256ba2004006008001000SE +/- 5.75, N = 2SE +/- 0.65, N = 2758.9879.61. (CXX) g++ options: -dynamic -Bstatic -static-libgcc -lgomp -lm -lrt -ldl -lquadmath -lstdc++ -pthread -fPIC -std=c++14 -O2 -fopenmp-simd -funroll-loops -ftree-vectorize -fdata-sections -ffunction-sections -fvisibility=hidden -msse4.2

Apache IoTDB

OpenBenchmarking.orgAverage Latency, More Is BetterApache IoTDB 1.1.2Device Count: 200 - Batch Size Per Write: 100 - Sensor Count: 200ba71421283531.6329.54MAX: 718.08MAX: 746.57

OpenBenchmarking.orgAverage Latency, More Is BetterApache IoTDB 1.1.2Device Count: 100 - Batch Size Per Write: 100 - Sensor Count: 500ba163248648073.5669.08MAX: 1309.93MAX: 1049.85

OpenBenchmarking.orgAverage Latency, More Is BetterApache IoTDB 1.1.2Device Count: 500 - Batch Size Per Write: 1 - Sensor Count: 500ba61218243021.6322.97MAX: 867.44MAX: 864.74

HeFFTe - Highly Efficient FFT for Exascale

HeFFTe is the Highly Efficient FFT for Exascale software developed as part of the Exascale Computing Project. This test profile uses HeFFTe's built-in speed benchmarks under a variety of configuration options and currently catering to CPU/processor testing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: c2c - Backend: Stock - Precision: double - X Y Z: 128ba1122334455SE +/- 3.39, N = 2SE +/- 0.26, N = 249.5246.641. (CXX) g++ options: -O3

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: BERT-Large, NLP Question Answering - Scenario: Asynchronous Multi-Streamba918273645SE +/- 0.38, N = 2SE +/- 0.01, N = 237.3335.15

Redis 7.0.12 + memtier_benchmark

Memtier_benchmark is a NoSQL Redis/Memcache traffic generation plus benchmarking tool developed by Redis Labs. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps/sec, More Is BetterRedis 7.0.12 + memtier_benchmark 2.0Protocol: Redis - Clients: 100 - Set To Get Ratio: 1:10ba500K1000K1500K2000K2500KSE +/- 12975.09, N = 2SE +/- 114392.77, N = 22304730.192447092.011. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

Apache IoTDB

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.1.2Device Count: 200 - Batch Size Per Write: 100 - Sensor Count: 200ba12M24M36M48M60M51199962.1154224351.10

Neural Magic DeepSparse

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: BERT-Large, NLP Question Answering - Scenario: Asynchronous Multi-Streamba100200300400500SE +/- 4.41, N = 2SE +/- 0.26, N = 2428.67453.48

Apache IoTDB

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.1.2Device Count: 100 - Batch Size Per Write: 100 - Sensor Count: 500ba13M26M39M52M65M56018457.8759041436.64

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.1.2Device Count: 500 - Batch Size Per Write: 1 - Sensor Count: 500ba400K800K1200K1600K2000K2009050.461916642.90

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Cloningba2K4K6K8K10KSE +/- 100.16, N = 2SE +/- 114.33, N = 29326.099740.571. (CXX) g++ options: -O2 -std=gnu99 -lc

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-Streamba306090120150SE +/- 0.80, N = 2SE +/- 4.10, N = 2143.44137.38

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-Streamba306090120150SE +/- 0.59, N = 2SE +/- 3.45, N = 2111.50116.38

Apache IoTDB

OpenBenchmarking.orgAverage Latency, More Is BetterApache IoTDB 1.1.2Device Count: 500 - Batch Size Per Write: 1 - Sensor Count: 200ba36912159.879.49MAX: 820.85MAX: 845.95

HeFFTe - Highly Efficient FFT for Exascale

HeFFTe is the Highly Efficient FFT for Exascale software developed as part of the Exascale Computing Project. This test profile uses HeFFTe's built-in speed benchmarks under a variety of configuration options and currently catering to CPU/processor testing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: r2c - Backend: Stock - Precision: float - X Y Z: 256ba4080120160200SE +/- 3.21, N = 2SE +/- 6.51, N = 2164.05157.871. (CXX) g++ options: -O3

Apache IoTDB

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.1.2Device Count: 500 - Batch Size Per Write: 1 - Sensor Count: 200ba300K600K900K1200K1500K1521587.401576432.25

HeFFTe - Highly Efficient FFT for Exascale

HeFFTe is the Highly Efficient FFT for Exascale software developed as part of the Exascale Computing Project. This test profile uses HeFFTe's built-in speed benchmarks under a variety of configuration options and currently catering to CPU/processor testing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: c2c - Backend: FFTW - Precision: double - X Y Z: 128ba1428425670SE +/- 2.41, N = 2SE +/- 2.73, N = 262.3064.431. (CXX) g++ options: -O3

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Futexba300K600K900K1200K1500KSE +/- 45385.58, N = 2SE +/- 56630.43, N = 21492979.461541676.361. (CXX) g++ options: -O2 -std=gnu99 -lc

srsRAN Project

srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMbps, More Is BettersrsRAN Project 23.5Test: PUSCH Processor Benchmark, Throughput Totalba12002400360048006000SE +/- 95.40, N = 2SE +/- 143.30, N = 25543.75372.91. (CXX) g++ options: -march=native -mfma -O3 -fno-trapping-math -fno-math-errno -lgtest

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Pipeba8M16M24M32M40MSE +/- 79631.10, N = 2SE +/- 1105250.10, N = 236852791.1235837711.851. (CXX) g++ options: -O2 -std=gnu99 -lc

HeFFTe - Highly Efficient FFT for Exascale

HeFFTe is the Highly Efficient FFT for Exascale software developed as part of the Exascale Computing Project. This test profile uses HeFFTe's built-in speed benchmarks under a variety of configuration options and currently catering to CPU/processor testing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: r2c - Backend: FFTW - Precision: float - X Y Z: 256ba306090120150SE +/- 1.59, N = 2SE +/- 3.76, N = 2154.05149.831. (CXX) g++ options: -O3

Apache IoTDB

OpenBenchmarking.orgAverage Latency, More Is BetterApache IoTDB 1.1.2Device Count: 100 - Batch Size Per Write: 1 - Sensor Count: 200ba4812162014.9814.58MAX: 612.21MAX: 679.89

OpenBenchmarking.orgAverage Latency, More Is BetterApache IoTDB 1.1.2Device Count: 200 - Batch Size Per Write: 1 - Sensor Count: 200ba369121512.1811.86MAX: 586.62MAX: 573.1

