net INTEL XEON PLATINUM 8592+ testing with a Quanta Cloud S6Q-MB-MPS (3B05.TEL4P1 BIOS) and ASPEED on Ubuntu 23.10 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 2312187-NE-NET50813920 a Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --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-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-XYspKM/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-XYspKM/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x21000161Java Notes: OpenJDK Runtime Environment (build 11.0.20+8-post-Ubuntu-1ubuntu1)Python Notes: Python 3.11.6Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
b Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --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-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-XYspKM/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-XYspKM/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x21000161Python Notes: Python 3.11.6Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
c Processor: INTEL XEON PLATINUM 8592+ @ 3.90GHz (64 Cores / 128 Threads), Motherboard: Quanta Cloud S6Q-MB-MPS (3B05.TEL4P1 BIOS), Chipset: Intel Device 1bce, Memory: 512GB, Disk: 3201GB Micron_7450_MTFDKCC3T2TFS + 0GB Virtual HDisk0 + 0GB Virtual HDisk2 + 0GB Virtual HDisk1 + 0GB Virtual HDisk3, Graphics: ASPEED
OS: Ubuntu 23.10, Kernel: 6.6.0-rc5-phx-patched (x86_64), Desktop: GNOME Shell 45.0, Display Server: X Server 1.21.1.7, Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1200
net OpenBenchmarking.org Phoronix Test Suite INTEL XEON PLATINUM 8592+ @ 3.90GHz (64 Cores / 128 Threads) Quanta Cloud S6Q-MB-MPS (3B05.TEL4P1 BIOS) Intel Device 1bce 512GB 3201GB Micron_7450_MTFDKCC3T2TFS + 0GB Virtual HDisk0 + 0GB Virtual HDisk2 + 0GB Virtual HDisk1 + 0GB Virtual HDisk3 ASPEED Ubuntu 23.10 6.6.0-rc5-phx-patched (x86_64) GNOME Shell 45.0 X Server 1.21.1.7 GCC 13.2.0 ext4 1920x1200 Processor Motherboard Chipset Memory Disk Graphics OS Kernel Desktop Display Server Compiler File-System Screen Resolution Net Benchmarks System Logs - Transparent Huge Pages: madvise - --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --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-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-XYspKM/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-XYspKM/gcc-13-13.2.0/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: 0x21000161 - a: OpenJDK Runtime Environment (build 11.0.20+8-post-Ubuntu-1ubuntu1) - Python 3.11.6 - gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
a b c Result Overview Phoronix Test Suite 100% 101% 101% 102% 102% LeelaChessZero SVT-AV1 Neural Magic DeepSparse Xmrig
net svt-av1: Preset 4 - Bosphorus 4K svt-av1: Preset 8 - Bosphorus 4K svt-av1: Preset 12 - Bosphorus 4K svt-av1: Preset 13 - Bosphorus 4K svt-av1: Preset 4 - Bosphorus 1080p svt-av1: Preset 8 - Bosphorus 1080p svt-av1: Preset 12 - Bosphorus 1080p svt-av1: Preset 13 - Bosphorus 1080p xmrig: KawPow - 1M xmrig: Monero - 1M xmrig: Wownero - 1M xmrig: GhostRider - 1M xmrig: CryptoNight-Heavy - 1M xmrig: CryptoNight-Femto UPX2 - 1M deepsparse: NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Stream deepsparse: NLP Document Classification, oBERT base uncased on IMDB - Synchronous Single-Stream deepsparse: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Asynchronous Multi-Stream deepsparse: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Synchronous Single-Stream deepsparse: ResNet-50, Baseline - Asynchronous Multi-Stream deepsparse: ResNet-50, Baseline - Synchronous Single-Stream deepsparse: ResNet-50, Sparse INT8 - Asynchronous Multi-Stream deepsparse: ResNet-50, Sparse INT8 - Synchronous Single-Stream deepsparse: CV Detection, YOLOv5s COCO - Asynchronous Multi-Stream deepsparse: CV Detection, YOLOv5s COCO - Synchronous Single-Stream deepsparse: BERT-Large, NLP Question Answering - Asynchronous Multi-Stream deepsparse: BERT-Large, NLP Question Answering - Synchronous Single-Stream deepsparse: CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Stream deepsparse: CV Classification, ResNet-50 ImageNet - Synchronous Single-Stream deepsparse: CV Detection, YOLOv5s COCO, Sparse INT8 - Asynchronous Multi-Stream deepsparse: CV Detection, YOLOv5s COCO, Sparse INT8 - Synchronous Single-Stream deepsparse: NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Stream deepsparse: NLP Text Classification, DistilBERT mnli - Synchronous Single-Stream deepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Asynchronous Multi-Stream deepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Synchronous Single-Stream deepsparse: BERT-Large, NLP Question Answering, Sparse INT8 - Asynchronous Multi-Stream deepsparse: BERT-Large, NLP Question Answering, Sparse INT8 - Synchronous Single-Stream deepsparse: NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Stream deepsparse: NLP Token Classification, BERT base uncased conll2003 - Synchronous Single-Stream lczero: BLAS lczero: Eigen deepsparse: NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Stream deepsparse: NLP Document Classification, oBERT base uncased on IMDB - Synchronous Single-Stream deepsparse: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Asynchronous Multi-Stream deepsparse: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Synchronous Single-Stream deepsparse: ResNet-50, Baseline - Asynchronous Multi-Stream deepsparse: ResNet-50, Baseline - Synchronous Single-Stream deepsparse: ResNet-50, Sparse INT8 - Asynchronous Multi-Stream deepsparse: ResNet-50, Sparse INT8 - Synchronous Single-Stream deepsparse: CV Detection, YOLOv5s COCO - Asynchronous Multi-Stream deepsparse: CV Detection, YOLOv5s COCO - Synchronous Single-Stream deepsparse: BERT-Large, NLP Question Answering - Asynchronous Multi-Stream deepsparse: BERT-Large, NLP Question Answering - Synchronous Single-Stream deepsparse: CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Stream deepsparse: CV Classification, ResNet-50 ImageNet - Synchronous Single-Stream deepsparse: CV Detection, YOLOv5s COCO, Sparse INT8 - Asynchronous Multi-Stream deepsparse: CV Detection, YOLOv5s COCO, Sparse INT8 - Synchronous Single-Stream deepsparse: NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Stream deepsparse: NLP Text Classification, DistilBERT mnli - Synchronous Single-Stream deepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Asynchronous Multi-Stream deepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Synchronous Single-Stream deepsparse: BERT-Large, NLP Question Answering, Sparse INT8 - Asynchronous Multi-Stream deepsparse: BERT-Large, NLP Question Answering, Sparse INT8 - Synchronous Single-Stream deepsparse: NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Stream deepsparse: NLP Token Classification, BERT base uncased conll2003 - Synchronous Single-Stream a b c 7.156 75.139 228.874 227.285 20.770 139.811 545.910 632.459 39936.2 40329.6 43088.3 6862.9 40394.9 40425.9 71.9864 34.4749 2323.9799 210.9638 934.6524 286.8882 5706.6964 701.7915 436.5226 164.9393 86.3024 31.2246 933.0074 290.0133 452.1951 166.6567 641.1047 186.0257 96.1343 35.1023 938.0661 67.8694 72.0462 34.6092 67 796 442.3119 29.0016 13.7504 4.7354 34.2126 3.4825 5.5953 1.4200 73.2660 6.0577 370.7287 32.0197 34.2750 3.4452 70.7373 5.9970 49.8955 5.3729 332.8063 28.4611 34.0725 14.7276 442.2210 28.8890 7.168 75.737 229.514 229.556 20.587 141.977 543.725 636.909 40344.7 40153.2 42998.7 6857.8 40282.2 40080.4 71.6759 34.4826 2324.7362 210.8100 928.4462 290.2826 5717.5185 705.8268 436.3286 164.8968 86.2914 31.1465 934.5224 288.6849 452.1641 166.5123 637.6092 188.5376 95.6329 34.6174 937.5775 68.1760 72.0875 34.6250 67 805 443.3971 28.9953 13.7462 4.7386 34.4463 3.4417 5.5844 1.4123 73.2935 6.0589 370.7761 32.0999 34.2173 3.4609 70.7412 6.0023 50.1708 5.3005 334.3153 28.8654 34.0905 14.6612 441.4776 28.8756 7.151 74.926 228.518 229.075 20.492 141.873 546.877 637.337 40242.5 40072.2 43080.4 6820.8 40286.6 40331.1 71.9685 34.2377 2322.5803 209.6095 939.5363 287.3469 5704.9530 701.7061 435.4932 163.9718 86.1709 31.1875 940.2198 289.8081 450.4918 166.3050 637.4175 188.2324 96.8521 33.7246 937.8630 67.5802 71.7827 34.4526 68 817 442.7525 29.2030 13.7597 4.7654 34.0380 3.4767 5.5965 1.4205 73.4267 6.0922 371.2957 32.0585 34.0080 3.4473 71.0046 6.0096 50.1864 5.3092 330.3352 29.6799 34.0792 14.7898 443.1582 29.0205 OpenBenchmarking.org
