1 | ID | Submitter | System | Processor | # | Accelerator | # | Software | Benchmark results (minutes) | Details | Code | Notes | ||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2 | Task | Image classification | Image segmentation (medical) | Object detection, light-weight | Object detection, heavy-weight | Speech recognition | LLM | NLP | Recom- mendation | |||||||||||
3 | Dataset | ImageNet | KiTS19 | OpenImages | COCO | LibriSpeech | C4 | Wikipedia | Criteo 4TB | |||||||||||
4 | Model | ResNet | 3D U-Net | RetinaNet | Mask R-CNN | RNN-T | GPT3 | BERT-large | DLRM-dcnv2 | |||||||||||
5 | Available Cloud | |||||||||||||||||||
6 | 3.0-2000 | NVIDIA+CoreWeave | coreweave_hgxh100_n192_ngc23.04_pytorch | Intel(R) Xeon(R) Platinum 8462Y+ | 384 | NVIDIA H100-SXM5-80GB | 1536 | NVIDIA NeMo Megatron Release 23.04 | 23.611 | details | code | |||||||||
7 | 3.0-2001 | NVIDIA+CoreWeave | coreweave_hgxh100_n384_ngc23.04_pytorch | Intel(R) Xeon(R) Platinum 8462Y+ | 768 | NVIDIA H100-SXM5-80GB | 3072 | PyTorch NVIDIA Release 23.04 | 0.134 | details | code | |||||||||
8 | 3.0-2002 | NVIDIA+CoreWeave | coreweave_hgxh100_n448_ngc23.04_mxnet | Intel(R) Xeon(R) Platinum 8462Y+ | 896 | NVIDIA H100-SXM5-80GB | 3584 | MXNet NVIDIA Release 23.04 | 0.183 | details | code | |||||||||
9 | 3.0-2003 | NVIDIA+CoreWeave | coreweave_hgxh100_n448_ngc23.04_pytorch | Intel(R) Xeon(R) Platinum 8462Y+ | 896 | NVIDIA H100-SXM5-80GB | 3584 | NVIDIA NeMo Megatron Release 23.04 | 10.940 | details | code | |||||||||
10 | 3.0-2004 | NVIDIA+CoreWeave | coreweave_hgxh100_n96_ngc23.04_pytorch | Intel(R) Xeon(R) Platinum 8462Y+ | 192 | NVIDIA H100-SXM5-80GB | 768 | NVIDIA NeMo Megatron Release 23.04 | 45.606 | details | code | |||||||||
11 | Available On premise | |||||||||||||||||||
12 | 3.0-2004 | ASUSTeK | ESC4000-E11-4xA100-PCIE-80GB | Intel(R) Xeon(R) Platinum 8462Y+ | 2 | NVIDIA A100-PCIe-80GB | 4 | NVIDIA Release 23.04 MxNet, pytorch | 57.711 | 47.715 | 171.378 | 86.668 | 63.216 | 45.793 | details | code | TDP=300W per GPU | |||
13 | 3.0-2005 | ASUSTeK | ESC4000-E11-4xA100-PCIE-80GB-NVBridge | Intel(R) Xeon(R) Platinum 8462Y+ | 2 | NVIDIA A100-PCIe-80GB | 4 | NVIDIA Release 23.04 MxNet, pytorch | 57.450 | 42.126 | 174.366 | 86.338 | 64.149 | 42.682 | details | code | TDP=300W per GPU | |||
14 | 3.0-2006 | ASUSTeK | ESC8000A-E12-8xH100-PCIE-80GB | AMD EPYC 9654 96-Core | 2 | NVIDIA H100-PCIe-80GB | 8 | NVIDIA Release 23.04 MxNet, pytorch | 20.764 | 19.232 | 54.928 | 28.579 | 23.296 | 10.510 | details | code | TDP=350W per GPU | |||
