Benchmarks for a future article.
Linux w DODT Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-12-Wbc0TK/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-Wbc0TK/gcc-12-12.2.0/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-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0xd000375Python Notes: Python 3.10.6Security Notes: dodt: Mitigation of DOITM + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Mitigation of Clear buffers; SMT vulnerable + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
doitm=off Processor: 2 x Intel Xeon Platinum 8380 @ 3.40GHz (80 Cores / 160 Threads), Motherboard: Intel M50CYP2SB2U (SE5C6200.86B.0022.D08.2103221623 BIOS), Chipset: Intel Ice Lake IEH, Memory: 512GB, Disk: 7682GB INTEL SSDPF2KX076TZ, Graphics: ASPEED, Monitor: VE228, Network: 2 x Intel X710 for 10GBASE-T + 2 x Intel E810-C for QSFP
OS: Ubuntu 22.10, Kernel: 6.2.0-rc5-phx-dodt (x86_64), Desktop: GNOME Shell, Display Server: X Server 1.21.1.3, Vulkan: 1.3.211, Compiler: GCC 12.2.0, File-System: ext4, Screen Resolution: 1920x1080
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-12-Wbc0TK/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-Wbc0TK/gcc-12-12.2.0/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-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0xd000375Python Notes: Python 3.10.6Security Notes: dodt: Vulnerable + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Mitigation of Clear buffers; SMT vulnerable + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
Xeon Platinum 8380 DODT Mitigation Impact OpenBenchmarking.org Phoronix Test Suite 2 x Intel Xeon Platinum 8380 @ 3.40GHz (80 Cores / 160 Threads) Intel M50CYP2SB2U (SE5C6200.86B.0022.D08.2103221623 BIOS) Intel Ice Lake IEH 512GB 7682GB INTEL SSDPF2KX076TZ ASPEED VE228 2 x Intel X710 for 10GBASE-T + 2 x Intel E810-C for QSFP Ubuntu 22.10 6.2.0-rc5-phx-dodt (x86_64) GNOME Shell X Server 1.21.1.3 1.3.211 GCC 12.2.0 ext4 1920x1080 Processor Motherboard Chipset Memory Disk Graphics Monitor Network OS Kernel Desktop Display Server Vulkan Compiler File-System Screen Resolution Xeon Platinum 8380 DODT Mitigation Impact Benchmarks System Logs - Transparent Huge Pages: madvise - --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-12-Wbc0TK/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-Wbc0TK/gcc-12-12.2.0/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-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: 0xd000375 - Python 3.10.6 - Linux w DODT: dodt: Mitigation of DOITM + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Mitigation of Clear buffers; SMT vulnerable + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected - doitm=off: dodt: Vulnerable + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Mitigation of Clear buffers; SMT vulnerable + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
OpenBenchmarking.org MiB/second, More Is Better Crypto++ 8.2 Test: Unkeyed Algorithms doitm=off Linux w DODT 80 160 240 320 400 SE +/- 0.02, N = 3 SE +/- 0.04, N = 3 361.88 361.67 1. (CXX) g++ options: -g2 -O3 -fPIC -pthread -pipe
miniBUDE MiniBUDE is a mini application for the the core computation of the Bristol University Docking Engine (BUDE). This test profile currently makes use of the OpenMP implementation of miniBUDE for CPU benchmarking. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org GFInst/s, More Is Better miniBUDE 20210901 Implementation: OpenMP - Input Deck: BM1 doitm=off Linux w DODT 500 1000 1500 2000 2500 SE +/- 10.72, N = 3 SE +/- 17.94, N = 3 2373.49 2368.12 1. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm