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: SENDFILEba130K260K390K520K650KSE +/- 243.97, N = 2SE +/- 6799.74, N = 2598173.56582724.631. (CXX) g++ options: -O2 -std=gnu99 -lc

Redis 7.0.12 + memtier_benchmark

Memtier_benchmark is a NoSQL Redis/Memcache traffic generation plus benchmarking tool developed by Redis Labs. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps/sec, More Is BetterRedis 7.0.12 + memtier_benchmark 2.0Protocol: Redis - Clients: 100 - Set To Get Ratio: 1:5ba500K1000K1500K2000K2500KSE +/- 3990.38, N = 2SE +/- 6000.63, N = 22227152.022285996.171. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Matrix Mathba30K60K90K120K150KSE +/- 332.46, N = 2SE +/- 2867.57, N = 2156668.43160653.441. (CXX) g++ options: -O2 -std=gnu99 -lc

Apache IoTDB

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.1.2Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 500ba14M28M42M56M70M65935725.6767607191.64

OpenBenchmarking.orgAverage Latency, More Is BetterApache IoTDB 1.1.2Device Count: 200 - Batch Size Per Write: 100 - Sensor Count: 500ba2040608010098.87101.25MAX: 3564.64MAX: 3631.89

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.1.2Device Count: 200 - Batch Size Per Write: 1 - Sensor Count: 500ba300K600K900K1200K1500K1469808.891505080.34

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.1.2Device Count: 200 - Batch Size Per Write: 100 - Sensor Count: 500ba10M20M30M40M50M46726912.4645677447.24

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 16 - Buffer Length: 256 - Filter Length: 512ba50M100M150M200M250MSE +/- 3170000.00, N = 2SE +/- 1950000.00, N = 22488200002439400001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

Apache IoTDB

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.1.2Device Count: 100 - Batch Size Per Write: 1 - Sensor Count: 200ba150K300K450K600K750K697217.55710382.44

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Streamba816243240SE +/- 0.03, N = 2SE +/- 0.07, N = 234.5433.94

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Streamba100200300400500SE +/- 0.20, N = 2SE +/- 1.46, N = 2460.67468.80

srsRAN Project

srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMbps, More Is BettersrsRAN Project 23.5Test: PUSCH Processor Benchmark, Throughput Threadba50100150200250SE +/- 0.10, N = 2SE +/- 3.55, N = 2236.3240.41. (CXX) g++ options: -march=native -mfma -O3 -fno-trapping-math -fno-math-errno -lgtest

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: IO_uringba300K600K900K1200K1500KSE +/- 5229.94, N = 2SE +/- 22482.34, N = 21503623.791529665.981. (CXX) g++ options: -O2 -std=gnu99 -lc

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 16 - Buffer Length: 256 - Filter Length: 57ba200M400M600M800M1000MSE +/- 695000.00, N = 2SE +/- 14365000.00, N = 28621950008484350001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

HeFFTe - Highly Efficient FFT for Exascale

HeFFTe is the Highly Efficient FFT for Exascale software developed as part of the Exascale Computing Project. This test profile uses HeFFTe's built-in speed benchmarks under a variety of configuration options and currently catering to CPU/processor testing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: r2c - Backend: Stock - Precision: double - X Y Z: 128ba20406080100SE +/- 0.09, N = 2SE +/- 0.90, N = 290.9992.401. (CXX) g++ options: -O3

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Streamba70140210280350SE +/- 0.51, N = 2SE +/- 0.05, N = 2299.93295.83

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Streamba1224364860SE +/- 0.09, N = 2SE +/- 0.01, N = 253.3354.07

Apache IoTDB

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.1.2Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 200ba12M24M36M48M60M56137174.7056894390.61

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Socket Activityba5K10K15K20K25KSE +/- 267.39, N = 2SE +/- 72.57, N = 225282.3124947.141. (CXX) g++ options: -O2 -std=gnu99 -lc

Apache IoTDB

OpenBenchmarking.orgAverage Latency, More Is BetterApache IoTDB 1.1.2Device Count: 200 - Batch Size Per Write: 1 - Sensor Count: 500ba61218243026.6426.29MAX: 636.93MAX: 620.79

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 32 - Buffer Length: 256 - Filter Length: 512ba80M160M240M320M400MSE +/- 4920000.00, N = 2SE +/- 1955000.00, N = 23786500003835550001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

VVenC

VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterVVenC 1.9Video Input: Bosphorus 4K - Video Preset: Fastba1.33132.66263.99395.32526.6565SE +/- 0.015, N = 2SE +/- 0.074, N = 25.9175.8421. (CXX) g++ options: -O3 -flto -fno-fat-lto-objects -flto=auto

Timed LLVM Compilation

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

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed LLVM Compilation 16.0Build System: Unix Makefilesba70140210280350SE +/- 5.88, N = 2SE +/- 5.08, N = 2319.85323.86

HeFFTe - Highly Efficient FFT for Exascale

HeFFTe is the Highly Efficient FFT for Exascale software developed as part of the Exascale Computing Project. This test profile uses HeFFTe's built-in speed benchmarks under a variety of configuration options and currently catering to CPU/processor testing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: r2c - Backend: Stock - Precision: float - X Y Z: 128ba306090120150SE +/- 1.24, N = 2SE +/- 1.93, N = 2151.80149.941. (CXX) g++ options: -O3

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: CV Detection, YOLOv5s COCO, Sparse INT8 - Scenario: Asynchronous Multi-Streamba50100150200250SE +/- 0.12, N = 2SE +/- 0.34, N = 2211.23208.85

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: CV Detection, YOLOv5s COCO, Sparse INT8 - Scenario: Asynchronous Multi-Streamba20406080100SE +/- 0.04, N = 2SE +/- 0.13, N = 275.7276.58

libxsmm

Libxsmm is an open-source library for specialized dense and sparse matrix operations and deep learning primitives. Libxsmm supports making use of Intel AMX, AVX-512, and other modern CPU instruction set capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOPS/s, More Is Betterlibxsmm 2-1.17-3645M N K: 128ba30060090012001500SE +/- 1.10, N = 2SE +/- 4.60, N = 21225.01211.81. (CXX) g++ options: -dynamic -Bstatic -static-libgcc -lgomp -lm -lrt -ldl -lquadmath -lstdc++ -pthread -fPIC -std=c++14 -O2 -fopenmp-simd -funroll-loops -ftree-vectorize -fdata-sections -ffunction-sections -fvisibility=hidden -msse4.2

