deepsparse tr AMD Ryzen Threadripper 3990X 64-Core testing with a Gigabyte TRX40 AORUS PRO WIFI (F4p BIOS) and AMD Radeon RX 5700 8GB on Ubuntu 22.10 via the Phoronix Test Suite.
HTML result view exported from: https://openbenchmarking.org/result/2301232-NE-DEEPSPARS10&rdt&grs .
deepsparse tr Processor Motherboard Chipset Memory Disk Graphics Audio Monitor Network OS Kernel Desktop Display Server OpenGL Vulkan Compiler File-System Screen Resolution a b c AMD Ryzen Threadripper 3990X 64-Core @ 2.90GHz (64 Cores / 128 Threads) Gigabyte TRX40 AORUS PRO WIFI (F4p BIOS) AMD Starship/Matisse 128GB Samsung SSD 970 EVO Plus 500GB AMD Radeon RX 5700 8GB (1750/875MHz) AMD Navi 10 HDMI Audio DELL P2415Q Intel I211 + Intel Wi-Fi 6 AX200 Ubuntu 22.10 6.1.0-rc8-phx-mglru (x86_64) GNOME Shell 43.0 X Server 1.21.1.4 + Wayland 4.6 Mesa 22.2.1 (LLVM 15.0.2 DRM 3.49) 1.3.224 GCC 12.2.0 + LLVM 15.0.2 ext4 3840x2160 OpenBenchmarking.org Kernel Details - Transparent Huge Pages: madvise Processor Details - Scaling Governor: acpi-cpufreq schedutil (Boost: Enabled) - CPU Microcode: 0x8301055 Python Details - Python 3.10.7 Security Details - itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + retbleed: Mitigation of untrained return thunk; SMT enabled with STIBP protection + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of usercopy/swapgs barriers and __user pointer sanitization + spectre_v2: Mitigation of Retpolines IBPB: conditional STIBP: always-on RSB filling PBRSB-eIBRS: Not affected + srbds: Not affected + tsx_async_abort: Not affected
deepsparse tr deepsparse: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Asynchronous Multi-Stream deepsparse: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Asynchronous Multi-Stream deepsparse: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Synchronous Single-Stream deepsparse: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Synchronous Single-Stream deepsparse: CV Detection, YOLOv5s COCO - Synchronous Single-Stream deepsparse: CV Detection, YOLOv5s COCO - Synchronous Single-Stream deepsparse: NLP Text Classification, DistilBERT mnli - Synchronous Single-Stream deepsparse: NLP Text Classification, DistilBERT mnli - Synchronous Single-Stream deepsparse: NLP Token Classification, BERT base uncased conll2003 - Synchronous Single-Stream deepsparse: NLP Token Classification, BERT base uncased conll2003 - Synchronous Single-Stream deepsparse: NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Stream deepsparse: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Synchronous Single-Stream deepsparse: NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Stream deepsparse: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Synchronous Single-Stream deepsparse: NLP Text Classification, BERT base uncased SST2 - Synchronous Single-Stream deepsparse: NLP Text Classification, BERT base uncased SST2 - Synchronous Single-Stream deepsparse: NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Stream deepsparse: NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Stream deepsparse: CV Classification, ResNet-50 ImageNet - Synchronous Single-Stream deepsparse: CV Classification, ResNet-50 ImageNet - Synchronous Single-Stream deepsparse: NLP Text Classification, BERT base uncased SST2 - Asynchronous Multi-Stream deepsparse: NLP Text Classification, BERT base uncased SST2 - Asynchronous Multi-Stream deepsparse: NLP Document Classification, oBERT base uncased on IMDB - Synchronous Single-Stream deepsparse: NLP Document Classification, oBERT base uncased on IMDB - Synchronous Single-Stream deepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Asynchronous Multi-Stream deepsparse: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Asynchronous Multi-Stream deepsparse: NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Stream deepsparse: CV Detection, YOLOv5s COCO - Asynchronous Multi-Stream deepsparse: CV Detection, YOLOv5s COCO - Asynchronous Multi-Stream deepsparse: NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Stream deepsparse: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Asynchronous Multi-Stream deepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Asynchronous Multi-Stream deepsparse: CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Stream deepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Synchronous Single-Stream deepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Synchronous Single-Stream deepsparse: CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Stream a b c 121.5987 263.2698 21.0843 47.4189 15.6921 63.6711 17.2750 57.8622 74.5910 13.4049 911.3941 48.0131 34.8860 20.8260 30.5467 32.7294 279.3454 114.4637 10.5637 94.5820 142.2132 224.9088 13.6680 73.1554 43.5414 360.5545 943.8212 171.8227 185.7590 33.4855 88.5450 728.3385 77.1694 54.6204 18.3022 414.4296 117.8674 271.2283 20.1109 49.7013 16.4399 60.7788 16.627 60.1136 72.1548 13.8574 916.5997 47.3506 34.4957 21.1142 30.4182 32.8655 284.1417 112.5672 10.5084 95.0838 143.9332 222.2642 13.6164 73.4315 43.8726 359.2449 941.5145 171.5536 185.9302 33.6307 88.7699 725.2629 76.9125 54.5493 18.3262 415.6349 115.7776 276.2048 20.1262 49.6628 16.0481 62.2604 16.7985 59.4981 72.8486 13.7254 929.8296 47.0812 34.2097 21.2348 31.0025 32.2466 283.3205 112.8945 10.6724 93.6235 143.714 222.6035 13.7707 72.6091 43.7343 358.0307 937.8174 172.622 184.7932 33.6845 89.0653 724.5184 77.1478 54.4505 18.3593 414.552 OpenBenchmarking.org