OpenBenchmarking.org Billion Interactions/s, More Is Better miniBUDE 20210901 Implementation: OpenMP - Input Deck: BM1 doitm=off Linux w DODT 20 40 60 80 100 SE +/- 0.43, N = 3 SE +/- 0.72, N = 3 94.94 94.73 1. (CC) gcc options: -std=c99 -Ofast -ffast-math -fopenmp -march=native -lm
nekRS nekRS is an open-source Navier Stokes solver based on the spectral element method. NekRS supports both CPU and GPU/accelerator support though this test profile is currently configured for CPU execution. NekRS is part of Nek5000 of the Mathematics and Computer Science MCS at Argonne National Laboratory. This nekRS benchmark is primarily relevant to large core count HPC servers and otherwise may be very time consuming. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org FLOP/s, More Is Better nekRS 22.0 Input: TurboPipe Periodic doitm=off Linux w DODT 60000M 120000M 180000M 240000M 300000M SE +/- 1474277940.03, N = 3 SE +/- 500223394.54, N = 3 298310333333 295621666667 1. (CXX) g++ options: -fopenmp -O2 -march=native -mtune=native -ftree-vectorize -lmpi_cxx -lmpi
OpenRadioss OpenRadioss is an open-source AGPL-licensed finite element solver for dynamic event analysis OpenRadioss is based on Altair Radioss and open-sourced in 2022. This open-source finite element solver is benchmarked with various example models available from https://www.openradioss.org/models/. This test is currently using a reference OpenRadioss binary build offered via GitHub. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better OpenRadioss 2022.10.13 Model: Bumper Beam doitm=off Linux w DODT 20 40 60 80 100 SE +/- 0.06, N = 3 SE +/- 0.28, N = 3 84.63 85.13
Xmrig Xmrig is an open-source cross-platform CPU/GPU miner for RandomX, KawPow, CryptoNight and AstroBWT. This test profile is setup to measure the Xmlrig CPU mining performance. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org H/s, More Is Better Xmrig 6.18.1 Variant: Monero - Hash Count: 1M doitm=off Linux w DODT 6K 12K 18K 24K 30K SE +/- 48.01, N = 3 SE +/- 76.73, N = 3 26578.0 26451.3 1. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc
OpenBenchmarking.org H/s, More Is Better Xmrig 6.18.1 Variant: Wownero - Hash Count: 1M doitm=off Linux w DODT 9K 18K 27K 36K 45K SE +/- 35.05, N = 3 SE +/- 149.86, N = 3 41317.2 41450.1 1. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc
OpenBenchmarking.org MP/s, More Is Better WebP Image Encode 1.2.4 Encode Settings: Quality 100, Highest Compression doitm=off Linux w DODT 0.6323 1.2646 1.8969 2.5292 3.1615 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 2.81 2.81 1. (CC) gcc options: -fvisibility=hidden -O2 -lm
Kvazaar This is a test of Kvazaar as a CPU-based H.265/HEVC video encoder written in the C programming language and optimized in Assembly. Kvazaar is the winner of the 2016 ACM Open-Source Software Competition and developed at the Ultra Video Group, Tampere University, Finland. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Frames Per Second, More Is Better Kvazaar 2.2 Video Input: Bosphorus 4K - Video Preset: Super Fast doitm=off Linux w DODT 11 22 33 44 55 SE +/- 0.63, N = 3 SE +/- 0.46, N = 15 46.47 47.06 1. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.org Frames Per Second, More Is Better Kvazaar 2.2 Video Input: Bosphorus 4K - Video Preset: Ultra Fast doitm=off Linux w DODT 11 22 33 44 55 SE +/- 0.60, N = 3 SE +/- 0.37, N = 10 47.75 47.95 1. (CC) gcc options: -pthread -ftree-vectorize -fvisibility=hidden -O2 -lpthread -lm -lrt