HeFFTe - Highly Efficient FFT for Exascale

HeFFTe is the Highly Efficient FFT for Exascale software developed as part of the Exascale Computing Project. This test profile uses HeFFTe's built-in speed benchmarks under a variety of configuration options and currently catering to CPU/processor testing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: c2c - Backend: FFTW - Precision: double - X Y Z: 256ba918273645SE +/- 0.16, N = 2SE +/- 0.25, N = 238.5238.931. (CXX) g++ options: -O3

libxsmm

Libxsmm is an open-source library for specialized dense and sparse matrix operations and deep learning primitives. Libxsmm supports making use of Intel AMX, AVX-512, and other modern CPU instruction set capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOPS/s, More Is Betterlibxsmm 2-1.17-3645M N K: 32ba100200300400500SE +/- 0.15, N = 2SE +/- 0.25, N = 2444.6440.01. (CXX) g++ options: -dynamic -Bstatic -static-libgcc -lgomp -lm -lrt -ldl -lquadmath -lstdc++ -pthread -fPIC -std=c++14 -O2 -fopenmp-simd -funroll-loops -ftree-vectorize -fdata-sections -ffunction-sections -fvisibility=hidden -msse4.2

Redis 7.0.12 + memtier_benchmark

Memtier_benchmark is a NoSQL Redis/Memcache traffic generation plus benchmarking tool developed by Redis Labs. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps/sec, More Is BetterRedis 7.0.12 + memtier_benchmark 2.0Protocol: Redis - Clients: 50 - Set To Get Ratio: 1:10ba500K1000K1500K2000K2500KSE +/- 4548.93, N = 2SE +/- 13610.76, N = 22293467.622316281.261. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

HeFFTe - Highly Efficient FFT for Exascale

HeFFTe is the Highly Efficient FFT for Exascale software developed as part of the Exascale Computing Project. This test profile uses HeFFTe's built-in speed benchmarks under a variety of configuration options and currently catering to CPU/processor testing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: c2c - Backend: FFTW - Precision: float - X Y Z: 256ba20406080100SE +/- 0.08, N = 2SE +/- 0.70, N = 275.3076.031. (CXX) g++ options: -O3

VVenC

VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterVVenC 1.9Video Input: Bosphorus 1080p - Video Preset: Fastba48121620SE +/- 0.02, N = 2SE +/- 0.17, N = 216.2516.101. (CXX) g++ options: -O3 -flto -fno-fat-lto-objects -flto=auto

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Atomicba306090120150SE +/- 0.20, N = 2SE +/- 1.05, N = 2132.61133.831. (CXX) g++ options: -O2 -std=gnu99 -lc

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Semaphoresba13M26M39M52M65MSE +/- 466593.23, N = 2SE +/- 2077286.42, N = 261651485.4362126446.211. (CXX) g++ options: -O2 -std=gnu99 -lc

HeFFTe - Highly Efficient FFT for Exascale

HeFFTe is the Highly Efficient FFT for Exascale software developed as part of the Exascale Computing Project. This test profile uses HeFFTe's built-in speed benchmarks under a variety of configuration options and currently catering to CPU/processor testing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: c2c - Backend: Stock - Precision: double - X Y Z: 256ba918273645SE +/- 0.07, N = 2SE +/- 0.07, N = 238.6838.961. (CXX) g++ options: -O3

libxsmm

Libxsmm is an open-source library for specialized dense and sparse matrix operations and deep learning primitives. Libxsmm supports making use of Intel AMX, AVX-512, and other modern CPU instruction set capabilities. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOPS/s, More Is Betterlibxsmm 2-1.17-3645M N K: 64ba2004006008001000SE +/- 0.20, N = 2SE +/- 1.05, N = 2839.9833.81. (CXX) g++ options: -dynamic -Bstatic -static-libgcc -lgomp -lm -lrt -ldl -lquadmath -lstdc++ -pthread -fPIC -std=c++14 -O2 -fopenmp-simd -funroll-loops -ftree-vectorize -fdata-sections -ffunction-sections -fvisibility=hidden -msse4.2

srsRAN Project

srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMbps, More Is BettersrsRAN Project 23.5Test: Downlink Processor Benchmarkba150300450600750SE +/- 1.60, N = 2SE +/- 5.15, N = 2710.9705.81. (CXX) g++ options: -march=native -mfma -O3 -fno-trapping-math -fno-math-errno -lgtest

Apache IoTDB

OpenBenchmarking.orgAverage Latency, More Is BetterApache IoTDB 1.1.2Device Count: 100 - Batch Size Per Write: 1 - Sensor Count: 500ba71421283528.4528.27MAX: 664.29MAX: 671.77

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: MMAPba2004006008001000SE +/- 2.06, N = 2SE +/- 3.32, N = 2856.14861.281. (CXX) g++ options: -O2 -std=gnu99 -lc

HeFFTe - Highly Efficient FFT for Exascale

HeFFTe is the Highly Efficient FFT for Exascale software developed as part of the Exascale Computing Project. This test profile uses HeFFTe's built-in speed benchmarks under a variety of configuration options and currently catering to CPU/processor testing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: r2c - Backend: FFTW - Precision: double - X Y Z: 128ba306090120150SE +/- 1.22, N = 2SE +/- 0.56, N = 2122.46121.791. (CXX) g++ options: -O3

Palabos

The Palabos library is a framework for general purpose Computational Fluid Dynamics (CFD). Palabos uses a kernel based on the Lattice Boltzmann method. This test profile uses the Palabos MPI-based Cavity3D benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMega Site Updates Per Second, More Is BetterPalabos 2.3Grid Size: 400ba60120180240300SE +/- 1.54, N = 2SE +/- 0.49, N = 2285.76287.271. (CXX) g++ options: -std=c++17 -pedantic -O3 -rdynamic -lcrypto -lcurl -lsz -lz -ldl -lm

Apache IoTDB

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.1.2Device Count: 100 - Batch Size Per Write: 1 - Sensor Count: 500ba300K600K900K1200K1500K1185338.021191500.88

HeFFTe - Highly Efficient FFT for Exascale

HeFFTe is the Highly Efficient FFT for Exascale software developed as part of the Exascale Computing Project. This test profile uses HeFFTe's built-in speed benchmarks under a variety of configuration options and currently catering to CPU/processor testing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: c2c - Backend: FFTW - Precision: float - X Y Z: 128ba306090120150SE +/- 0.61, N = 2SE +/- 0.77, N = 2130.98131.661. (CXX) g++ options: -O3