SVT-AV1 This is a benchmark of the SVT-AV1 open-source video encoder/decoder. SVT-AV1 was originally developed by Intel as part of their Open Visual Cloud / Scalable Video Technology (SVT). Development of SVT-AV1 has since moved to the Alliance for Open Media as part of upstream AV1 development. SVT-AV1 is a CPU-based multi-threaded video encoder for the AV1 video format with a sample YUV video file. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 1.8 Encoder Mode: Preset 4 - Input: Bosphorus 4K c a b 2 4 6 8 10 SE +/- 0.071, N = 3 SE +/- 0.022, N = 3 SE +/- 0.025, N = 3 7.151 7.156 7.168 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 1.8 Encoder Mode: Preset 8 - Input: Bosphorus 4K c a b 20 40 60 80 100 SE +/- 0.15, N = 3 SE +/- 0.38, N = 3 SE +/- 0.27, N = 3 74.93 75.14 75.74 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 1.8 Encoder Mode: Preset 12 - Input: Bosphorus 4K c a b 50 100 150 200 250 SE +/- 1.46, N = 3 SE +/- 0.42, N = 3 SE +/- 1.07, N = 3 228.52 228.87 229.51 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 1.8 Encoder Mode: Preset 13 - Input: Bosphorus 4K a c b 50 100 150 200 250 SE +/- 2.23, N = 3 SE +/- 0.19, N = 3 SE +/- 0.75, N = 3 227.29 229.08 229.56 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 1.8 Encoder Mode: Preset 4 - Input: Bosphorus 1080p c b a 5 10 15 20 25 SE +/- 0.25, N = 3 SE +/- 0.01, N = 3 SE +/- 0.08, N = 3 20.49 20.59 20.77 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 1.8 Encoder Mode: Preset 8 - Input: Bosphorus 1080p a c b 30 60 90 120 150 SE +/- 0.16, N = 3 SE +/- 1.86, N = 3 SE +/- 1.29, N = 7 139.81 141.87 141.98 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 1.8 Encoder Mode: Preset 12 - Input: Bosphorus 1080p b a c 120 240 360 480 600 SE +/- 3.52, N = 3 SE +/- 4.95, N = 3 SE +/- 4.85, N = 3 543.73 545.91 546.88 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
OpenBenchmarking.org Frames Per Second, More Is Better SVT-AV1 1.8 Encoder Mode: Preset 13 - Input: Bosphorus 1080p a b c 140 280 420 560 700 SE +/- 3.46, N = 3 SE +/- 4.83, N = 3 SE +/- 4.46, N = 3 632.46 636.91 637.34 1. (CXX) g++ options: -march=native -mno-avx -mavx2 -mavx512f -mavx512bw -mavx512dq
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 Xmrig CPU mining performance. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org H/s, More Is Better Xmrig 6.21 Variant: KawPow - Hash Count: 1M a c b 9K 18K 27K 36K 45K SE +/- 185.86, N = 3 SE +/- 102.95, N = 3 SE +/- 97.76, N = 3 39936.2 40242.5 40344.7 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 6.21 Variant: Monero - Hash Count: 1M c b a 9K 18K 27K 36K 45K SE +/- 234.22, N = 3 SE +/- 234.97, N = 3 SE +/- 106.27, N = 3 40072.2 40153.2 40329.6 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 6.21 Variant: Wownero - Hash Count: 1M b c a 9K 18K 27K 36K 45K SE +/- 81.91, N = 3 SE +/- 104.95, N = 3 SE +/- 76.66, N = 3 42998.7 43080.4 43088.3 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 6.21 Variant: GhostRider - Hash Count: 1M c b a 1500 3000 4500 6000 7500 SE +/- 18.86, N = 3 SE +/- 29.78, N = 3 SE +/- 5.37, N = 3 6820.8 6857.8 6862.9 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 6.21 Variant: CryptoNight-Heavy - Hash Count: 1M b c a 9K 18K 27K 36K 45K SE +/- 64.66, N = 3 SE +/- 169.87, N = 3 SE +/- 42.10, N = 3 40282.2 40286.6 40394.9 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 6.21 Variant: CryptoNight-Femto UPX2 - Hash Count: 1M b c a 9K 18K 27K 36K 45K SE +/- 72.11, N = 3 SE +/- 88.91, N = 3 SE +/- 55.21, N = 3 40080.4 40331.1 40425.9 1. (CXX) g++ options: -fexceptions -fno-rtti -maes -O3 -Ofast -static-libgcc -static-libstdc++ -rdynamic -lssl -lcrypto -luv -lpthread -lrt -ldl -lhwloc
