deepsparse unv

Intel Core i7-1165G7 testing with a Dell 0GG9PT (3.11.0 BIOS) and Intel Xe TGL GT2 15GB on Ubuntu 22.10 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 2302015-NE-DEEPSPARS80
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

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
Hide
Result
Result
Identifier
View Logs
Performance Per
Dollar
Date
Run
  Test
  Duration
a
January 31 2023
  31 Minutes
b
January 31 2023
  30 Minutes
c
January 31 2023
  6 Hours, 45 Minutes
d
February 01 2023
  6 Hours, 1 Minute
Invert Hiding All Results Option
  3 Hours, 27 Minutes

Only show results where is faster than
Only show results matching title/arguments (delimit multiple options with a comma):
Do not show results matching title/arguments (delimit multiple options with a comma):

a

Kernel Notes: Transparent Huge Pages: madvise
Processor Notes: Scaling Governor: intel_pstate powersave (EPP: balance_performance) - CPU Microcode: 0xa4 - Thermald 2.5.1
Python Notes: Python 3.10.7
Security Notes: 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 + 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

b

c

d

Processor: Intel Core i7-1165G7 @ 4.70GHz (4 Cores / 8 Threads), Motherboard: Dell 0GG9PT (3.11.0 BIOS), Chipset: Intel Tiger Lake-LP, Memory: 16GB, Disk: Kioxia KBG40ZNS256G NVMe 256GB, Graphics: Intel Xe TGL GT2 15GB (1300MHz), Audio: Realtek ALC289, Network: Intel Wi-Fi 6 AX201

OS: Ubuntu 22.10, Kernel: 5.19.0-29-generic (x86_64), Desktop: GNOME Shell 43.0, Display Server: X Server + Wayland, OpenGL: 4.6 Mesa 22.2.1, Vulkan: 1.3.224, Compiler: GCC 12.2.0, File-System: ext4, Screen Resolution: 1920x1200

deepsparse unvOpenBenchmarking.orgPhoronix Test SuiteIntel Core i7-1165G7 @ 4.70GHz (4 Cores / 8 Threads)Dell 0GG9PT (3.11.0 BIOS)Intel Tiger Lake-LP16GBKioxia KBG40ZNS256G NVMe 256GBIntel Xe TGL GT2 15GB (1300MHz)Realtek ALC289Intel Wi-Fi 6 AX201Ubuntu 22.105.19.0-29-generic (x86_64)GNOME Shell 43.0X Server + Wayland4.6 Mesa 22.2.11.3.224GCC 12.2.0ext41920x1200ProcessorMotherboardChipsetMemoryDiskGraphicsAudioNetworkOSKernelDesktopDisplay ServerOpenGLVulkanCompilerFile-SystemScreen ResolutionDeepsparse Unv BenchmarksSystem Logs- Transparent Huge Pages: madvise- Scaling Governor: intel_pstate powersave (EPP: balance_performance) - CPU Microcode: 0xa4 - Thermald 2.5.1 - Python 3.10.7- 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 + 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

abcdResult OverviewPhoronix Test Suite100%105%109%114%119%Neural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseNeural Magic DeepSparseUnvanquishedUnvanquishedUnvanquishedUnvanquishedUnvanquishedUnvanquishedNeural Magic DeepSparseNeural Magic DeepSparseC.C.R.5.I - A.M.SC.C.R.5.I - A.M.SN.T.C.B.b.u.c - A.M.SN.T.C.B.b.u.c - A.M.SN.T.C.D.m - S.S.SN.Q.A.B.b.u.S.1.P - A.M.SN.Q.A.B.b.u.S.1.P - A.M.SN.T.C.D.m - A.M.SN.T.C.B.b.u.c - S.S.SN.T.C.B.b.u.c - S.S.SN.T.C.D.m - S.S.SN.T.C.D.m - A.M.SN.D.C.o.b.u.o.I - A.M.SC.D.Y.C - S.S.SN.D.C.o.b.u.o.I - A.M.SC.D.Y.C - S.S.SN.T.C.B.b.u.S - A.M.SN.T.C.B.b.u.S - A.M.SN.D.C.o.b.u.o.I - S.S.SN.D.C.o.b.u.o.I - S.S.SN.S.A.8.P.Q.B.B.U - A.M.SN.S.A.8.P.Q.B.B.U - A.M.SC.S.9.P.Y.P - A.M.SC.S.9.P.Y.P - A.M.SN.T.C.B.b.u.S - S.S.SN.T.C.B.b.u.S - S.S.SN.S.A.8.P.Q.B.B.U - S.S.SC.D.Y.C - A.M.SN.Q.A.B.b.u.S.1.P - S.S.SN.S.A.8.P.Q.B.B.U - S.S.SN.Q.A.B.b.u.S.1.P - S.S.SC.D.Y.C - A.M.SC.S.9.P.Y.P - S.S.SC.S.9.P.Y.P - S.S.S1920 x 1200 - Medium1920 x 1080 - Medium1920 x 1080 - High1920 x 1080 - Ultra1920 x 1200 - Ultra1920 x 1200 - HighC.C.R.5.I - S.S.SC.C.R.5.I - S.S.S