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: r2c - Backend: FFTW - Precision: float - X Y Z: 128ba50100150200250SE +/- 0.19, N = 2SE +/- 0.61, N = 2206.22207.241. (CXX) g++ options: -O3

Laghos

Laghos (LAGrangian High-Order Solver) is a miniapp that solves the time-dependent Euler equations of compressible gas dynamics in a moving Lagrangian frame using unstructured high-order finite element spatial discretization and explicit high-order time-stepping. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMajor Kernels Total Rate, More Is BetterLaghos 3.1Test: Triple Point Problemba4080120160200SE +/- 0.02, N = 2SE +/- 0.13, N = 2176.92177.781. (CXX) g++ options: -O3 -std=c++11 -lmfem -lHYPRE -lmetis -lrt -lmpi_cxx -lmpi

Apache IoTDB

OpenBenchmarking.orgAverage Latency, More Is BetterApache IoTDB 1.1.2Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 500ba153045607568.0168.34MAX: 1606.75MAX: 2006.68

Neural Magic DeepSparse

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Streamba80160240320400SE +/- 1.63, N = 2SE +/- 0.15, N = 2343.52345.15

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Streamba1122334455SE +/- 0.20, N = 2SE +/- 0.02, N = 246.5546.33

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Fused Multiply-Addba7M14M21M28M35MSE +/- 285.63, N = 2SE +/- 137631.48, N = 234050669.2334197705.631. (CXX) g++ options: -O2 -std=gnu99 -lc

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 3.1X Y Z: 160 160 160 - RT: 60ba612182430SE +/- 0.07, N = 2SE +/- 0.03, N = 227.4027.511. (CXX) g++ options: -O3 -ffast-math -ftree-vectorize -lmpi_cxx -lmpi

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Function Callba5K10K15K20K25KSE +/- 74.09, N = 2SE +/- 80.03, N = 222106.4922028.031. (CXX) g++ options: -O2 -std=gnu99 -lc

Apache IoTDB

OpenBenchmarking.orgAverage Latency, More Is BetterApache IoTDB 1.1.2Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 200ba71421283531.6931.58MAX: 1610.79MAX: 1920.32

HeFFTe - Highly Efficient FFT for Exascale

HeFFTe is the Highly Efficient FFT for Exascale software developed as part of the Exascale Computing Project. This test profile uses HeFFTe's built-in speed benchmarks under a variety of configuration options and currently catering to CPU/processor testing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: r2c - Backend: FFTW - Precision: double - X Y Z: 512ba20406080100SE +/- 0.16, N = 2SE +/- 0.48, N = 274.7174.471. (CXX) g++ options: -O3

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 32 - Buffer Length: 256 - Filter Length: 57ba300M600M900M1200M1500MSE +/- 4400000.00, N = 2SE +/- 300000.00, N = 2132390000013281000001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: NUMAba90180270360450SE +/- 0.05, N = 2SE +/- 0.88, N = 2392.08390.871. (CXX) g++ options: -O2 -std=gnu99 -lc

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Mutexba3M6M9M12M15MSE +/- 2864.48, N = 2SE +/- 23940.47, N = 215192892.5915147444.511. (CXX) g++ options: -O2 -std=gnu99 -lc

HeFFTe - Highly Efficient FFT for Exascale

HeFFTe is the Highly Efficient FFT for Exascale software developed as part of the Exascale Computing Project. This test profile uses HeFFTe's built-in speed benchmarks under a variety of configuration options and currently catering to CPU/processor testing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: c2c - Backend: Stock - Precision: float - X Y Z: 128ba20406080100SE +/- 0.88, N = 2SE +/- 1.30, N = 285.4985.741. (CXX) g++ options: -O3

Neural Magic DeepSparse

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-Streamba306090120150SE +/- 0.22, N = 2SE +/- 0.05, N = 2131.07131.45

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-Streamba306090120150SE +/- 0.22, N = 2SE +/- 0.05, N = 2122.04121.69

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Wide Vector Mathba400K800K1200K1600K2000KSE +/- 4139.63, N = 2SE +/- 918.08, N = 21750003.431745029.271. (CXX) g++ options: -O2 -std=gnu99 -lc

Apache IoTDB

OpenBenchmarking.orgpoint/sec, More Is BetterApache IoTDB 1.1.2Device Count: 200 - Batch Size Per Write: 1 - Sensor Count: 200ba200K400K600K800K1000K1042859.031045806.81

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 64 - Buffer Length: 256 - Filter Length: 57ba400M800M1200M1600M2000MSE +/- 900000.00, N = 2SE +/- 550000.00, N = 2173370000017288500001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Asynchronous Multi-Streamba90180270360450SE +/- 0.12, N = 2SE +/- 1.01, N = 2391.91390.91

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Asynchronous Multi-Streamba918273645SE +/- 0.01, N = 2SE +/- 0.11, N = 240.8140.91

VVenC

VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterVVenC 1.9Video Input: Bosphorus 4K - Video Preset: Fasterba3691215SE +/- 0.03, N = 2SE +/- 0.00, N = 210.9911.021. (CXX) g++ options: -O3 -flto -fno-fat-lto-objects -flto=auto

Redis 7.0.12 + memtier_benchmark

Memtier_benchmark is a NoSQL Redis/Memcache traffic generation plus benchmarking tool developed by Redis Labs. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOps/sec, More Is BetterRedis 7.0.12 + memtier_benchmark 2.0Protocol: Redis - Clients: 50 - Set To Get Ratio: 1:5ba500K1000K1500K2000K2500KSE +/- 39004.04, N = 2SE +/- 31848.80, N = 22217192.122211638.651. (CXX) g++ options: -O2 -levent_openssl -levent -lcrypto -lssl -lpthread -lz -lpcre

Timed GDB GNU Debugger Compilation

This test times how long it takes to build the GNU Debugger (GDB) in a default configuration. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed GDB GNU Debugger Compilation 10.2Time To Compileba1020304050SE +/- 0.12, N = 2SE +/- 0.06, N = 242.0141.91

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Glibc C String Functionsba6M12M18M24M30MSE +/- 69329.81, N = 2SE +/- 150617.25, N = 226125214.8426067360.601. (CXX) g++ options: -O2 -std=gnu99 -lc

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 3.1X Y Z: 104 104 104 - RT: 60ba714212835SE +/- 0.01, N = 2SE +/- 0.03, N = 227.8427.781. (CXX) g++ options: -O3 -ffast-math -ftree-vectorize -lmpi_cxx -lmpi