OpenBenchmarking.org Frames Per Second, More Is Better uvg266 0.4.1 Video Input: Bosphorus 4K - Video Preset: Ultra Fast doitm=off Linux w DODT 10 20 30 40 50 SE +/- 0.18, N = 3 SE +/- 0.45, N = 3 42.62 42.61
OpenBenchmarking.org Seconds, Fewer Is Better libavif avifenc 0.11 Encoder Speed: 6 doitm=off Linux w DODT 0.7918 1.5836 2.3754 3.1672 3.959 SE +/- 0.014, N = 3 SE +/- 0.031, N = 15 3.499 3.519 1. (CXX) g++ options: -O3 -fPIC -lm
OpenSSL OpenSSL is an open-source toolkit that implements SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. The system/openssl test profiles relies on benchmarking the system/OS-supplied openssl binary rather than the pts/openssl test profile that uses the locally-built OpenSSL for benchmarking. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org sign/s, More Is Better OpenSSL doitm=off Linux w DODT 4K 8K 12K 16K 20K SE +/- 20.72, N = 3 SE +/- 63.70, N = 3 17837.9 17781.6 1. OpenSSL 3.0.5 5 Jul 2022 (Library: OpenSSL 3.0.5 5 Jul 2022)
OpenBenchmarking.org verify/s, More Is Better OpenSSL doitm=off Linux w DODT 300K 600K 900K 1200K 1500K SE +/- 1026.50, N = 3 SE +/- 1656.25, N = 3 1188684.8 1188255.4 1. OpenSSL 3.0.5 5 Jul 2022 (Library: OpenSSL 3.0.5 5 Jul 2022)
ClickHouse ClickHouse is an open-source, high performance OLAP data management system. This test profile uses ClickHouse's standard benchmark recommendations per https://clickhouse.com/docs/en/operations/performance-test/ / https://github.com/ClickHouse/ClickBench/tree/main/clickhouse with the 100 million rows web analytics dataset. The reported value is the query processing time using the geometric mean of all separate queries performed as an aggregate. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Queries Per Minute, Geo Mean, More Is Better ClickHouse 22.12.3.5 100M Rows Hits Dataset, First Run / Cold Cache doitm=off Linux w DODT 90 180 270 360 450 SE +/- 1.76, N = 3 SE +/- 3.64, N = 3 392.39 398.75 MIN: 33.75 / MAX: 5000 MIN: 34.82 / MAX: 4615.38
OpenBenchmarking.org Queries Per Minute, Geo Mean, More Is Better ClickHouse 22.12.3.5 100M Rows Hits Dataset, Second Run doitm=off Linux w DODT 90 180 270 360 450 SE +/- 1.56, N = 3 SE +/- 1.99, N = 3 421.39 423.13 MIN: 34.86 / MAX: 5454.55 MIN: 35.89 / MAX: 5000
OpenBenchmarking.org Queries Per Minute, Geo Mean, More Is Better ClickHouse 22.12.3.5 100M Rows Hits Dataset, Third Run doitm=off Linux w DODT 90 180 270 360 450 SE +/- 0.79, N = 3 SE +/- 3.25, N = 3 424.96 427.42 MIN: 36.19 / MAX: 5454.55 MIN: 35.99 / MAX: 5454.55
OpenBenchmarking.org Iterations Per Second, More Is Better Cryptsetup PBKDF2-whirlpool doitm=off Linux w DODT 120K 240K 360K 480K 600K SE +/- 1624.74, N = 3 SE +/- 1294.67, N = 3 582551 583195
OpenBenchmarking.org MiB/s, More Is Better Cryptsetup AES-XTS 256b Encryption doitm=off Linux w DODT 800 1600 2400 3200 4000 SE +/- 7.45, N = 3 SE +/- 16.70, N = 3 3876.2 3864.3
OpenBenchmarking.org MiB/s, More Is Better Cryptsetup AES-XTS 256b Decryption doitm=off Linux w DODT 800 1600 2400 3200 4000 SE +/- 3.34, N = 3 SE +/- 17.78, N = 3 3886.1 3891.5
OpenBenchmarking.org MiB/s, More Is Better Cryptsetup Serpent-XTS 256b Encryption doitm=off Linux w DODT 120 240 360 480 600 SE +/- 1.35, N = 3 SE +/- 1.27, N = 3 559.1 559.5
OpenBenchmarking.org MiB/s, More Is Better Cryptsetup Serpent-XTS 256b Decryption doitm=off Linux w DODT 110 220 330 440 550 SE +/- 0.09, N = 3 SE +/- 0.27, N = 3 527.3 527.6
OpenBenchmarking.org MiB/s, More Is Better Cryptsetup Twofish-XTS 256b Encryption doitm=off Linux w DODT 80 160 240 320 400 SE +/- 0.88, N = 3 SE +/- 0.92, N = 3 351.5 351.4