deepsparse unvunvanquished: 1920 x 1080 - Highunvanquished: 1920 x 1200 - Highunvanquished: 1920 x 1080 - Ultraunvanquished: 1920 x 1200 - Ultraunvanquished: 1920 x 1080 - Mediumunvanquished: 1920 x 1200 - Mediumdeepsparse: NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Streamdeepsparse: NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Streamdeepsparse: NLP Document Classification, oBERT base uncased on IMDB - Synchronous Single-Streamdeepsparse: NLP Document Classification, oBERT base uncased on IMDB - Synchronous Single-Streamdeepsparse: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Asynchronous Multi-Streamdeepsparse: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Asynchronous Multi-Streamdeepsparse: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Synchronous Single-Streamdeepsparse: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Synchronous Single-Streamdeepsparse: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Asynchronous Multi-Streamdeepsparse: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Asynchronous Multi-Streamdeepsparse: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Synchronous Single-Streamdeepsparse: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Synchronous Single-Streamdeepsparse: CV Detection, YOLOv5s COCO - Asynchronous Multi-Streamdeepsparse: CV Detection, YOLOv5s COCO - Asynchronous Multi-Streamdeepsparse: CV Detection, YOLOv5s COCO - Synchronous Single-Streamdeepsparse: CV Detection, YOLOv5s COCO - Synchronous Single-Streamdeepsparse: CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Streamdeepsparse: CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Streamdeepsparse: CV Classification, ResNet-50 ImageNet - Synchronous Single-Streamdeepsparse: CV Classification, ResNet-50 ImageNet - Synchronous Single-Streamdeepsparse: NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Streamdeepsparse: NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Streamdeepsparse: NLP Text Classification, DistilBERT mnli - Synchronous Single-Streamdeepsparse: NLP Text Classification, DistilBERT mnli - Synchronous Single-Streamdeepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Asynchronous Multi-Streamdeepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Asynchronous Multi-Streamdeepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Synchronous Single-Streamdeepsparse: CV Segmentation, 90% Pruned YOLACT Pruned - Synchronous Single-Streamdeepsparse: NLP Text Classification, BERT base uncased SST2 - Asynchronous Multi-Streamdeepsparse: NLP Text Classification, BERT base uncased SST2 - Asynchronous Multi-Streamdeepsparse: NLP Text Classification, BERT base uncased SST2 - Synchronous Single-Streamdeepsparse: NLP Text Classification, BERT base uncased SST2 - Synchronous Single-Streamdeepsparse: NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Streamdeepsparse: NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Streamdeepsparse: NLP Token Classification, BERT base uncased conll2003 - Synchronous Single-Streamdeepsparse: NLP Token Classification, BERT base uncased conll2003 - Synchronous Single-Streamabcd288.9282.8218.9208.8313.5308.13.6314550.72573.5177284.264968.607929.119662.156916.078317.6171113.467217.007358.787721.813891.652420.594548.544258.778733.995549.511620.18731.644363.179430.029833.2936.5581304.92596.9658143.541314.589137.064913.591773.56693.6554547.10183.5135284.6107289.8283.2220.1208.2318.2315.23.617552.873.5154284.45169.847628.596459.988916.659418.1684109.995617.305257.774622.197990.06620.614648.49760.164533.211549.681320.117332.23862.015130.736832.52787.3184273.2216.5168153.429715.5764128.36914.16270.6033.1909626.70153.0315329.8659287.4282.0217.0207.0313.3310.53.1630634.16973.1185321.171564.580330.999558.211417.206615.7258127.215116.107762.153420.888495.964718.420454.538852.023438.481850.104219.994228.571770.383327.315136.80496.7233297.54076.6878149.682714.1127142.244113.149676.24133.1323639.67503.1354319.5068284.4279.0216.0205.3311.8308.83.2394619.92273.1391319.179962.332032.048757.209017.497515.6403127.838515.929962.792420.460897.902817.967555.791650.638339.498949.846120.077127.882772.027026.534037.80806.6054302.30216.5320153.074413.7617145.708513.005777.00293.0903648.15933.0976323.2343OpenBenchmarking.org