15 | 3.0-2007 | H3C | R4900G6x2A30-PCIE-24GB | Intel(R) Xeon(R) Platinum 8490H CPU @ 1.90GHz | 2 | NVIDIA A30-PCIE-24GB | 2 | NGC MXNet 23.04 , NGC PyTorch 23.04 , NGC TensorFlow 23.04-tf1 | 237.273 | details | code | N/A;N/A | ||||||||
16 | 3.0-2008 | H3C | R5300G6x8A30-PCIE-24GB | Intel(R) Xeon(R) Platinum 8458P | 2 | NVIDIA A30-PCIE-24GB | 8 | NGC MXNet 23.04 , NGC PyTorch 23.04 , NGC TensorFlow 23.04-tf1 | 60.061 | 87.966 | 208.635 | 78.639 | details | code | N/A;N/A | |||||
17 | 3.0-2009 | H3C | R5350G6x8A30-PCIE-24GB | AMD EPYC 9754 128-Core Processor | 2 | NVIDIA A30-PCIE-24GB | 8 | NGC MXNet 23.04 , NGC PyTorch 23.04 , NGC TensorFlow 23.04-tf1 | 58.915 | 74.246 | 192.350 | 75.150 | details | code | N/A;N/A | |||||
18 | 3.0-2010 | H3C | R5350G6x8A30-PCIE-24GB | AMD EPYC 9754 128-Core Processor | 2 | NVIDIA A30-PCIE-24GB | 10 | NGC MXNet 23.04 , NGC PyTorch 23.04 , NGC TensorFlow 23.04-tf1 | 47.360 | 142.884 | 68.652 | details | code | N/A;N/A | ||||||
19 | 3.0-2011 | Intel | 16-nodes-SPR-pytorch | Intel(R) Xeon(R) Platinum 8480+ @ 2.00GHz | 32 | N/A | 0 | Pytorch | 88.173 | 232.405 | 47.929 | details | code | |||||||
20 | 3.0-2012 | Intel | 8-nodes-SPR-pytorch | Intel(R) Xeon(R) Platinum 8480+ @ 2.00GHz | 16 | N/A | 0 | Pytorch | 88.103 | details | code | |||||||||
21 | 3.0-2013 | Intel-HabanaLabs | HLS-Gaudi2-N32-PT | Intel(R) Xeon(R) Platinum 8380 | 64 | Habana Gaudi2 | 256 | PyTorch 1.13.1a0 | 442.578 | details | code | |||||||||
22 | 3.0-2014 | Intel-HabanaLabs | HLS-Gaudi2-N48-PT | Intel(R) Xeon(R) Platinum 8380 | 96 | Habana Gaudi2 | 384 | PyTorch 1.13.1a0 | 311.945 | details | code | |||||||||
23 | 3.0-2015 | Intel-HabanaLabs | HLS-Gaudi2-N8-PT | Intel(R) Xeon(R) Platinum 8380 | 16 | Habana Gaudi2 | 64 | PyTorch 1.13.1a0 | 2.103 | details | code | |||||||||
24 | 3.0-2016 | Intel-HabanaLabs | HLS-Gaudi2-PT | Intel(R) Xeon(R) Platinum 8380 | 2 | Habana Gaudi2 | 8 | PyTorch 1.13.1a0 | 16.460 | 20.516 | 14.794 | details | code | |||||||
25 | 3.0-2017 | Intel-HabanaLabs | HLS-Gaudi2-TF | Intel(R) Xeon(R) Platinum 8380 | 2 | Habana Gaudi2 | 8 | TensorFlow 2.12.0 | 15.991 | 14.116 | details | code | ||||||||
26 | 3.0-2018 | Lenovo | Lenovo ThinkSystem SR670 V2 Server with 4x 40GB SXM4 A100 | Intel(R) Xeon(R) Platinum 8360Y CPU @ 2.40GHz | 2 | NVIDIA A100-SXM4-40GB | 4 | NGC MLPerf v2.1/2.0 | 61.999 | 91.941 | details | code | TDP=400W per GPU, Air Cooled | |||||||