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 1.6 Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Stream b c a 16 32 48 64 80 SE +/- 0.17, N = 3 SE +/- 0.20, N = 3 SE +/- 0.16, N = 3 71.68 71.97 71.99
OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.6 Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Synchronous Single-Stream c a b 8 16 24 32 40 SE +/- 0.07, N = 3 SE +/- 0.02, N = 3 SE +/- 0.05, N = 3 34.24 34.47 34.48
OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.6 Model: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Scenario: Asynchronous Multi-Stream c a b 500 1000 1500 2000 2500 SE +/- 0.54, N = 3 SE +/- 2.42, N = 3 SE +/- 2.92, N = 3 2322.58 2323.98 2324.74
OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.6 Model: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Scenario: Synchronous Single-Stream c b a 50 100 150 200 250 SE +/- 0.49, N = 3 SE +/- 0.83, N = 3 SE +/- 1.74, N = 3 209.61 210.81 210.96
OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.6 Model: ResNet-50, Sparse INT8 - Scenario: Asynchronous Multi-Stream c a b 1200 2400 3600 4800 6000 SE +/- 16.89, N = 3 SE +/- 15.58, N = 3 SE +/- 10.52, N = 3 5704.95 5706.70 5717.52
OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.6 Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Stream a b c 200 400 600 800 1000 SE +/- 3.85, N = 3 SE +/- 4.66, N = 3 SE +/- 0.19, N = 3 933.01 934.52 940.22
OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.6 Model: CV Classification, ResNet-50 ImageNet - Scenario: Synchronous Single-Stream b c a 60 120 180 240 300 SE +/- 0.99, N = 3 SE +/- 0.98, N = 3 SE +/- 1.63, N = 3 288.68 289.81 290.01
OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.6 Model: CV Detection, YOLOv5s COCO, Sparse INT8 - Scenario: Asynchronous Multi-Stream c b a 100 200 300 400 500 SE +/- 0.54, N = 3 SE +/- 0.02, N = 3 SE +/- 0.30, N = 3 450.49 452.16 452.20
OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.6 Model: CV Detection, YOLOv5s COCO, Sparse INT8 - Scenario: Synchronous Single-Stream c b a 40 80 120 160 200 SE +/- 0.19, N = 3 SE +/- 0.25, N = 3 SE +/- 0.33, N = 3 166.31 166.51 166.66
OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.6 Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Stream c b a 140 280 420 560 700 SE +/- 2.11, N = 3 SE +/- 1.94, N = 3 SE +/- 0.45, N = 3 637.42 637.61 641.10
OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.6 Model: NLP Text Classification, DistilBERT mnli - Scenario: Synchronous Single-Stream a c b 40 80 120 160 200 SE +/- 1.59, N = 3 SE +/- 0.41, N = 3 SE +/- 0.69, N = 3 186.03 188.23 188.54
OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.6 Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Stream b a c 20 40 60 80 100 SE +/- 0.27, N = 3 SE +/- 0.29, N = 3 SE +/- 0.08, N = 3 95.63 96.13 96.85
OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.6 Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Synchronous Single-Stream c b a 8 16 24 32 40 SE +/- 0.38, N = 15 SE +/- 0.35, N = 3 SE +/- 0.16, N = 3 33.72 34.62 35.10
OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.6 Model: BERT-Large, NLP Question Answering, Sparse INT8 - Scenario: Asynchronous Multi-Stream b c a 200 400 600 800 1000 SE +/- 0.09, N = 3 SE +/- 0.07, N = 3 SE +/- 0.34, N = 3 937.58 937.86 938.07
OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.6 Model: BERT-Large, NLP Question Answering, Sparse INT8 - Scenario: Synchronous Single-Stream c a b 15 30 45 60 75 SE +/- 0.11, N = 3 SE +/- 0.27, N = 3 SE +/- 0.12, N = 3 67.58 67.87 68.18
OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.6 Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Stream c a b 16 32 48 64 80 SE +/- 0.07, N = 3 SE +/- 0.17, N = 3 SE +/- 0.04, N = 3 71.78 72.05 72.09
OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.6 Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Synchronous Single-Stream c a b 8 16 24 32 40 SE +/- 0.03, N = 3 SE +/- 0.04, N = 3 SE +/- 0.02, N = 3 34.45 34.61 34.63
OpenBenchmarking.org Nodes Per Second, More Is Better LeelaChessZero 0.30 Backend: Eigen a b c 200 400 600 800 1000 SE +/- 21.82, N = 9 SE +/- 11.08, N = 9 SE +/- 32.94, N = 6 796 805 817 1. (CXX) g++ options: -flto -pthread
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 ms/batch, Fewer Is Better Neural Magic DeepSparse 1.6 Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Stream b c a 100 200 300 400 500 SE +/- 0.20, N = 3 SE +/- 0.77, N = 3 SE +/- 0.57, N = 3 443.40 442.75 442.31
OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.6 Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Synchronous Single-Stream c a b 7 14 21 28 35 SE +/- 0.06, N = 3 SE +/- 0.02, N = 3 SE +/- 0.05, N = 3 29.20 29.00 29.00
OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.6 Model: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Scenario: Asynchronous Multi-Stream c a b 4 8 12 16 20 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 SE +/- 0.02, N = 3 13.76 13.75 13.75
OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.6 Model: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Scenario: Synchronous Single-Stream c b a 1.0722 2.1444 3.2166 4.2888 5.361 SE +/- 0.0109, N = 3 SE +/- 0.0188, N = 3 SE +/- 0.0394, N = 3 4.7654 4.7386 4.7354
OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.6 Model: ResNet-50, Baseline - Scenario: Synchronous Single-Stream a c b 0.7836 1.5672 2.3508 3.1344 3.918 SE +/- 0.0119, N = 3 SE +/- 0.0147, N = 3 SE +/- 0.0077, N = 3 3.4825 3.4767 3.4417
OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.6 Model: ResNet-50, Sparse INT8 - Scenario: Asynchronous Multi-Stream c a b 1.2592 2.5184 3.7776 5.0368 6.296 SE +/- 0.0160, N = 3 SE +/- 0.0153, N = 3 SE +/- 0.0102, N = 3 5.5965 5.5953 5.5844
OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.6 Model: ResNet-50, Sparse INT8 - Scenario: Synchronous Single-Stream c a b 0.3196 0.6392 0.9588 1.2784 1.598 SE +/- 0.0007, N = 3 SE +/- 0.0038, N = 3 SE +/- 0.0057, N = 3 1.4205 1.4200 1.4123
OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.6 Model: BERT-Large, NLP Question Answering - Scenario: Asynchronous Multi-Stream c b a 80 160 240 320 400 SE +/- 0.22, N = 3 SE +/- 0.36, N = 3 SE +/- 0.59, N = 3 371.30 370.78 370.73
OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.6 Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Stream a b c 8 16 24 32 40 SE +/- 0.14, N = 3 SE +/- 0.17, N = 3 SE +/- 0.01, N = 3 34.28 34.22 34.01
OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.6 Model: CV Classification, ResNet-50 ImageNet - Scenario: Synchronous Single-Stream b c a 0.7787 1.5574 2.3361 3.1148 3.8935 SE +/- 0.0119, N = 3 SE +/- 0.0117, N = 3 SE +/- 0.0193, N = 3 3.4609 3.4473 3.4452
OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.6 Model: CV Detection, YOLOv5s COCO, Sparse INT8 - Scenario: Asynchronous Multi-Stream c b a 16 32 48 64 80 SE +/- 0.09, N = 3 SE +/- 0.00, N = 3 SE +/- 0.05, N = 3 71.00 70.74 70.74
OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.6 Model: CV Detection, YOLOv5s COCO, Sparse INT8 - Scenario: Synchronous Single-Stream c b a 2 4 6 8 10 SE +/- 0.0069, N = 3 SE +/- 0.0087, N = 3 SE +/- 0.0119, N = 3 6.0096 6.0023 5.9970
OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.6 Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Stream c b a 11 22 33 44 55 SE +/- 0.17, N = 3 SE +/- 0.15, N = 3 SE +/- 0.03, N = 3 50.19 50.17 49.90
OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.6 Model: NLP Text Classification, DistilBERT mnli - Scenario: Synchronous Single-Stream a c b 1.2089 2.4178 3.6267 4.8356 6.0445 SE +/- 0.0459, N = 3 SE +/- 0.0116, N = 3 SE +/- 0.0193, N = 3 5.3729 5.3092 5.3005
OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.6 Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Stream b a c 70 140 210 280 350 SE +/- 0.71, N = 3 SE +/- 1.00, N = 3 SE +/- 0.28, N = 3 334.32 332.81 330.34
OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.6 Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Synchronous Single-Stream c b a 7 14 21 28 35 SE +/- 0.34, N = 15 SE +/- 0.29, N = 3 SE +/- 0.12, N = 3 29.68 28.87 28.46
OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.6 Model: BERT-Large, NLP Question Answering, Sparse INT8 - Scenario: Asynchronous Multi-Stream b c a 8 16 24 32 40 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 34.09 34.08 34.07
OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.6 Model: BERT-Large, NLP Question Answering, Sparse INT8 - Scenario: Synchronous Single-Stream c a b 4 8 12 16 20 SE +/- 0.02, N = 3 SE +/- 0.06, N = 3 SE +/- 0.03, N = 3 14.79 14.73 14.66
OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.6 Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Stream c a b 100 200 300 400 500 SE +/- 0.25, N = 3 SE +/- 0.73, N = 3 SE +/- 0.59, N = 3 443.16 442.22 441.48
OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.6 Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Synchronous Single-Stream c a b 7 14 21 28 35 SE +/- 0.03, N = 3 SE +/- 0.03, N = 3 SE +/- 0.02, N = 3 29.02 28.89 28.88
a Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --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-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-XYspKM/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-XYspKM/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x21000161Java Notes: OpenJDK Runtime Environment (build 11.0.20+8-post-Ubuntu-1ubuntu1)Python Notes: Python 3.11.6Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 18 December 2023 11:29 by user phoronix.
b Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --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-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-XYspKM/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-XYspKM/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x21000161Python Notes: Python 3.11.6Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 18 December 2023 14:45 by user phoronix.
c Processor: INTEL XEON PLATINUM 8592+ @ 3.90GHz (64 Cores / 128 Threads), Motherboard: Quanta Cloud S6Q-MB-MPS (3B05.TEL4P1 BIOS), Chipset: Intel Device 1bce, Memory: 512GB, Disk: 3201GB Micron_7450_MTFDKCC3T2TFS + 0GB Virtual HDisk0 + 0GB Virtual HDisk2 + 0GB Virtual HDisk1 + 0GB Virtual HDisk3, Graphics: ASPEED
OS: Ubuntu 23.10, Kernel: 6.6.0-rc5-phx-patched (x86_64), Desktop: GNOME Shell 45.0, Display Server: X Server 1.21.1.7, Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 1920x1200
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=x86_64-linux-gnu --disable-vtable-verify --disable-werror --enable-bootstrap --enable-cet --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --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-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-XYspKM/gcc-13-13.2.0/debian/tmp-nvptx/usr,amdgcn-amdhsa=/build/gcc-13-XYspKM/gcc-13-13.2.0/debian/tmp-gcn/usr --enable-plugin --enable-shared --enable-threads=posix --host=x86_64-linux-gnu --program-prefix=x86_64-linux-gnu- --target=x86_64-linux-gnu --with-abi=m64 --with-arch-32=i686 --with-build-config=bootstrap-lto-lean --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-multilib-list=m32,m64,mx32 --with-target-system-zlib=auto --with-tune=generic --without-cuda-driver -vProcessor Notes: Scaling Governor: intel_pstate performance (EPP: performance) - CPU Microcode: 0x21000161Python Notes: Python 3.11.6Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Enhanced / Automatic IBRS IBPB: conditional RSB filling PBRSB-eIBRS: SW sequence + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 18 December 2023 17:34 by user phoronix.