Neural Magic DeepSparse Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Asynchronous Multi-Stream 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 a b c 30 60 90 120 150 SE +/- 1.25, N = 15 121.60 117.87 115.78
Neural Magic DeepSparse Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Asynchronous Multi-Stream 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 a b c 60 120 180 240 300 SE +/- 2.74, N = 15 263.27 271.23 276.20
Neural Magic DeepSparse Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Synchronous Single-Stream OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.3.2 Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Synchronous Single-Stream a b c 5 10 15 20 25 SE +/- 0.22, N = 3 21.08 20.11 20.13
Neural Magic DeepSparse Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Synchronous Single-Stream OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.3.2 Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Synchronous Single-Stream a b c 11 22 33 44 55 SE +/- 0.50, N = 3 47.42 49.70 49.66
Neural Magic DeepSparse Model: CV Detection, YOLOv5s COCO - Scenario: Synchronous Single-Stream OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.3.2 Model: CV Detection, YOLOv5s COCO - Scenario: Synchronous Single-Stream a b c 4 8 12 16 20 SE +/- 0.03, N = 3 15.69 16.44 16.05
Neural Magic DeepSparse Model: CV Detection, YOLOv5s COCO - Scenario: Synchronous Single-Stream OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.3.2 Model: CV Detection, YOLOv5s COCO - Scenario: Synchronous Single-Stream a b c 14 28 42 56 70 SE +/- 0.11, N = 3 63.67 60.78 62.26
Neural Magic DeepSparse Model: NLP Text Classification, DistilBERT mnli - Scenario: Synchronous Single-Stream OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.3.2 Model: NLP Text Classification, DistilBERT mnli - Scenario: Synchronous Single-Stream a b c 4 8 12 16 20 SE +/- 0.10, N = 3 17.28 16.63 16.80
Neural Magic DeepSparse Model: NLP Text Classification, DistilBERT mnli - Scenario: Synchronous Single-Stream OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.3.2 Model: NLP Text Classification, DistilBERT mnli - Scenario: Synchronous Single-Stream a b c 13 26 39 52 65 SE +/- 0.34, N = 3 57.86 60.11 59.50
Neural Magic DeepSparse Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Synchronous Single-Stream OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.3.2 Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Synchronous Single-Stream a b c 20 40 60 80 100 SE +/- 0.09, N = 3 74.59 72.15 72.85
Neural Magic DeepSparse Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Synchronous Single-Stream OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.3.2 Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Synchronous Single-Stream a b c 4 8 12 16 20 SE +/- 0.02, N = 3 13.40 13.86 13.73
Neural Magic DeepSparse Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Stream 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 a b c 200 400 600 800 1000 SE +/- 8.05, N = 3 911.39 916.60 929.83
Neural Magic DeepSparse Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Synchronous Single-Stream OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.3.2 Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Synchronous Single-Stream a b c 11 22 33 44 55 SE +/- 0.42, N = 3 48.01 47.35 47.08
Neural Magic DeepSparse Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Stream 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 a b c 8 16 24 32 40 SE +/- 0.38, N = 3 34.89 34.50 34.21
Neural Magic DeepSparse Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Synchronous Single-Stream OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.3.2 Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Synchronous Single-Stream a b c 5 10 15 20 25 SE +/- 0.18, N = 3 20.83 21.11 21.23
Neural Magic DeepSparse Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Synchronous Single-Stream OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.3.2 Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Synchronous Single-Stream a b c 7 14 21 28 35 SE +/- 0.12, N = 3 30.55 30.42 31.00
Neural Magic DeepSparse Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Synchronous Single-Stream OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.3.2 Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Synchronous Single-Stream a b c 8 16 24 32 40 SE +/- 0.13, N = 3 32.73 32.87 32.25
Neural Magic DeepSparse Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Stream OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.3.2 Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Stream a b c 60 120 180 240 300 SE +/- 0.58, N = 3 279.35 284.14 283.32