27 | 3.0-2019 | Lenovo | Lenovo ThinkSystem SR670 V2 Server with 8x 80GB PCIe A100 | Intel(R) Xeon(R) Platinum 8360Y CPU @ 2.40GHz | 2 | NVIDIA A100-PCIe-80GB | 8 | NGC MLPerf v2.1/2.0 | 32.744 | 47.175 | details | code | TDP=300W per GPU, Air Cooling | |||||||
28 | 3.0-2020 | Lenovo | Lenovo ThinkSystem SR670 V2 Server with 8x 80GB PCIe H100 | Intel(R) Xeon(R) Platinum 8360Y CPU @ 2.40GHz | 2 | NVIDIA H100-PCIe | 8 | NGC MLPerf Training v3.0 | 24.223 | 21.374 | 35.736 | 27.961 | details | code | TDP=310W per GPU, Air Cooling | |||||
29 | 3.0-2021 | Supermicro | AS-4125GS-TNRT | AMD EPYC 9554 64-Core Processor | 2 | NVIDIA H100-PCIe-80GB | 8 | CUDA 12.0 | 21.043 | 18.270 | 56.118 | 28.693 | 27.180 | 10.630 | details | code | NVIDIA H100-PCIe-80GB | |||
30 | 3.0-2022 | Supermicro | AS-8125GS-TNHR | AMD EPYC 9634 | 2 | NVIDIA H100-SXM5-80GB | 8 | MXNet NVIDIA Release 23.04, PyTorch NVIDIA Release 23.04, HugeCTR NVIDIA Release 23.04 | 13.603 | 12.989 | 37.501 | 20.439 | 19.235 | 5.389 | details | code | ||||
31 | 3.0-2023 | Supermicro | SYS-421GU-TNX | Intel(R) Xeon(R) Platinum 8460H | 2 | NVIDIA H100-SXM5-80GB | 4 | MXNet NVIDIA Release 23.04, PyTorch NVIDIA Release 23.04, HugeCTR NVIDIA Release 23.04 | 27.382 | 22.786 | 72.593 | 40.254 | 27.890 | 11.222 | 8.821 | details | code | |||
32 | 3.0-2024 | Supermicro | SYS-421GU-TNXR | Intel(R) Xeon(R) Platinum 8480+ | 2 | NVIDIA H100-SXM5-80GB | 4 | MxNet NVIDIA Release 22.04 | 27.121 | 22.643 | 73.874 | 30.285 | 11.214 | details | code | |||||
33 | 3.0-2025 | Supermicro | SYS-820GH-TNR2 | Intel(R) Xeon(R) Platinum 8380 | 2 | Habana Gaudi2 | 8 | tensorflow 2.11.0 | 16.427 | 13.951 | details | code | ||||||||
34 | 3.0-2026 | Supermicro | SYS-821GE-TNHR | Intel(R) Xeon(R) Platinum 8490H | 2 | NVIDIA H100-SXM5-80GB | 8 | MXNet NVIDIA Release 23.04, PyTorch NVIDIA Release 23.04, HugeCTR NVIDIA Release 23.04 | 13.501 | 12.037 | 37.353 | 21.493 | 17.919 | 5.383 | details | code | ||||
35 | 3.0-2027 | Dell | 16xXE8545x4A100-SXM-40GB | AMD EPYC 7713 64-Core Processor | 32 | NVIDIA A100-SXM-40GB | 64 | NGC MXNet 22.09 , NGC PyTorch 22.09 , NGC TensorFlow 22.09-tf1 | 5.532 | 10.301 | 3.335 | details | code | GPU TDP:400W;N/A | ||||||
36 | 3.0-2028 | Dell | 2xR750xax4A100-PCIE-80GB | Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz | 4 | NVIDIA A100-PCIe-80GB | 8 | NGC MXNet 22.09 , NGC PyTorch 22.09 , NGC TensorFlow 22.09-tf1 | 35.187 | 31.374 | details | code | GPU TDP:300W;N/A | |||||||
37 | 3.0-2029 | Dell | 2xR750xax4A100-PCIE-80GB-1opa | Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz | 4 | NVIDIA A100-PCIe-80GB | 8 | NGC MXNet 22.09 , NGC PyTorch 22.09 , NGC TensorFlow 22.09-tf1 | 36.459 | 31.387 | details | code | GPU TDP:300W;N/A | |||||||