HeFFTe - Highly Efficient FFT for Exascale

HeFFTe is the Highly Efficient FFT for Exascale software developed as part of the Exascale Computing Project. This test profile uses HeFFTe's built-in speed benchmarks under a variety of configuration options and currently catering to CPU/processor testing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: c2c - Backend: Stock - Precision: float - X Y Z: 256ba20406080100SE +/- 0.10, N = 2SE +/- 0.48, N = 274.9375.091. (CXX) g++ options: -O3

OpenFOAM

OpenFOAM is the leading free, open-source software for computational fluid dynamics (CFD). This test profile currently uses the drivaerFastback test case for analyzing automotive aerodynamics or alternatively the older motorBike input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 10Input: drivaerFastback, Small Mesh Size - Execution Timeba1530456075SE +/- 0.11, N = 2SE +/- 0.09, N = 267.5667.711. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfiniteVolume -lmeshTools -lparallel -llagrangian -lregionModels -lgenericPatchFields -lOpenFOAM -ldl -lm

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: ResNet-50, Sparse INT8 - Scenario: Asynchronous Multi-Streamba7001400210028003500SE +/- 3.51, N = 2SE +/- 8.40, N = 23233.963227.10

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: ResNet-50, Sparse INT8 - Scenario: Asynchronous Multi-Streamba1.11192.22383.33574.44765.5595SE +/- 0.0056, N = 2SE +/- 0.0128, N = 24.93124.9416

HeFFTe - Highly Efficient FFT for Exascale

HeFFTe is the Highly Efficient FFT for Exascale software developed as part of the Exascale Computing Project. This test profile uses HeFFTe's built-in speed benchmarks under a variety of configuration options and currently catering to CPU/processor testing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: c2c - Backend: Stock - Precision: double - X Y Z: 512ba918273645SE +/- 0.00, N = 2SE +/- 0.05, N = 240.6640.741. (CXX) g++ options: -O3

Palabos

The Palabos library is a framework for general purpose Computational Fluid Dynamics (CFD). Palabos uses a kernel based on the Lattice Boltzmann method. This test profile uses the Palabos MPI-based Cavity3D benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMega Site Updates Per Second, More Is BetterPalabos 2.3Grid Size: 500ba70140210280350SE +/- 1.17, N = 2SE +/- 1.63, N = 2300.86300.281. (CXX) g++ options: -std=c++17 -pedantic -O3 -rdynamic -lcrypto -lcurl -lsz -lz -ldl -lm

HeFFTe - Highly Efficient FFT for Exascale

HeFFTe is the Highly Efficient FFT for Exascale software developed as part of the Exascale Computing Project. This test profile uses HeFFTe's built-in speed benchmarks under a variety of configuration options and currently catering to CPU/processor testing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: r2c - Backend: Stock - Precision: double - X Y Z: 256ba20406080100SE +/- 0.65, N = 2SE +/- 0.40, N = 277.0376.901. (CXX) g++ options: -O3

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: c2c - Backend: FFTW - Precision: float - X Y Z: 512ba20406080100SE +/- 0.06, N = 2SE +/- 0.36, N = 278.9678.831. (CXX) g++ options: -O3

OpenFOAM

OpenFOAM is the leading free, open-source software for computational fluid dynamics (CFD). This test profile currently uses the drivaerFastback test case for analyzing automotive aerodynamics or alternatively the older motorBike input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 10Input: drivaerFastback, Medium Mesh Size - Mesh Timeba306090120150SE +/- 0.08, N = 2SE +/- 0.01, N = 2144.94144.701. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfiniteVolume -lmeshTools -lparallel -llagrangian -lregionModels -lgenericPatchFields -lOpenFOAM -ldl -lm

HeFFTe - Highly Efficient FFT for Exascale

HeFFTe is the Highly Efficient FFT for Exascale software developed as part of the Exascale Computing Project. This test profile uses HeFFTe's built-in speed benchmarks under a variety of configuration options and currently catering to CPU/processor testing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: r2c - Backend: FFTW - Precision: float - X Y Z: 512ba306090120150SE +/- 0.20, N = 2SE +/- 0.63, N = 2141.19141.411. (CXX) g++ options: -O3

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: ResNet-50, Baseline - Scenario: Asynchronous Multi-Streamba100200300400500SE +/- 0.54, N = 2SE +/- 0.12, N = 2480.52479.79

Laghos

Laghos (LAGrangian High-Order Solver) is a miniapp that solves the time-dependent Euler equations of compressible gas dynamics in a moving Lagrangian frame using unstructured high-order finite element spatial discretization and explicit high-order time-stepping. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMajor Kernels Total Rate, More Is BetterLaghos 3.1Test: Sedov Blast Wave, ube_922_hex.meshba50100150200250SE +/- 0.18, N = 2SE +/- 0.24, N = 2217.19216.861. (CXX) g++ options: -O3 -std=c++11 -lmfem -lHYPRE -lmetis -lrt -lmpi_cxx -lmpi

Neural Magic DeepSparse

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: ResNet-50, Baseline - Scenario: Asynchronous Multi-Streamba816243240SE +/- 0.04, N = 2SE +/- 0.01, N = 233.2833.33

Blender

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.6Blend File: BMW27 - Compute: CPU-Onlyba1122334455SE +/- 0.08, N = 2SE +/- 0.02, N = 247.2247.15

HeFFTe - Highly Efficient FFT for Exascale

HeFFTe is the Highly Efficient FFT for Exascale software developed as part of the Exascale Computing Project. This test profile uses HeFFTe's built-in speed benchmarks under a variety of configuration options and currently catering to CPU/processor testing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: r2c - Backend: Stock - Precision: float - X Y Z: 512ba306090120150SE +/- 0.33, N = 2SE +/- 0.00, N = 2137.74137.541. (CXX) g++ options: -O3

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: AVL Treeba60120180240300SE +/- 0.85, N = 2SE +/- 0.32, N = 2294.66294.261. (CXX) g++ options: -O2 -std=gnu99 -lc

Palabos

The Palabos library is a framework for general purpose Computational Fluid Dynamics (CFD). Palabos uses a kernel based on the Lattice Boltzmann method. This test profile uses the Palabos MPI-based Cavity3D benchmark. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgMega Site Updates Per Second, More Is BetterPalabos 2.3Grid Size: 100ba50100150200250SE +/- 0.34, N = 2SE +/- 0.02, N = 2234.87235.191. (CXX) g++ options: -std=c++17 -pedantic -O3 -rdynamic -lcrypto -lcurl -lsz -lz -ldl -lm