Neural Magic DeepSparse Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Stream OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.3.2 Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Stream a b c 30 60 90 120 150 SE +/- 0.20, N = 3 114.46 112.57 112.89
Neural Magic DeepSparse Model: CV Classification, ResNet-50 ImageNet - Scenario: Synchronous Single-Stream OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.3.2 Model: CV Classification, ResNet-50 ImageNet - Scenario: Synchronous Single-Stream a b c 3 6 9 12 15 SE +/- 0.02, N = 3 10.56 10.51 10.67
Neural Magic DeepSparse Model: CV Classification, ResNet-50 ImageNet - Scenario: Synchronous Single-Stream OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.3.2 Model: CV Classification, ResNet-50 ImageNet - Scenario: Synchronous Single-Stream a b c 20 40 60 80 100 SE +/- 0.14, N = 3 94.58 95.08 93.62
Neural Magic DeepSparse Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-Stream 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 a b c 30 60 90 120 150 SE +/- 0.04, N = 3 142.21 143.93 143.71
Neural Magic DeepSparse Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-Stream 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 a b c 50 100 150 200 250 SE +/- 0.04, N = 3 224.91 222.26 222.60
Neural Magic DeepSparse Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Synchronous Single-Stream OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.3.2 Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Synchronous Single-Stream a b c 4 8 12 16 20 SE +/- 0.03, N = 3 13.67 13.62 13.77
Neural Magic DeepSparse Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Synchronous Single-Stream OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.3.2 Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Synchronous Single-Stream a b c 16 32 48 64 80 SE +/- 0.18, N = 3 73.16 73.43 72.61
Neural Magic DeepSparse Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Stream OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.3.2 Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Stream a b c 10 20 30 40 50 SE +/- 0.05, N = 3 43.54 43.87 43.73
Neural Magic DeepSparse Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-Stream 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 a b c 80 160 240 320 400 SE +/- 1.72, N = 3 360.55 359.24 358.03
Neural Magic DeepSparse Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Stream 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 a b c 200 400 600 800 1000 SE +/- 1.04, N = 3 943.82 941.51 937.82
Neural Magic DeepSparse Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-Stream OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.3.2 Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-Stream a b c 40 80 120 160 200 SE +/- 0.14, N = 3 171.82 171.55 172.62
Neural Magic DeepSparse Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-Stream OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.3.2 Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-Stream a b c 40 80 120 160 200 SE +/- 0.12, N = 3 185.76 185.93 184.79
Neural Magic DeepSparse Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Stream 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 a b c 8 16 24 32 40 SE +/- 0.05, N = 3 33.49 33.63 33.68
Neural Magic DeepSparse Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-Stream 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 a b c 20 40 60 80 100 SE +/- 0.36, N = 3 88.55 88.77 89.07
Neural Magic DeepSparse Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Stream OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.3.2 Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Stream a b c 160 320 480 640 800 SE +/- 0.38, N = 3 728.34 725.26 724.52
Neural Magic DeepSparse Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Stream OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.3.2 Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Stream a b c 20 40 60 80 100 SE +/- 0.02, N = 3 77.17 76.91 77.15
Neural Magic DeepSparse Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Synchronous Single-Stream OpenBenchmarking.org ms/batch, Fewer Is Better Neural Magic DeepSparse 1.3.2 Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Synchronous Single-Stream a b c 12 24 36 48 60 SE +/- 0.03, N = 3 54.62 54.55 54.45
Neural Magic DeepSparse Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Synchronous Single-Stream OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.3.2 Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Synchronous Single-Stream a b c 5 10 15 20 25 SE +/- 0.01, N = 3 18.30 18.33 18.36
Neural Magic DeepSparse Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Stream OpenBenchmarking.org items/sec, More Is Better Neural Magic DeepSparse 1.3.2 Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Stream a b c 90 180 270 360 450 SE +/- 0.08, N = 3 414.43 415.63 414.55
Phoronix Test Suite v10.8.5