38 | 3.0-2030 | Dell | 2xR750xax4A100-PCIE-80GB-2opa | Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz | 4 | NVIDIA A100-PCIe-80GB | 8 | NGC MXNet 22.09 , NGC PyTorch 22.09 , NGC TensorFlow 22.09-tf1 | 35.789 | 28.891 | details | code | GPU TDP:300W;N/A | |||||||
39 | 3.0-2031 | Dell | 2xXE8545x4A100-SXM-40GB | AMD EPYC 7713 64-Core Processor | 4 | NVIDIA A100-SXM-40GB | 8 | NGC MXNet 22.09 , NGC PyTorch 22.09 , NGC TensorFlow 22.09-tf1 | 33.451 | 45.109 | 24.176 | details | code | GPU TDP:400W;N/A | ||||||
40 | 3.0-2032 | Dell | 2xXE8545x4A100-SXM-80GB | AMD EPYC 7713 64-Core Processor | 4 | NVIDIA A100-SXM-80GB CTS | 8 | NGC MXNet 22.09 , NGC PyTorch 22.09 , NGC TensorFlow 22.09-tf1 | 28.805 | 44.614 | 23.586 | details | code | GPU TDP:500W;N/A | ||||||
41 | 3.0-2033 | Dell | 2xXE9680x8A100-SXM-80GB | Intel(R) Xeon(R) Platinum 8470 | 4 | NVIDIA A100-SXM-80GB CTS | 16 | NGC MXNet 23.04, NGC Pytorch 23.04, NGC HugeCTR 23.04 | 14.397 | 10.961 | 41.627 | details | code | GPU TDP:500W;N/A | ||||||
42 | 3.0-2034 | Dell | 2xXE9680x8H100-SXM-80GB | Intel(R) Xeon(R) Platinum 8470 | 4 | NVIDIA H100-SXM5-80GB | 16 | NGC MXNet 23.04, NGC PyTorch 23.04, NGC HugeCTR 23.04 | 7.847 | 7.619 | details | code | GPU TDP:700W;N/A | |||||||
43 | 3.0-2035 | Dell | 32xXE8545x4A100-SXM-40GB | AMD EPYC 7713 64-Core Processor | 64 | NVIDIA A100-SXM-40GB | 128 | NGC MXNet 22.09 , NGC PyTorch 22.09 , NGC TensorFlow 22.09-tf1 | 3.260 | 2.655 | details | code | GPU TDP:400W;N/A | |||||||
44 | 3.0-2036 | Dell | 4xR750xax4A100-PCIE-80GB | Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz | 8 | NVIDIA A100-PCIe-80GB | 16 | NGC MXNet 22.09 , NGC PyTorch 22.09 , NGC TensorFlow 22.09-tf1 | 19.478 | 17.588 | details | code | GPU TDP:300W;N/A | |||||||
45 | 3.0-2037 | Dell | 4xR750xax4A100-PCIE-80GB-1opa | Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz | 8 | NVIDIA A100-PCIe-80GB | 16 | NGC MXNet 22.09 , NGC PyTorch 22.09 , NGC TensorFlow 22.09-tf1 | 20.651 | 22.324 | details | code | GPU TDP:300W;N/A | |||||||
46 | 3.0-2038 | Dell | 4xR750xax4A100-PCIE-80GB-2opa | Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz | 8 | NVIDIA A100-PCIe-80GB | 16 | NGC MXNet 22.09 , NGC PyTorch 22.09 , NGC TensorFlow 22.09-tf1 | 19.088 | 16.947 | details | code | GPU TDP:300W;N/A | |||||||