Unvanquished

Unvanquished is a modern fork of the Tremulous first person shooter. Unvanquished is powered by the Daemon engine, a combination of the ioquake3 (id Tech 3) engine with the graphically-beautiful XreaL engine. Unvanquished supports a modern OpenGL 3 renderer and other advanced graphics features for this open-source, cross-platform shooter game. Learn more via the OpenBenchmarking.org test page.

OpenBenchmarking.orgFrames Per Second, More Is BetterUnvanquished 0.54Resolution: 1920 x 1080 - Effects Quality: Highabcd60120180240300SE +/- 0.82, N = 3SE +/- 0.31, N = 3288.9289.8287.4284.4
OpenBenchmarking.orgFrames Per Second, More Is BetterUnvanquished 0.54Resolution: 1920 x 1080 - Effects Quality: Highabcd50100150200250Min: 286.3 / Avg: 287.4 / Max: 289Min: 283.8 / Avg: 284.4 / Max: 284.8

OpenBenchmarking.orgFrames Per Second, More Is BetterUnvanquished 0.54Resolution: 1920 x 1200 - Effects Quality: Highabcd60120180240300SE +/- 0.41, N = 3SE +/- 0.70, N = 3282.8283.2282.0279.0
OpenBenchmarking.orgFrames Per Second, More Is BetterUnvanquished 0.54Resolution: 1920 x 1200 - Effects Quality: Highabcd50100150200250Min: 281.4 / Avg: 282.03 / Max: 282.8Min: 278.3 / Avg: 279 / Max: 280.4

OpenBenchmarking.orgFrames Per Second, More Is BetterUnvanquished 0.54Resolution: 1920 x 1080 - Effects Quality: Ultraabcd50100150200250SE +/- 0.73, N = 3SE +/- 0.22, N = 3218.9220.1217.0216.0
OpenBenchmarking.orgFrames Per Second, More Is BetterUnvanquished 0.54Resolution: 1920 x 1080 - Effects Quality: Ultraabcd4080120160200Min: 216.3 / Avg: 217.03 / Max: 218.5Min: 215.7 / Avg: 215.97 / Max: 216.4

OpenBenchmarking.orgFrames Per Second, More Is BetterUnvanquished 0.54Resolution: 1920 x 1200 - Effects Quality: Ultraabcd50100150200250SE +/- 0.40, N = 3SE +/- 0.12, N = 3208.8208.2207.0205.3
OpenBenchmarking.orgFrames Per Second, More Is BetterUnvanquished 0.54Resolution: 1920 x 1200 - Effects Quality: Ultraabcd4080120160200Min: 206.5 / Avg: 207 / Max: 207.8Min: 205.1 / Avg: 205.3 / Max: 205.5

OpenBenchmarking.orgFrames Per Second, More Is BetterUnvanquished 0.54Resolution: 1920 x 1080 - Effects Quality: Mediumabcd70140210280350SE +/- 1.01, N = 3SE +/- 0.64, N = 3313.5318.2313.3311.8
OpenBenchmarking.orgFrames Per Second, More Is BetterUnvanquished 0.54Resolution: 1920 x 1080 - Effects Quality: Mediumabcd60120180240300Min: 311.5 / Avg: 313.27 / Max: 315Min: 310.7 / Avg: 311.83 / Max: 312.9