OpenBenchmarking.org MiB/s, More Is Better Cryptsetup Twofish-XTS 256b Decryption doitm=off Linux w DODT 80 160 240 320 400 SE +/- 0.21, N = 3 SE +/- 0.17, N = 3 358.5 357.7
OpenBenchmarking.org MiB/s, More Is Better Cryptsetup AES-XTS 512b Encryption doitm=off Linux w DODT 700 1400 2100 2800 3500 SE +/- 3.15, N = 3 SE +/- 13.51, N = 3 3458.6 3445.4
OpenBenchmarking.org MiB/s, More Is Better Cryptsetup AES-XTS 512b Decryption doitm=off Linux w DODT 700 1400 2100 2800 3500 SE +/- 2.54, N = 3 SE +/- 13.35, N = 3 3446.3 3449.8
OpenBenchmarking.org MiB/s, More Is Better Cryptsetup Serpent-XTS 512b Encryption doitm=off Linux w DODT 120 240 360 480 600 SE +/- 0.06, N = 3 SE +/- 0.45, N = 2 560.5 561.0
OpenBenchmarking.org MiB/s, More Is Better Cryptsetup Twofish-XTS 512b Encryption doitm=off Linux w DODT 80 160 240 320 400 SE +/- 0.03, N = 3 SE +/- 0.12, N = 3 352.2 352.2
OpenBenchmarking.org MiB/s, More Is Better Cryptsetup Twofish-XTS 512b Decryption doitm=off Linux w DODT 80 160 240 320 400 SE +/- 0.23, N = 3 SE +/- 0.25, N = 2 358.1 357.5
PostgreSQL This is a benchmark of PostgreSQL using the integrated pgbench for facilitating the database benchmarks. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org TPS, More Is Better PostgreSQL 15 Scaling Factor: 100 - Clients: 1000 - Mode: Read Only doitm=off Linux w DODT 400K 800K 1200K 1600K 2000K SE +/- 80874.14, N = 12 SE +/- 89751.30, N = 12 2020130 2021337 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.org ms, Fewer Is Better PostgreSQL 15 Scaling Factor: 100 - Clients: 1000 - Mode: Read Only - Average Latency doitm=off Linux w DODT 0.1139 0.2278 0.3417 0.4556 0.5695 SE +/- 0.021, N = 12 SE +/- 0.024, N = 12 0.504 0.506 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.org TPS, More Is Better PostgreSQL 15 Scaling Factor: 100 - Clients: 800 - Mode: Read Write doitm=off Linux w DODT 16K 32K 48K 64K 80K SE +/- 295.95, N = 3 SE +/- 197.73, N = 3 75123 75223 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.org ms, Fewer Is Better PostgreSQL 15 Scaling Factor: 100 - Clients: 800 - Mode: Read Write - Average Latency doitm=off Linux w DODT 3 6 9 12 15 SE +/- 0.04, N = 3 SE +/- 0.03, N = 3 10.65 10.64 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.org TPS, More Is Better PostgreSQL 15 Scaling Factor: 100 - Clients: 1000 - Mode: Read Write doitm=off Linux w DODT 15K 30K 45K 60K 75K SE +/- 295.57, N = 3 SE +/- 95.69, N = 3 68055 68328 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.org ms, Fewer Is Better PostgreSQL 15 Scaling Factor: 100 - Clients: 1000 - Mode: Read Write - Average Latency doitm=off Linux w DODT 4 8 12 16 20 SE +/- 0.06, N = 3 SE +/- 0.02, N = 3 14.70 14.64 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
Neural Magic DeepSparse This is a benchmark of Neural Magic's DeepSparse using its built-in deepsparse.benchmark utility and various models from their SparseZoo (https://sparsezoo.neuralmagic.com/). Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.3.2 Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Stream doitm=off Linux w DODT 11 22 33 44 55 SE +/- 0.02, N = 3 SE +/- 0.01, N = 3 48.45 48.51
OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.3.2 Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Stream doitm=off Linux w DODT 200 400 600 800 1000 SE +/- 0.15, N = 3 SE +/- 0.30, N = 3 821.52 823.50
OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.3.2 Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Asynchronous Multi-Stream doitm=off Linux w DODT 200 400 600 800 1000 SE +/- 3.12, N = 3 SE +/- 2.39, N = 3 902.06 890.95
OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.3.2 Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Asynchronous Multi-Stream doitm=off Linux w DODT 10 20 30 40 50 SE +/- 0.16, N = 3 SE +/- 0.12, N = 3 44.31 44.86
OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.3.2 Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-Stream doitm=off Linux w DODT 50 100 150 200 250 SE +/- 2.51, N = 3 SE +/- 0.42, N = 3 218.78 219.31
OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.3.2 Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-Stream doitm=off Linux w DODT 40 80 120 160 200 SE +/- 2.12, N = 3 SE +/- 0.39, N = 3 182.70 182.20
OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.3.2 Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-Stream doitm=off Linux w DODT 50 100 150 200 250 SE +/- 0.28, N = 3 SE +/- 0.20, N = 3 224.05 223.65
OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.3.2 Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-Stream doitm=off Linux w DODT 40 80 120 160 200 SE +/- 0.20, N = 3 SE +/- 0.19, N = 3 178.28 178.62
OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.3.2 Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Stream doitm=off Linux w DODT 11 22 33 44 55 SE +/- 0.03, N = 3 SE +/- 0.18, N = 3 48.49 48.46
OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.3.2 Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Stream doitm=off Linux w DODT 200 400 600 800 1000 SE +/- 2.36, N = 3 SE +/- 0.79, N = 3 818.30 821.21
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.14.06 Test: MEMFD doitm=off Linux w DODT 800 1600 2400 3200 4000 SE +/- 1.55, N = 3 SE +/- 2.24, N = 3 3624.42 3592.88 1. (CC) gcc options: -O2 -std=gnu99 -lm -fuse-ld=gold -lapparmor -latomic -lc -lcrypt -ldl -ljpeg -lrt -lz -pthread
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.14.06 Test: Mutex doitm=off Linux w DODT 9M 18M 27M 36M 45M SE +/- 34997.86, N = 3 SE +/- 162165.00, N = 3 41978074.41 41364067.75 1. (CC) gcc options: -O2 -std=gnu99 -lm -fuse-ld=gold -lapparmor -latomic -lc -lcrypt -ldl -ljpeg -lrt -lz -pthread
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.14.06 Test: Crypto doitm=off Linux w DODT 20K 40K 60K 80K 100K SE +/- 991.74, N = 3 SE +/- 445.46, N = 3 84975.84 85533.02 1. (CC) gcc options: -O2 -std=gnu99 -lm -fuse-ld=gold -lapparmor -latomic -lc -lcrypt -ldl -ljpeg -lrt -lz -pthread
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.14.06 Test: Malloc doitm=off Linux w DODT 40M 80M 120M 160M 200M SE +/- 454146.57, N = 3 SE +/- 655475.74, N = 3 201328390.34 195951844.81 1. (CC) gcc options: -O2 -std=gnu99 -lm -fuse-ld=gold -lapparmor -latomic -lc -lcrypt -ldl -ljpeg -lrt -lz -pthread
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.14.06 Test: IO_uring doitm=off Linux w DODT 6K 12K 18K 24K 30K SE +/- 1.30, N = 3 SE +/- 16.56, N = 3 26532.30 26508.09 1. (CC) gcc options: -O2 -std=gnu99 -lm -fuse-ld=gold -lapparmor -latomic -lc -lcrypt -ldl -ljpeg -lrt -lz -pthread
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.14.06 Test: SENDFILE doitm=off Linux w DODT 200K 400K 600K 800K 1000K SE +/- 8323.86, N = 3 SE +/- 8091.69, N = 3 1146358.16 1128077.91 1. (CC) gcc options: -O2 -std=gnu99 -lm -fuse-ld=gold -lapparmor -latomic -lc -lcrypt -ldl -ljpeg -lrt -lz -pthread
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.14.06 Test: x86_64 RdRand doitm=off Linux w DODT 140K 280K 420K 560K 700K SE +/- 67.49, N = 3 SE +/- 68.80, N = 3 669178.00 669198.10 1. (CC) gcc options: -O2 -std=gnu99 -lm -fuse-ld=gold -lapparmor -latomic -lc -lcrypt -ldl -ljpeg -lrt -lz -pthread