47 | 3.0-2039 | Dell | 4xXE8545x4A100-SXM-40GB | AMD EPYC 7713 64-Core Processor | 8 | NVIDIA A100-SXM-40GB | 16 | NGC MXNet 22.09 , NGC PyTorch 22.09 , NGC TensorFlow 22.09-tf1 | 17.171 | 25.459 | 17.874 | details | code | GPU TDP:400W;N/A | ||||||
48 | 3.0-2040 | Dell | 4xXE8545x4A100-SXM-80GB | AMD EPYC 7713 64-Core Processor | 8 | NVIDIA A100-SXM-80GB CTS | 16 | NGC MXNet 22.09 , NGC PyTorch 22.09 , NGC TensorFlow 22.09-tf1 | 15.933 | 23.654 | 16.006 | details | code | GPU TDP:500W;N/A | ||||||
49 | 3.0-2041 | Dell | 8xR750xax4A100-PCIE-80GB | Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz | 16 | NVIDIA A100-PCIe-80GB | 32 | NGC MXNet 22.09 , NGC PyTorch 22.09 , NGC TensorFlow 22.09-tf1 | 10.058 | 9.385 | details | code | GPU TDP:300W;N/A | |||||||
50 | 3.0-2042 | Dell | 8xR750xax4A100-PCIE-80GB-1opa | Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz | 16 | NVIDIA A100-PCIe-80GB | 32 | NGC MXNet 22.09 , NGC PyTorch 22.09 , NGC TensorFlow 22.09-tf1 | 10.459 | 9.638 | details | code | GPU TDP:300W;N/A | |||||||
51 | 3.0-2043 | Dell | 8xR750xax4A100-PCIE-80GB-2opa | Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz | 16 | NVIDIA A100-PCIe-80GB | 32 | NGC MXNet 22.09 , NGC PyTorch 22.09 , NGC TensorFlow 22.09-tf1 | 10.224 | 8.927 | details | code | GPU TDP:300W;N/A | |||||||
52 | 3.0-2044 | Dell | 8xXE8545x4A100-SXM-40GB | AMD EPYC 7713 64-Core Processor | 16 | NVIDIA A100-SXM-40GB | 32 | NGC MXNet 22.09 , NGC PyTorch 22.09 , NGC TensorFlow 22.09-tf1 | 9.830 | 15.633 | 6.898 | details | code | GPU TDP:400W;N/A | ||||||
53 | 3.0-2045 | Dell | R750xax4A100-PCIE-80GB | Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz | 2 | NVIDIA A100-PCIe-80GB | 4 | NGC MXNet 23.04 , NGC PyTorch 23.04, NGC HugeCTR 23.04 | 61.357 | 48.054 | 176.844 | 81.860 | 64.049 | 51.590 | details | code | GPU TDP:300W;N/A | |||
54 | 3.0-2046 | Dell | R750xax4A100-PCIE-80GB-NVBRIDGE | Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz | 2 | NVIDIA A100-PCIe-80GB | 4 | NGC MXNet 22.09 , NGC PyTorch 22.09 , NGC TensorFlow 22.09-tf1 | 61.275 | 207.566 | 66.200 | 48.957 | details | code | GPU TDP:300W;N/A | |||||
55 | 3.0-2047 | Dell | R750xax4H100-PCIE-80GB | Intel(R) Xeon(R) Gold 6338 | 2 | NVIDIA H100-PCIe-80GB | 4 | NGC MXNet 23.04 , NGC PyTorch 23.04, NGC HugeCTR 23.04 | 45.147 | 32.885 | 114.693 | 62.092 | 51.924 | details | code | GPU TDP:310W;N/A | ||||
56 | 3.0-2048 | Dell | R760xax4H100-PCIE-80GB | Intel(R) Xeon(R) Platinum 8480+ | 2 | NVIDIA H100-PCIe-80GB | 4 | NGC MXNet 23.04 , NGC PyTorch 23.04, NGC HugeCTR 23.04 | 39.918 | 31.999 | 107.463 | 55.181 | 45.924 | details | code | GPU TDP:350W;N/A | ||||