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Floating Pointba2K4K6K8K10KSE +/- 17.77, N = 2SE +/- 1.07, N = 210601.1010587.481. (CXX) g++ options: -O2 -std=gnu99 -lc

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 16 - Buffer Length: 256 - Filter Length: 32ba120M240M360M480M600MSE +/- 605000.00, N = 2SE +/- 2065000.00, N = 25586550005579450001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

HeFFTe - Highly Efficient FFT for Exascale

HeFFTe is the Highly Efficient FFT for Exascale software developed as part of the Exascale Computing Project. This test profile uses HeFFTe's built-in speed benchmarks under a variety of configuration options and currently catering to CPU/processor testing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: r2c - Backend: FFTW - Precision: double - X Y Z: 256ba1632486480SE +/- 0.12, N = 2SE +/- 0.44, N = 272.2072.291. (CXX) g++ options: -O3

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Mallocba20M40M60M80M100MSE +/- 83929.32, N = 2SE +/- 129754.02, N = 299251227.2899373474.311. (CXX) g++ options: -O2 -std=gnu99 -lc

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Hashba1.2M2.4M3.6M4.8M6MSE +/- 2865.25, N = 2SE +/- 3166.95, N = 25583978.145577252.321. (CXX) g++ options: -O2 -std=gnu99 -lc

Neural Magic DeepSparse

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-Streamba20406080100SE +/- 0.04, N = 2SE +/- 0.03, N = 276.4776.56

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 3.1X Y Z: 144 144 144 - RT: 60ba612182430SE +/- 0.06, N = 2SE +/- 0.01, N = 227.3927.421. (CXX) g++ options: -O3 -ffast-math -ftree-vectorize -lmpi_cxx -lmpi

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Scenario: Asynchronous Multi-Streamba2004006008001000SE +/- 1.01, N = 2SE +/- 0.57, N = 21075.961074.82

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: BERT-Large, NLP Question Answering, Sparse INT8 - Scenario: Asynchronous Multi-Streamba110220330440550SE +/- 0.12, N = 2SE +/- 0.18, N = 2505.13504.61

Timed LLVM Compilation

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

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed LLVM Compilation 16.0Build System: Ninjaba60120180240300SE +/- 0.15, N = 2SE +/- 0.15, N = 2262.88263.15

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Pthreadba30K60K90K120K150KSE +/- 102.07, N = 2SE +/- 971.78, N = 2136709.81136846.011. (CXX) g++ options: -O2 -std=gnu99 -lc

Blender

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.6Blend File: Fishy Cat - Compute: CPU-Onlyba1428425670SE +/- 0.20, N = 2SE +/- 0.08, N = 264.0164.07

HeFFTe - Highly Efficient FFT for Exascale

HeFFTe is the Highly Efficient FFT for Exascale software developed as part of the Exascale Computing Project. This test profile uses HeFFTe's built-in speed benchmarks under a variety of configuration options and currently catering to CPU/processor testing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: c2c - Backend: FFTW - Precision: double - X Y Z: 512ba1020304050SE +/- 0.02, N = 2SE +/- 0.04, N = 244.0143.971. (CXX) g++ options: -O3

OpenFOAM

OpenFOAM is the leading free, open-source software for computational fluid dynamics (CFD). This test profile currently uses the drivaerFastback test case for analyzing automotive aerodynamics or alternatively the older motorBike input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 10Input: drivaerFastback, Medium Mesh Size - Execution Timeba130260390520650SE +/- 0.03, N = 2SE +/- 0.42, N = 2615.46615.991. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfiniteVolume -lmeshTools -lparallel -llagrangian -lregionModels -lgenericPatchFields -lOpenFOAM -ldl -lm

Neural Magic DeepSparse

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Scenario: Asynchronous Multi-Streamba48121620SE +/- 0.01, N = 2SE +/- 0.01, N = 214.8514.86

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: BERT-Large, NLP Question Answering, Sparse INT8 - Scenario: Asynchronous Multi-Streamba714212835SE +/- 0.01, N = 2SE +/- 0.01, N = 231.6531.68

Timed PHP Compilation

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

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed PHP Compilation 8.1.9Time To Compileba1020304050SE +/- 0.48, N = 2SE +/- 0.34, N = 242.3842.35

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: MEMFDba120240360480600SE +/- 1.20, N = 2SE +/- 1.31, N = 2549.55549.941. (CXX) g++ options: -O2 -std=gnu99 -lc

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 32 - Buffer Length: 256 - Filter Length: 32ba200M400M600M800M1000MSE +/- 85000.00, N = 2SE +/- 25000.00, N = 28476750008470850001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Context Switchingba600K1200K1800K2400K3000KSE +/- 604.17, N = 2SE +/- 678.57, N = 22571092.692572801.751. (CXX) g++ options: -O2 -std=gnu99 -lc

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Streamba816243240SE +/- 0.12, N = 2SE +/- 0.06, N = 234.5534.53

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Streamba100200300400500SE +/- 0.02, N = 2SE +/- 0.05, N = 2479.22478.91

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Streamba816243240SE +/- 0.00, N = 2SE +/- 0.00, N = 233.3733.39

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Pollba800K1600K2400K3200K4000KSE +/- 1953.54, N = 2SE +/- 2536.76, N = 23671617.973669281.691. (CXX) g++ options: -O2 -std=gnu99 -lc

VVenC

VVenC is the Fraunhofer Versatile Video Encoder as a fast/efficient H.266/VVC encoder. The vvenc encoder makes use of SIMD Everywhere (SIMDe). The vvenc software is published under the Clear BSD License. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterVVenC 1.9Video Input: Bosphorus 1080p - Video Preset: Fasterba714212835SE +/- 0.04, N = 2SE +/- 0.06, N = 230.9330.951. (CXX) g++ options: -O3 -flto -fno-fat-lto-objects -flto=auto

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Memory Copyingba15003000450060007500SE +/- 11.04, N = 2SE +/- 8.71, N = 27180.437176.191. (CXX) g++ options: -O2 -std=gnu99 -lc

OpenFOAM

OpenFOAM is the leading free, open-source software for computational fluid dynamics (CFD). This test profile currently uses the drivaerFastback test case for analyzing automotive aerodynamics or alternatively the older motorBike input. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgSeconds, Fewer Is BetterOpenFOAM 10Input: drivaerFastback, Small Mesh Size - Mesh Timeba714212835SE +/- 0.05, N = 2SE +/- 0.02, N = 227.9527.971. (CXX) g++ options: -std=c++14 -m64 -O3 -ftemplate-depth-100 -fPIC -fuse-ld=bfd -Xlinker --add-needed --no-as-needed -lfiniteVolume -lmeshTools -lparallel -llagrangian -lregionModels -lgenericPatchFields -lOpenFOAM -ldl -lm