OpenBenchmarking.orgFrames Per Second, More Is BetterUnvanquished 0.54Resolution: 1920 x 1200 - Effects Quality: Mediumabcd70140210280350SE +/- 0.74, N = 3SE +/- 1.17, N = 3308.1315.2310.5308.8
OpenBenchmarking.orgFrames Per Second, More Is BetterUnvanquished 0.54Resolution: 1920 x 1200 - Effects Quality: Mediumabcd60120180240300Min: 309.4 / Avg: 310.47 / Max: 311.9Min: 307.5 / Avg: 308.77 / Max: 311.1

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.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Streamabcd0.81711.63422.45133.26844.0855SE +/- 0.0481, N = 15SE +/- 0.0558, N = 153.63143.61703.16303.2394
OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Streamabcd246810Min: 2.97 / Avg: 3.16 / Max: 3.61Min: 2.96 / Avg: 3.24 / Max: 3.6

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Streamabcd140280420560700SE +/- 9.29, N = 15SE +/- 10.87, N = 15550.73552.87634.17619.92
OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Streamabcd110220330440550Min: 554.21 / Avg: 634.17 / Max: 673.2Min: 555.29 / Avg: 619.92 / Max: 675.66

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Synchronous Single-Streamabcd0.79151.5832.37453.1663.9575SE +/- 0.0338, N = 15SE +/- 0.0375, N = 153.51773.51543.11853.1391
OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Synchronous Single-Streamabcd246810Min: 3.01 / Avg: 3.12 / Max: 3.36Min: 3 / Avg: 3.14 / Max: 3.36

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Synchronous Single-Streamabcd70140210280350SE +/- 3.37, N = 15SE +/- 3.74, N = 15284.26284.45321.17319.18
OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Synchronous Single-Streamabcd60120180240300Min: 297.39 / Avg: 321.17 / Max: 332.73Min: 297.56 / Avg: 319.18 / Max: 332.89

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Asynchronous Multi-Streamabcd1632486480SE +/- 0.82, N = 15SE +/- 0.24, N = 368.6169.8564.5862.33
OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Asynchronous Multi-Streamabcd1428425670Min: 61.81 / Avg: 64.58 / Max: 72.07Min: 61.87 / Avg: 62.33 / Max: 62.68

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Asynchronous Multi-Streamabcd714212835SE +/- 0.37, N = 15SE +/- 0.12, N = 329.1228.6031.0032.05
OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Asynchronous Multi-Streamabcd714212835Min: 27.71 / Avg: 31 / Max: 32.32Min: 31.87 / Avg: 32.05 / Max: 32.29

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Synchronous Single-Streamabcd1428425670SE +/- 0.74, N = 15SE +/- 0.63, N = 1562.1659.9958.2157.21
OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Synchronous Single-Streamabcd1224364860Min: 55.55 / Avg: 58.21 / Max: 62.35Min: 55.26 / Avg: 57.21 / Max: 62.07

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Synchronous Single-Streamabcd48121620SE +/- 0.21, N = 15SE +/- 0.18, N = 1516.0816.6617.2117.50
OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Synchronous Single-Streamabcd48121620Min: 16.03 / Avg: 17.21 / Max: 17.99Min: 16.1 / Avg: 17.5 / Max: 18.09

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-Streamabcd48121620SE +/- 0.15, N = 13SE +/- 0.12, N = 1317.6218.1715.7315.64
OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-Streamabcd510152025Min: 15.44 / Avg: 15.73 / Max: 17.5Min: 15.42 / Avg: 15.64 / Max: 17.12

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-Streamabcd306090120150SE +/- 1.12, N = 13SE +/- 0.93, N = 13113.47110.00127.22127.84
OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-Streamabcd20406080100Min: 114.23 / Avg: 127.22 / Max: 129.41Min: 116.79 / Avg: 127.84 / Max: 129.51

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Synchronous Single-Streamabcd48121620SE +/- 0.16, N = 15SE +/- 0.10, N = 1317.0117.3116.1115.93
OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Synchronous Single-Streamabcd48121620Min: 15.65 / Avg: 16.11 / Max: 18.11Min: 15.72 / Avg: 15.93 / Max: 17.09

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Synchronous Single-Streamabcd1428425670SE +/- 0.58, N = 15SE +/- 0.37, N = 1358.7957.7762.1562.79
OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Synchronous Single-Streamabcd1224364860Min: 55.22 / Avg: 62.15 / Max: 63.89Min: 58.52 / Avg: 62.79 / Max: 63.62

