Intel Core i9-10980XE testing with a ASRock X299 Steel Legend (P1.50 BIOS) and llvmpipe on Ubuntu 22.04 via the Phoronix Test Suite.
Compare your own system(s) to this result file with the
Phoronix Test Suite by running the command:
phoronix-test-suite benchmark 2403167-PTS-DGGG234526
dggg,
"Neural Magic DeepSparse 1.7 - Model: CV Classification, ResNet-50 ImageNet - Scenario: Synchronous Single-Stream",
Lower Results Are Better
"a",7.5655,7.5194,7.4467
"b",
"Neural Magic DeepSparse 1.7 - Model: CV Classification, ResNet-50 ImageNet - Scenario: Synchronous Single-Stream",
Higher Results Are Better
"a",131.9729,132.7713,134.0805
"b",
"Neural Magic DeepSparse 1.7 - Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Stream",
Lower Results Are Better
"a",409.0273,413.5935,409.4113
"b",
"Neural Magic DeepSparse 1.7 - Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Stream",
Higher Results Are Better
"a",22.001,21.7581,21.9804
"b",
"Neural Magic DeepSparse 1.7 - Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Stream",
Higher Results Are Better
"a",18.4556,18.229,18.2957
"b",
"Neural Magic DeepSparse 1.7 - Model: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Scenario: Synchronous Single-Stream",
Higher Results Are Better
"a",145.923,142.6011,143.4329
"b",
"Neural Magic DeepSparse 1.7 - Model: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Scenario: Synchronous Single-Stream",
Lower Results Are Better
"a",6.8434,7.0035,6.9628
"b",
"Neural Magic DeepSparse 1.7 - Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Synchronous Single-Stream",
Higher Results Are Better
"a",21.6868,21.719,21.3973
"b",
"Neural Magic DeepSparse 1.7 - Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Synchronous Single-Stream",
Lower Results Are Better
"a",46.0939,46.0249,46.7184
"b",
"SVT-AV1 2.0 - Encoder Mode: Preset 13 - Input: Bosphorus 1080p",
Higher Results Are Better
"a",199.998,199.104,200.966
"b",
"Neural Magic DeepSparse 1.7 - Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Stream",
Lower Results Are Better
"a",486.2755,493.2904,489.006
"b",
"Neural Magic DeepSparse 1.7 - Model: ResNet-50, Baseline - Scenario: Synchronous Single-Stream",
Lower Results Are Better
"a",7.5593,7.4745,7.7999
"b",
"Neural Magic DeepSparse 1.7 - Model: ResNet-50, Baseline - Scenario: Synchronous Single-Stream",
Higher Results Are Better
"a",132.0864,133.5727,128.0144
"b",
"Timed Linux Kernel Compilation 6.8 - Build: defconfig",
Lower Results Are Better
"a",75.659,73.652,73.769
"b",
"SVT-AV1 2.0 - Encoder Mode: Preset 8 - Input: Bosphorus 1080p",
Higher Results Are Better
"a",51.375,50.555,51.337
"b",
"Neural Magic DeepSparse 1.7 - Model: ResNet-50, Sparse INT8 - Scenario: Synchronous Single-Stream",
Higher Results Are Better
"a",558.8584,573.0228,553.268
"b",
"Neural Magic DeepSparse 1.7 - Model: ResNet-50, Sparse INT8 - Scenario: Synchronous Single-Stream",
Lower Results Are Better
"a",1.7802,1.7355,1.7986
"b",
"Neural Magic DeepSparse 1.7 - Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Stream",
Higher Results Are Better
"a",18.4688,18.221,18.2421
"b",
"SVT-AV1 2.0 - Encoder Mode: Preset 12 - Input: Bosphorus 4K",
Higher Results Are Better
"a",81.616,81.595,82.815
"b",
"Neural Magic DeepSparse 1.7 - Model: NLP Text Classification, DistilBERT mnli - Scenario: Synchronous Single-Stream",
Lower Results Are Better
"a",12.2636,12.4417,11.5324,11.4418,12.1495,12.1825,11.3061,11.574,11.4237,11.9526,11.2535,11.1499,12.3085,11.2865,11.0286