spaCy The spaCy library is an open-source solution for advanced neural language processing (NLP). The spaCy library leverages Python and is a leading neural language processing solution. This test profile times the spaCy CPU performance with various models. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org tokens/sec, More Is Better spaCy 3.4.1 Model: en_core_web_lg doitm=off Linux w DODT 2K 4K 6K 8K 10K SE +/- 69.82, N = 3 SE +/- 10.12, N = 3 10888 10971
OpenBenchmarking.org tokens/sec, More Is Better spaCy 3.4.1 Model: en_core_web_trf doitm=off Linux w DODT 700 1400 2100 2800 3500 SE +/- 48.89, N = 3 SE +/- 30.17, N = 3 3225 3169
Blender Blender is an open-source 3D creation and modeling software project. This test is of Blender's Cycles performance with various sample files. GPU computing via NVIDIA OptiX and NVIDIA CUDA is currently supported as well as HIP for AMD Radeon GPUs and Intel oneAPI for Intel Graphics. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.4 Blend File: BMW27 - Compute: CPU-Only doitm=off Linux w DODT 6 12 18 24 30 SE +/- 0.04, N = 3 SE +/- 0.07, N = 3 23.30 23.44
OpenBenchmarking.org Seconds, Fewer Is Better Blender 3.4 Blend File: Classroom - Compute: CPU-Only doitm=off Linux w DODT 14 28 42 56 70 SE +/- 0.06, N = 3 SE +/- 0.10, N = 3 62.68 62.71
RocksDB This is a benchmark of Meta/Facebook's RocksDB as an embeddable persistent key-value store for fast storage based on Google's LevelDB. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Op/s, More Is Better RocksDB 7.9.2 Test: Random Read doitm=off Linux w DODT 60M 120M 180M 240M 300M SE +/- 2757403.57, N = 3 SE +/- 2251545.82, N = 3 278886510 278841535 1. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread
OpenBenchmarking.org Op/s, More Is Better RocksDB 7.9.2 Test: Update Random doitm=off Linux w DODT 130K 260K 390K 520K 650K SE +/- 3603.95, N = 3 SE +/- 1503.74, N = 3 588787 592576 1. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti -lpthread
Linux w DODT Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-12-Wbc0TK/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-Wbc0TK/gcc-12-12.2.0/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-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0xd000375Python Notes: Python 3.10.6Security Notes: dodt: Mitigation of DOITM + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Mitigation of Clear buffers; SMT vulnerable + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 26 January 2023 14:43 by user phoronix.
doitm=off Processor: 2 x Intel Xeon Platinum 8380 @ 3.40GHz (80 Cores / 160 Threads), Motherboard: Intel M50CYP2SB2U (SE5C6200.86B.0022.D08.2103221623 BIOS), Chipset: Intel Ice Lake IEH, Memory: 512GB, Disk: 7682GB INTEL SSDPF2KX076TZ, Graphics: ASPEED, Monitor: VE228, Network: 2 x Intel X710 for 10GBASE-T + 2 x Intel E810-C for QSFP
OS: Ubuntu 22.10, Kernel: 6.2.0-rc5-phx-dodt (x86_64), Desktop: GNOME Shell, Display Server: X Server 1.21.1.3, Vulkan: 1.3.211, Compiler: GCC 12.2.0, File-System: ext4, Screen Resolution: 1920x1080
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-multilib --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-12-Wbc0TK/gcc-12-12.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-12-Wbc0TK/gcc-12-12.2.0/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-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0xd000375Python Notes: Python 3.10.6Security Notes: dodt: Vulnerable + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Mitigation of Clear buffers; SMT vulnerable + retbleed: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 26 January 2023 21:10 by user phoronix.