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-Streamabcd510152025SE +/- 0.29, N = 15SE +/- 0.25, N = 1521.8122.2020.8920.46
OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-Streamabcd510152025Min: 19.82 / Avg: 20.89 / Max: 23.26Min: 19.8 / Avg: 20.46 / Max: 22.75

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-Streamabcd20406080100SE +/- 1.30, N = 15SE +/- 1.14, N = 1591.6590.0795.9697.90
OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-Streamabcd20406080100Min: 85.94 / Avg: 95.96 / Max: 100.86Min: 87.89 / Avg: 97.9 / Max: 100.97

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: CV Detection, YOLOv5s COCO - Scenario: Synchronous Single-Streamabcd510152025SE +/- 0.35, N = 15SE +/- 0.25, N = 1520.5920.6118.4217.97
OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: CV Detection, YOLOv5s COCO - Scenario: Synchronous Single-Streamabcd510152025Min: 17.23 / Avg: 18.42 / Max: 20.66Min: 17.23 / Avg: 17.97 / Max: 19.69

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: CV Detection, YOLOv5s COCO - Scenario: Synchronous Single-Streamabcd1326395265SE +/- 1.00, N = 15SE +/- 0.75, N = 1548.5448.5054.5455.79
OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: CV Detection, YOLOv5s COCO - Scenario: Synchronous Single-Streamabcd1122334455Min: 48.39 / Avg: 54.54 / Max: 58.01Min: 50.79 / Avg: 55.79 / Max: 58.03

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Streamabcd1326395265SE +/- 0.62, N = 15SE +/- 0.42, N = 1558.7860.1652.0250.64
OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Streamabcd1224364860Min: 49.89 / Avg: 52.02 / Max: 57.28Min: 49.99 / Avg: 50.64 / Max: 56.41

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Streamabcd918273645SE +/- 0.44, N = 15SE +/- 0.30, N = 1534.0033.2138.4839.50
OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Streamabcd816243240Min: 34.88 / Avg: 38.48 / Max: 40.06Min: 35.43 / Avg: 39.5 / Max: 39.98

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: CV Classification, ResNet-50 ImageNet - Scenario: Synchronous Single-Streamabcd1122334455SE +/- 0.66, N = 15SE +/- 0.48, N = 1549.5149.6850.1049.85
OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: CV Classification, ResNet-50 ImageNet - Scenario: Synchronous Single-Streamabcd1020304050Min: 48.63 / Avg: 50.1 / Max: 55.52Min: 48.59 / Avg: 49.85 / Max: 54.14

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: CV Classification, ResNet-50 ImageNet - Scenario: Synchronous Single-Streamabcd510152025SE +/- 0.25, N = 15SE +/- 0.19, N = 1520.1920.1219.9920.08
OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: CV Classification, ResNet-50 ImageNet - Scenario: Synchronous Single-Streamabcd510152025Min: 18 / Avg: 19.99 / Max: 20.55Min: 18.46 / Avg: 20.08 / Max: 20.57

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Streamabcd714212835SE +/- 0.58, N = 15SE +/- 0.51, N = 1531.6432.2428.5727.88
OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Streamabcd714212835Min: 25.9 / Avg: 28.57 / Max: 32.12Min: 25.96 / Avg: 27.88 / Max: 31.76

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Streamabcd1632486480SE +/- 1.43, N = 15SE +/- 1.27, N = 1563.1862.0270.3872.03
OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Streamabcd1428425670Min: 62.23 / Avg: 70.38 / Max: 77.19Min: 62.94 / Avg: 72.03 / Max: 77.01

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Text Classification, DistilBERT mnli - Scenario: Synchronous Single-Streamabcd714212835SE +/- 0.55, N = 15SE +/- 0.42, N = 1530.0330.7427.3226.53
OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Text Classification, DistilBERT mnli - Scenario: Synchronous Single-Streamabcd714212835Min: 25.25 / Avg: 27.32 / Max: 30.72Min: 25.04 / Avg: 26.53 / Max: 29.47

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Text Classification, DistilBERT mnli - Scenario: Synchronous Single-Streamabcd918273645SE +/- 0.72, N = 15SE +/- 0.58, N = 1533.2932.5336.8037.81
OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Text Classification, DistilBERT mnli - Scenario: Synchronous Single-Streamabcd816243240Min: 32.55 / Avg: 36.8 / Max: 39.6Min: 33.92 / Avg: 37.81 / Max: 39.94