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Matrix 3D Mathba2K4K6K8K10KSE +/- 4.08, N = 2SE +/- 34.45, N = 29605.309599.931. (CXX) g++ options: -O2 -std=gnu99 -lc

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Forkingba20K40K60K80K100KSE +/- 421.24, N = 2SE +/- 469.20, N = 289966.2989918.211. (CXX) g++ options: -O2 -std=gnu99 -lc

Neural Magic DeepSparse

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.5Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-Streamba50100150200250SE +/- 0.05, N = 2SE +/- 0.10, N = 2208.99208.90

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Glibc Qsort Data Sortingba150300450600750SE +/- 0.46, N = 2SE +/- 0.40, N = 2696.92696.651. (CXX) g++ options: -O2 -std=gnu99 -lc

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Zlibba6001200180024003000SE +/- 0.65, N = 2SE +/- 0.06, N = 22648.812647.811. (CXX) g++ options: -O2 -std=gnu99 -lc

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: System V Message Passingba1.3M2.6M3.9M5.2M6.5MSE +/- 9802.94, N = 2SE +/- 7174.98, N = 25854201.785852281.711. (CXX) g++ options: -O2 -std=gnu99 -lc

Blender

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.6Blend File: Barbershop - Compute: CPU-Onlyba110220330440550SE +/- 0.42, N = 2SE +/- 0.22, N = 2493.61493.45

Timed Linux Kernel Compilation

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

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 6.1Build: defconfigba918273645SE +/- 0.69, N = 2SE +/- 0.72, N = 240.4540.44

HeFFTe - Highly Efficient FFT for Exascale

HeFFTe is the Highly Efficient FFT for Exascale software developed as part of the Exascale Computing Project. This test profile uses HeFFTe's built-in speed benchmarks under a variety of configuration options and currently catering to CPU/processor testing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: c2c - Backend: Stock - Precision: float - X Y Z: 512ba1632486480SE +/- 0.00, N = 2SE +/- 0.21, N = 272.5472.561. (CXX) g++ options: -O3

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Vector Mathba30K60K90K120K150KSE +/- 5.98, N = 2SE +/- 47.16, N = 2151431.15151386.311. (CXX) g++ options: -O2 -std=gnu99 -lc

Liquid-DSP

LiquidSDR's Liquid-DSP is a software-defined radio (SDR) digital signal processing library. This test profile runs a multi-threaded benchmark of this SDR/DSP library focused on embedded platform usage. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 64 - Buffer Length: 256 - Filter Length: 32ba300M600M900M1200M1500MSE +/- 450000.00, N = 2SE +/- 300000.00, N = 2157685000015773000001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

OpenBenchmarking.orgsamples/s, More Is BetterLiquid-DSP 1.6Threads: 64 - Buffer Length: 256 - Filter Length: 512ba110M220M330M440M550MSE +/- 800000.00, N = 2SE +/- 385000.00, N = 25130400005131350001. (CC) gcc options: -O3 -pthread -lm -lc -lliquid

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Vector Floating Pointba12K24K36K48K60KSE +/- 4.11, N = 2SE +/- 30.71, N = 258232.7058243.381. (CXX) g++ options: -O2 -std=gnu99 -lc

Blender

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.6Blend File: Classroom - Compute: CPU-Onlyba306090120150SE +/- 0.13, N = 2SE +/- 0.05, N = 2127.76127.78

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: CPU Stressba14K28K42K56K70KSE +/- 38.95, N = 2SE +/- 12.73, N = 264118.8764111.111. (CXX) g++ options: -O2 -std=gnu99 -lc

HeFFTe - Highly Efficient FFT for Exascale

HeFFTe is the Highly Efficient FFT for Exascale software developed as part of the Exascale Computing Project. This test profile uses HeFFTe's built-in speed benchmarks under a variety of configuration options and currently catering to CPU/processor testing. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgGFLOP/s, More Is BetterHeFFTe - Highly Efficient FFT for Exascale 2.3Test: r2c - Backend: Stock - Precision: double - X Y Z: 512ba20406080100SE +/- 0.11, N = 2SE +/- 0.01, N = 276.6076.611. (CXX) g++ options: -O3

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Cryptoba11K22K33K44K55KSE +/- 18.13, N = 2SE +/- 3.65, N = 250243.4850240.091. (CXX) g++ options: -O2 -std=gnu99 -lc

Neural Magic DeepSparse

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.5Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Streamba100200300400500SE +/- 2.44, N = 2SE +/- 0.42, N = 2460.76460.78

Stress-NG

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

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: x86_64 RdRandba70K140K210K280K350KSE +/- 1.14, N = 2SE +/- 2.35, N = 2331423.04331416.521. (CXX) g++ options: -O2 -std=gnu99 -lc

OpenBenchmarking.orgBogo Ops/s, More Is BetterStress-NG 0.15.10Test: Vector Shuffleba40K80K120K160K200KSE +/- 6.04, N = 2SE +/- 6.63, N = 2167202.07167204.211. (CXX) g++ options: -O2 -std=gnu99 -lc

Timed Linux Kernel Compilation

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

OpenBenchmarking.orgSeconds, Fewer Is BetterTimed Linux Kernel Compilation 6.1Build: allmodconfigba100200300400500SE +/- 1.13, N = 2SE +/- 1.46, N = 2445.38445.39

BRL-CAD

BRL-CAD is a cross-platform, open-source solid modeling system with built-in benchmark mode. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgVGR Performance Metric, More Is BetterBRL-CAD 7.36VGR Performance Metrica100K200K300K400K500KSE +/- 3768.50, N = 24666861. (CXX) g++ options: -std=c++14 -pipe -fvisibility=hidden -fno-strict-aliasing -fno-common -fexceptions -ftemplate-depth-128 -m64 -ggdb3 -O3 -fipa-pta -fstrength-reduce -finline-functions -flto -ltcl8.6 -lregex_brl -lz_brl -lnetpbm -ldl -lm -ltk8.6

Apache Cassandra

This is a benchmark of the Apache Cassandra NoSQL database management system making use of cassandra-stress. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgOp/s, More Is BetterApache Cassandra 4.1.3Test: Writesa30K60K90K120K150KSE +/- 803.50, N = 2155626

Blender

OpenBenchmarking.orgSeconds, Fewer Is BetterBlender 3.6Blend File: Pabellon Barcelona - Compute: CPU-Onlya4080120160200SE +/- 0.04, N = 2159.94

Redis 7.0.12 + memtier_benchmark

Memtier_benchmark is a NoSQL Redis/Memcache traffic generation plus benchmarking tool developed by Redis Labs. Learn more via the OpenBenchmarking.org test page.