57 | 3.0-2049 | Dell | XE8545x4A100-SXM-40GB | AMD EPYC 7713 64-Core Processor | 2 | NVIDIA A100-SXM-40GB | 4 | NGC MXNet 22.09 , NGC PyTorch 22.09 , NGC TensorFlow 22.09-tf1 | 87.739 | 38.399 | details | code | GPU TDP:400W;N/A | |||||||
58 | 3.0-2050 | Dell | XE8545x4A100-SXM-80GB | AMD EPYC 7763 64-Core Processor | 2 | NVIDIA A100-SXM-80GB CTS | 4 | NGC MXNet 22.09 , NGC PyTorch 22.09 , NGC TensorFlow 22.09-tf1 | 54.231 | 47.253 | 222.199 | 83.712 | 55.086 | 32.792 | details | code | GPU TDP:500W;N/A | |||
59 | 3.0-2051 | Dell | XE8640x4H100-SXM-80GB | Intel(R) Xeon(R) Platinum 8468 | 2 | NVIDIA H100-SXM5-80GB | 4 | NGC MXNet 23.04, NGC Pytorch 23.04, NGC HugeCTR 23.04 | 26.524 | 22.270 | 72.601 | 27.958 | 11.010 | 8.112 | details | code | GPU TDP:700W;N/A | |||
60 | 3.0-2052 | Dell | XE9680x8A100-SXM-80GB | Intel(R) Xeon(R) Platinum 8480+ | 2 | NVIDIA A100-SXM-80GB CTS | 8 | NGC MXNet 23.04, NGC Pytorch 23.04, NGC HugeCTR 23.04 | 27.048 | 23.193 | 79.835 | 38.579 | 29.224 | 15.752 | 8.412 | details | code | GPU TDP:500W;N/A | ||
61 | 3.0-2053 | Dell | XE9680x8H100-SXM-80GB | Intel(R) Xeon(R) Platinum 8470 | 2 | NVIDIA H100-SXM5-80GB | 8 | NGC MXNet 23.04, NGC Pytorch 23.04, NGC HugeCTR 23.04 | 13.466 | 12.204 | 37.406 | 19.985 | 16.846 | 5.363 | 4.277 | details | code | GPU TDP:700W;N/A | ||
62 | 3.0-2054 | GIGABYTE | G593-SD0 | Intel(R) Xeon(R) Platinum 8480+ | 2 | NVIDIA H100-SXM5-80GB | 8 | Mxnet | 13.500 | 11.796 | details | code | 700W | |||||||
63 | 3.0-2055 | GIGABYTE | G593-SD0 | Intel(R) Xeon(R) Platinum 8480+ | 2 | NVIDIA H100-SXM5-80GB | 8 | PyTorch | 37.527 | 19.244 | 16.733 | 5.349 | details | code | 700W | |||||
64 | 3.0-2056 | GIGABYTE | G593-SD0 | Intel(R) Xeon(R) Platinum 8480+ | 2 | NVIDIA H100-SXM5-80GB | 8 | hugectr | 4.256 | details | code | 700W | ||||||||
65 | 3.0-2057 | IEI | NF5468M6 | Intel(R) Xeon(R) Platinum 8380 | 2 | NVIDIA A40 | 8 | hugectr | 41.431 | details | code | |||||||||
66 | 3.0-2058 | IEI | NF5468M6 | Intel(R) Xeon(R) Platinum 8380 | 2 | NVIDIA A40 | 8 | mxnet | 66.699 | details | code | |||||||||
67 | 3.0-2059 | IEI | NF5468M6 | Intel(R) Xeon(R) Platinum 8380 | 2 | NVIDIA A40 | 8 | pytorch | 177.112 | 87.885 | 77.536 | details | code | |||||||
68 | 3.0-2060 | Krai | Dell Precision 7920 Tower with 2x A5000 using MxNet 22.04 | Intel(R) Xeon(R) Gold 6240R CPU @ 2.40GHz | 1 | NVIDIA RTX A5000 | 2 | MxNet NVIDIA Release 22.04 | 284.038 | details | code | |||||||||