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Streamabcd246810SE +/- 0.0680, N = 15SE +/- 0.0457, N = 136.55817.31846.72336.6054
OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Streamabcd3691215Min: 6.52 / Avg: 6.72 / Max: 7.4Min: 6.5 / Avg: 6.61 / Max: 7.06

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Streamabcd70140210280350SE +/- 2.86, N = 15SE +/- 1.96, N = 13304.93273.22297.54302.30
OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Streamabcd50100150200250Min: 270.1 / Avg: 297.54 / Max: 306.09Min: 283.11 / Avg: 302.3 / Max: 307.11

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Synchronous Single-Streamabcd246810SE +/- 0.0619, N = 15SE +/- 0.0031, N = 36.96586.51686.68786.5320
OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Synchronous Single-Streamabcd3691215Min: 6.5 / Avg: 6.69 / Max: 7.13Min: 6.53 / Avg: 6.53 / Max: 6.54

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Synchronous Single-Streamabcd306090120150SE +/- 1.35, N = 15SE +/- 0.07, N = 3143.54153.43149.68153.07
OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Synchronous Single-Streamabcd306090120150Min: 140.15 / Avg: 149.68 / Max: 153.89Min: 152.95 / Avg: 153.07 / Max: 153.2

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-Streamabcd48121620SE +/- 0.24, N = 15SE +/- 0.21, N = 1514.5915.5814.1113.76
OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-Streamabcd48121620Min: 13.05 / Avg: 14.11 / Max: 15.74Min: 13.15 / Avg: 13.76 / Max: 15.26

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-Streamabcd306090120150SE +/- 2.35, N = 15SE +/- 2.11, N = 15137.06128.37142.24145.71
OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-Streamabcd306090120150Min: 127 / Avg: 142.24 / Max: 153.18Min: 131.02 / Avg: 145.71 / Max: 152.11

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Synchronous Single-Streamabcd48121620SE +/- 0.18, N = 15SE +/- 0.14, N = 1513.5914.1613.1513.01
OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Synchronous Single-Streamabcd48121620Min: 12.6 / Avg: 13.15 / Max: 14.38Min: 12.55 / Avg: 13.01 / Max: 14.35

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Synchronous Single-Streamabcd20406080100SE +/- 1.03, N = 15SE +/- 0.80, N = 1573.5770.6076.2477.00
OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Synchronous Single-Streamabcd1530456075Min: 69.54 / Avg: 76.24 / Max: 79.38Min: 69.69 / Avg: 77 / Max: 79.65

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Streamabcd0.82251.6452.46753.294.1125SE +/- 0.0413, N = 15SE +/- 0.0362, N = 153.65543.19093.13233.0903
OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Streamabcd246810Min: 2.99 / Avg: 3.13 / Max: 3.55Min: 2.97 / Avg: 3.09 / Max: 3.37

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Streamabcd140280420560700SE +/- 7.96, N = 15SE +/- 7.31, N = 15547.10626.70639.68648.16
OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Streamabcd110220330440550Min: 562.99 / Avg: 639.68 / Max: 669.08Min: 592.62 / Avg: 648.16 / Max: 674.09

OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Synchronous Single-Streamabcd0.79051.5812.37153.1623.9525SE +/- 0.0360, N = 15SE +/- 0.0301, N = 153.51353.03153.13543.0976
OpenBenchmarking.orgitems/sec, More Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Synchronous Single-Streamabcd246810Min: 3.02 / Avg: 3.14 / Max: 3.36Min: 3.02 / Avg: 3.1 / Max: 3.35

OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Synchronous Single-Streamabcd70140210280350SE +/- 3.57, N = 15SE +/- 3.02, N = 15284.61329.87319.51323.23
OpenBenchmarking.orgms/batch, Fewer Is BetterNeural Magic DeepSparse 1.3.2Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Synchronous Single-Streamabcd60120180240300Min: 297.77 / Avg: 319.51 / Max: 331.41Min: 298.85 / Avg: 323.23 / Max: 331.47