"b",
"SVT-AV1 2.0 - Encoder Mode: Preset 4 - Input: Bosphorus 1080p",
Higher Results Are Better
"a",8.484,8.595,8.579
"b",
"Neural Magic DeepSparse 1.7 - Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Stream",
Lower Results Are Better
"a",485.0268,489.7974,493.3003
"b",
"SVT-AV1 2.0 - Encoder Mode: Preset 12 - Input: Bosphorus 1080p",
Higher Results Are Better
"a",170.396,172.856,166.811
"b",
"Neural Magic DeepSparse 1.7 - Model: NLP Text Classification, DistilBERT mnli - Scenario: Synchronous Single-Stream",
Higher Results Are Better
"a",81.478,80.3161,86.6446,87.3273,82.2421,82.0254,88.3767,86.3329,87.4689,83.5995,88.7885,89.6137,81.1856,88.529,90.5985
"b",
"Neural Magic DeepSparse 1.7 - Model: BERT-Large, NLP Question Answering, Sparse INT8 - Scenario: Synchronous Single-Stream",
Higher Results Are Better
"a",58.3778,59.4375,54.5459,58.1435,57.5183,55.3076,53.6212,57.9124,53.5399,55.1765,52.628,56.3096,51.6762,53.7694,50.0358
"b",
"srsRAN Project 23.10.1-20240219 - Test: PDSCH Processor Benchmark, Throughput Total",
Higher Results Are Better
"a",5536.4,5565.5,5536.2
"b",
"Neural Magic DeepSparse 1.7 - Model: BERT-Large, NLP Question Answering, Sparse INT8 - Scenario: Asynchronous Multi-Stream",
Higher Results Are Better
"a",209.3901,209.2325,206.2167
"b",
"Neural Magic DeepSparse 1.7 - Model: BERT-Large, NLP Question Answering, Sparse INT8 - Scenario: Asynchronous Multi-Stream",
Lower Results Are Better
"a",42.9589,42.9914,43.5948
"b",
"Neural Magic DeepSparse 1.7 - Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Synchronous Single-Stream",
Lower Results Are Better
"a",62.434,62.6259,64.4449
"b",
"Neural Magic DeepSparse 1.7 - Model: Llama2 Chat 7b Quantized - Scenario: Synchronous Single-Stream",
Lower Results Are Better
"a",124.8471,122.7848,122.6119
"b",
"Parallel BZIP2 Compression 1.1.13 - FreeBSD-13.0-RELEASE-amd64-memstick.img Compression",
Lower Results Are Better
"a",6.217068,5.919299,6.093497
"b",
"Neural Magic DeepSparse 1.7 - Model: Llama2 Chat 7b Quantized - Scenario: Synchronous Single-Stream",
Higher Results Are Better
"a",8.0081,8.1425,8.1541
"b",
"Neural Magic DeepSparse 1.7 - Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Synchronous Single-Stream",
Higher Results Are Better
"a",16.0146,15.9653,15.5149
"b",
"Neural Magic DeepSparse 1.7 - Model: BERT-Large, NLP Question Answering, Sparse INT8 - Scenario: Synchronous Single-Stream",
Lower Results Are Better
"a",17.1174,16.8128,18.3204,17.186,17.3726,18.0678,18.6361,17.2539,18.6649,18.1099,18.9879,17.7455,19.338,18.5845,19.9722
"b",
"SVT-AV1 2.0 - Encoder Mode: Preset 4 - Input: Bosphorus 4K",
Higher Results Are Better
"a",3.22,3.209,3.231
"b",
"Neural Magic DeepSparse 1.7 - Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Stream",
Lower Results Are Better
"a",35.5771,35.2009,35.5984
"b",
"Neural Magic DeepSparse 1.7 - Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Stream",
Higher Results Are Better
"a",252.8154,255.4901,252.6615
"b",
"Neural Magic DeepSparse 1.7 - Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Synchronous Single-Stream",
Lower Results Are Better
"a",62.0155,62.6219,64.2325
"b",
"SVT-AV1 2.0 - Encoder Mode: Preset 13 - Input: Bosphorus 4K",
Higher Results Are Better
"a",82.613,82.49,83.408
"b",
"Neural Magic DeepSparse 1.7 - Model: ResNet-50, Baseline - Scenario: Asynchronous Multi-Stream",