Protocol: Redis - Clients: 500 - Set To Get Ratio: 1:10

a: The test run did not produce a result.

b: The test run did not produce a result.

Protocol: Redis - Clients: 500 - Set To Get Ratio: 1:5

a: The test run did not produce a result.

b: The test run did not produce a result.

164 Results Shown

Apache IoTDB:
  100 - 100 - 200:
    Average Latency
    point/sec
Stress-NG
libxsmm
Apache IoTDB:
  200 - 100 - 200
  100 - 100 - 500
  500 - 1 - 500
HeFFTe - Highly Efficient FFT for Exascale
Neural Magic DeepSparse
Redis 7.0.12 + memtier_benchmark
Apache IoTDB
Neural Magic DeepSparse
Apache IoTDB:
  100 - 100 - 500
  500 - 1 - 500
Stress-NG
Neural Magic DeepSparse:
  NLP Text Classification, BERT base uncased SST2 - Asynchronous Multi-Stream:
    items/sec
    ms/batch
Apache IoTDB
HeFFTe - Highly Efficient FFT for Exascale
Apache IoTDB
HeFFTe - Highly Efficient FFT for Exascale
Stress-NG
srsRAN Project
Stress-NG
HeFFTe - Highly Efficient FFT for Exascale
Apache IoTDB:
  100 - 1 - 200
  200 - 1 - 200
Stress-NG
Redis 7.0.12 + memtier_benchmark
Stress-NG
Apache IoTDB:
  500 - 100 - 500
  200 - 100 - 500
  200 - 1 - 500
  200 - 100 - 500
Liquid-DSP
Apache IoTDB
Neural Magic DeepSparse:
  NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Stream:
    items/sec
    ms/batch
srsRAN Project
Stress-NG
Liquid-DSP
HeFFTe - Highly Efficient FFT for Exascale
Neural Magic DeepSparse:
  NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Stream:
    items/sec
    ms/batch
Apache IoTDB
Stress-NG
Apache IoTDB
Liquid-DSP
VVenC
Timed LLVM Compilation
HeFFTe - Highly Efficient FFT for Exascale
Neural Magic DeepSparse:
  CV Detection, YOLOv5s COCO, Sparse INT8 - Asynchronous Multi-Stream:
    items/sec
    ms/batch
libxsmm
HeFFTe - Highly Efficient FFT for Exascale
libxsmm
Redis 7.0.12 + memtier_benchmark
HeFFTe - Highly Efficient FFT for Exascale
VVenC
Stress-NG:
  Atomic
  Semaphores
HeFFTe - Highly Efficient FFT for Exascale
libxsmm
srsRAN Project
Apache IoTDB
Stress-NG
HeFFTe - Highly Efficient FFT for Exascale
Palabos
Apache IoTDB
HeFFTe - Highly Efficient FFT for Exascale:
  c2c - FFTW - float - 128
  r2c - FFTW - float - 128
Laghos
Apache IoTDB
Neural Magic DeepSparse:
  CV Segmentation, 90% Pruned YOLACT Pruned - Asynchronous Multi-Stream:
    ms/batch
    items/sec
Stress-NG
High Performance Conjugate Gradient
Stress-NG
Apache IoTDB
HeFFTe - Highly Efficient FFT for Exascale
Liquid-DSP
Stress-NG:
  NUMA
  Mutex
HeFFTe - Highly Efficient FFT for Exascale
Neural Magic DeepSparse:
  NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Asynchronous Multi-Stream:
    ms/batch
    items/sec
Stress-NG
Apache IoTDB
Liquid-DSP
Neural Magic DeepSparse:
  NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Asynchronous Multi-Stream:
    items/sec
    ms/batch
VVenC
Redis 7.0.12 + memtier_benchmark
Timed GDB GNU Debugger Compilation
Stress-NG
High Performance Conjugate Gradient
HeFFTe - Highly Efficient FFT for Exascale
OpenFOAM
Neural Magic DeepSparse:
  ResNet-50, Sparse INT8 - Asynchronous Multi-Stream:
    items/sec
    ms/batch
HeFFTe - Highly Efficient FFT for Exascale
Palabos
HeFFTe - Highly Efficient FFT for Exascale:
  r2c - Stock - double - 256
  c2c - FFTW - float - 512
OpenFOAM
HeFFTe - Highly Efficient FFT for Exascale
Neural Magic DeepSparse
Laghos
Neural Magic DeepSparse
Blender
HeFFTe - Highly Efficient FFT for Exascale
Stress-NG
Palabos
Stress-NG
Liquid-DSP
HeFFTe - Highly Efficient FFT for Exascale
Stress-NG:
  Malloc
  Hash
Neural Magic DeepSparse
High Performance Conjugate Gradient
Neural Magic DeepSparse:
  NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Asynchronous Multi-Stream
  BERT-Large, NLP Question Answering, Sparse INT8 - Asynchronous Multi-Stream
Timed LLVM Compilation
Stress-NG
Blender
HeFFTe - Highly Efficient FFT for Exascale
OpenFOAM
Neural Magic DeepSparse:
  NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Asynchronous Multi-Stream
  BERT-Large, NLP Question Answering, Sparse INT8 - Asynchronous Multi-Stream
Timed PHP Compilation
Stress-NG
Liquid-DSP
Stress-NG
Neural Magic DeepSparse:
  NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Stream
  CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Stream
  CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Stream
Stress-NG
VVenC
Stress-NG
OpenFOAM
Stress-NG:
  Matrix 3D Math
  Forking
Neural Magic DeepSparse
Stress-NG:
  Glibc Qsort Data Sorting
  Zlib
  System V Message Passing
Blender
Timed Linux Kernel Compilation
HeFFTe - Highly Efficient FFT for Exascale
Stress-NG
Liquid-DSP:
  64 - 256 - 32
  64 - 256 - 512
Stress-NG
Blender
Stress-NG
HeFFTe - Highly Efficient FFT for Exascale
Stress-NG
Neural Magic DeepSparse
Stress-NG:
  x86_64 RdRand
  Vector Shuffle
Timed Linux Kernel Compilation
BRL-CAD
Apache Cassandra
Blender