42 Results Shown

Unvanquished:
  1920 x 1080 - High
  1920 x 1200 - High
  1920 x 1080 - Ultra
  1920 x 1200 - Ultra
  1920 x 1080 - Medium
  1920 x 1200 - Medium
Neural Magic DeepSparse:
  NLP Document Classification, oBERT base uncased on IMDB - Asynchronous Multi-Stream:
    items/sec
    ms/batch
  NLP Document Classification, oBERT base uncased on IMDB - Synchronous Single-Stream:
    items/sec
    ms/batch
  NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Asynchronous Multi-Stream:
    items/sec
    ms/batch
  NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Synchronous Single-Stream:
    items/sec
    ms/batch
  NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Asynchronous Multi-Stream:
    items/sec
    ms/batch
  NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Synchronous Single-Stream:
    items/sec
    ms/batch
  CV Detection, YOLOv5s COCO - Asynchronous Multi-Stream:
    items/sec
    ms/batch
  CV Detection, YOLOv5s COCO - Synchronous Single-Stream:
    items/sec
    ms/batch
  CV Classification, ResNet-50 ImageNet - Asynchronous Multi-Stream:
    items/sec
    ms/batch
  CV Classification, ResNet-50 ImageNet - Synchronous Single-Stream:
    items/sec
    ms/batch
  NLP Text Classification, DistilBERT mnli - Asynchronous Multi-Stream:
    items/sec
    ms/batch
  NLP Text Classification, DistilBERT mnli - Synchronous Single-Stream:
    items/sec
    ms/batch
  CV Segmentation, 90% Pruned YOLACT Pruned - Asynchronous Multi-Stream:
    items/sec
    ms/batch
  CV Segmentation, 90% Pruned YOLACT Pruned - Synchronous Single-Stream:
    items/sec
    ms/batch
  NLP Text Classification, BERT base uncased SST2 - Asynchronous Multi-Stream:
    items/sec
    ms/batch
  NLP Text Classification, BERT base uncased SST2 - Synchronous Single-Stream:
    items/sec
    ms/batch
  NLP Token Classification, BERT base uncased conll2003 - Asynchronous Multi-Stream:
    items/sec
    ms/batch
  NLP Token Classification, BERT base uncased conll2003 - Synchronous Single-Stream:
    items/sec
    ms/batch

a

Kernel Notes: Transparent Huge Pages: madvise
Processor Notes: Scaling Governor: intel_pstate powersave (EPP: balance_performance) - CPU Microcode: 0xa4 - Thermald 2.5.1
Python Notes: Python 3.10.7
Security Notes: 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 + 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 31 January 2023 20:07 by user phoronix.

b

Kernel Notes: Transparent Huge Pages: madvise
Processor Notes: Scaling Governor: intel_pstate powersave (EPP: balance_performance) - CPU Microcode: 0xa4 - Thermald 2.5.1
Python Notes: Python 3.10.7
Security Notes: 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 + 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 31 January 2023 20:44 by user phoronix.

c

Kernel Notes: Transparent Huge Pages: madvise
Processor Notes: Scaling Governor: intel_pstate powersave (EPP: balance_performance) - CPU Microcode: 0xa4 - Thermald 2.5.1
Python Notes: Python 3.10.7
Security Notes: 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 + 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 31 January 2023 21:05 by user phoronix.

d

Processor: Intel Core i7-1165G7 @ 4.70GHz (4 Cores / 8 Threads), Motherboard: Dell 0GG9PT (3.11.0 BIOS), Chipset: Intel Tiger Lake-LP, Memory: 16GB, Disk: Kioxia KBG40ZNS256G NVMe 256GB, Graphics: Intel Xe TGL GT2 15GB (1300MHz), Audio: Realtek ALC289, Network: Intel Wi-Fi 6 AX201

OS: Ubuntu 22.10, Kernel: 5.19.0-29-generic (x86_64), Desktop: GNOME Shell 43.0, Display Server: X Server + Wayland, OpenGL: 4.6 Mesa 22.2.1, Vulkan: 1.3.224, Compiler: GCC 12.2.0, File-System: ext4, Screen Resolution: 1920x1200

Kernel Notes: Transparent Huge Pages: madvise
Processor Notes: Scaling Governor: intel_pstate powersave (EPP: balance_performance) - CPU Microcode: 0xa4 - Thermald 2.5.1
Python Notes: Python 3.10.7
Security Notes: 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 + 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 1 February 2023 08:01 by user phoronix.