Higher Results Are Better
"a",258.8235,258.0103,258.9618
"b",
"Neural Magic DeepSparse 1.7 - Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Synchronous Single-Stream",
Higher Results Are Better
"a",16.1225,15.9662,15.566
"b",
"Neural Magic DeepSparse 1.7 - Model: ResNet-50, Baseline - Scenario: Asynchronous Multi-Stream",
Lower Results Are Better
"a",34.7505,34.8601,34.732
"b",
"Neural Magic DeepSparse 1.7 - Model: CV Detection, YOLOv5s COCO, Sparse INT8 - Scenario: Synchronous Single-Stream",
Higher Results Are Better
"a",84.7549,82.9407,82.9691
"b",
"Neural Magic DeepSparse 1.7 - Model: CV Detection, YOLOv5s COCO, Sparse INT8 - Scenario: Synchronous Single-Stream",
Lower Results Are Better
"a",11.7883,12.0466,12.0422
"b",
"Google Draco 1.5.6 - Model: Church Facade",
Lower Results Are Better
"a",8553,8573,8574
"b",
"Neural Magic DeepSparse 1.7 - Model: Llama2 Chat 7b Quantized - Scenario: Asynchronous Multi-Stream",
Higher Results Are Better
"a",2.6605,2.6688,2.6658
"b",
"SVT-AV1 2.0 - Encoder Mode: Preset 8 - Input: Bosphorus 4K",
Higher Results Are Better
"a",31.392,31.89,31.549
"b",
"Neural Magic DeepSparse 1.7 - Model: ResNet-50, Sparse INT8 - Scenario: Asynchronous Multi-Stream",
Higher Results Are Better
"a",1226.8185,1223.9024,1229.1039
"b",
"Neural Magic DeepSparse 1.7 - Model: ResNet-50, Sparse INT8 - Scenario: Asynchronous Multi-Stream",
Lower Results Are Better
"a",7.3169,7.3337,7.3034
"b",
"Neural Magic DeepSparse 1.7 - Model: Llama2 Chat 7b Quantized - Scenario: Asynchronous Multi-Stream",
Lower Results Are Better
"a",3228.954,3220.5042,3222.9056
"b",
"Timed Linux Kernel Compilation 6.8 - Build: allmodconfig",
Lower Results Are Better
"a",945.632,943.144,942.097
"b",
"Neural Magic DeepSparse 1.7 - Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Stream",
Lower Results Are Better
"a",60.4359,60.3898,60.0747
"b",
"Neural Magic DeepSparse 1.7 - Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Stream",
Higher Results Are Better
"a",148.8629,148.9777,149.7586
"b",
"Neural Magic DeepSparse 1.7 - Model: CV Detection, YOLOv5s COCO, Sparse INT8 - Scenario: Asynchronous Multi-Stream",
Higher Results Are Better
"a",106.6823,106.7373,107.1074
"b",
"Neural Magic DeepSparse 1.7 - Model: CV Detection, YOLOv5s COCO, Sparse INT8 - Scenario: Asynchronous Multi-Stream",
Lower Results Are Better
"a",84.3406,84.2972,84.0053
"b",
"Google Draco 1.5.6 - Model: Lion",
Lower Results Are Better
"a",6059,6066,6053
"b",
"Primesieve 12.1 - Length: 1e12",
Lower Results Are Better
"a",11.09,11.096,11.114
"b",
"Neural Magic DeepSparse 1.7 - Model: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Scenario: Asynchronous Multi-Stream",
Higher Results Are Better
"a",456.0336,456.5353,453.5072
"b",
"Neural Magic DeepSparse 1.7 - Model: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Scenario: Asynchronous Multi-Stream",
Lower Results Are Better
"a",19.7122,19.6902,19.8218
"b",
"Primesieve 12.1 - Length: 1e13",
Lower Results Are Better
"a",128.885,128.886,128.858
"b",
"srsRAN Project 23.10.1-20240219 - Test: PUSCH Processor Benchmark, Throughput Thread",
Higher Results Are Better
"a",134.5,134.8,134.2
"srsRAN Project 23.10.1-20240219 - Test: PDSCH Processor Benchmark, Throughput Thread",
Higher Results Are Better
"a",399.3,399.5,401.8
"srsRAN Project 23.10.1-20240219 - Test: PUSCH Processor Benchmark, Throughput Total",
Higher Results Are Better
"a",1101.9,1107.5,1106.9