69 | 3.0-2061 | Krai | Dell Precision 7920 Tower with 2x A5000 using MxNet 22.08 | Intel(R) Xeon(R) Gold 6240R CPU @ 2.40GHz | 1 | NVIDIA RTX A5000 | 2 | MxNet NVIDIA Release 22.08 | 319.092 | details | code | |||||||||
70 | 3.0-2062 | NVIDIA | dgxh100 | Intel(R) Xeon(R) Platinum 8480C | 2 | NVIDIA H100-SXM5-80GB | 8 | MXNet NVIDIA Release 23.04 | 13.601 | 12.103 | details | code | ||||||||
71 | 3.0-2063 | NVIDIA | dgxh100 | Intel(R) Xeon(R) Platinum 8480C | 2 | NVIDIA H100-SXM5-80GB | 8 | NVIDIA Merlin HugeCTR Release 23.04 | 4.184 | details | code | |||||||||
72 | 3.0-2064 | NVIDIA | dgxh100 | Intel(R) Xeon(R) Platinum 8480C | 2 | NVIDIA H100-SXM5-80GB | 8 | PyTorch NVIDIA Release 23.04 | 37.009 | 19.180 | 16.686 | 5.469 | details | code | ||||||
73 | 3.0-2065 | NVIDIA | dgxh100_n16 | Intel(R) Xeon(R) Platinum 8480C | 32 | NVIDIA H100-SXM5-80GB | 128 | NVIDIA Merlin HugeCTR Release 23.04 | 1.613 | details | code | |||||||||
74 | 3.0-2066 | NVIDIA | dgxh100_n48 | Intel(R) Xeon(R) Platinum 8480C | 96 | NVIDIA H100-SXM5-80GB | 384 | PyTorch NVIDIA Release 23.04 | 1.466 | details | code | |||||||||
75 | 3.0-2067 | NVIDIA | dgxh100_n54 | Intel(R) Xeon(R) Platinum 8480C | 108 | NVIDIA H100-SXM5-80GB | 432 | MXNet NVIDIA Release 23.04 | 0.818 | details | code | |||||||||
76 | 3.0-2068 | NVIDIA | dgxh100_n64 | Intel(R) Xeon(R) Platinum 8480C | 128 | NVIDIA H100-SXM5-80GB | 512 | MXNet NVIDIA Release 23.04 | 0.509 | details | code | |||||||||
77 | 3.0-2069 | NVIDIA | dgxh100_n64 | Intel(R) Xeon(R) Platinum 8480C | 128 | NVIDIA H100-SXM5-80GB | 512 | NVIDIA NeMo Framework | 64.264 | details | code | |||||||||
78 | 3.0-2070 | NVIDIA | dgxh100_n64 | Intel(R) Xeon(R) Platinum 8480C | 128 | NVIDIA H100-SXM5-80GB | 512 | PyTorch NVIDIA Release 23.04 | 1.906 | 1.649 | 0.344 | details | code | |||||||
79 | 3.0-2071 | NVIDIA | dgxh100_n8 | Intel(R) Xeon(R) Platinum 8480C | 16 | NVIDIA H100-SXM5-80GB | 64 | MXNet NVIDIA Release 23.04 | 2.664 | details | code | |||||||||
80 | 3.0-2072 | NVIDIA | dgxh100_n8 | Intel(R) Xeon(R) Platinum 8480C | 16 | NVIDIA H100-SXM5-80GB | 64 | NVIDIA Merlin HugeCTR Release 23.04 | 1.760 | details | code | |||||||||
81 | 3.0-2073 | NVIDIA | dgxh100_n8 | Intel(R) Xeon(R) Platinum 8480C | 16 | NVIDIA H100-SXM5-80GB | 64 | PyTorch NVIDIA Release 23.04 | 6.511 | 4.264 | 4.231 | 0.898 | details | code | ||||||
82 | 3.0-2074 | NVIDIA | dgxh100_n9 | Intel(R) Xeon(R) Platinum 8480C | 18 | NVIDIA H100-SXM5-80GB | 72 | MXNet NVIDIA Release 23.04 | 1.853 | details | code | |||||||||
83 | 3.0-2075 | NVIDIA | dgxh100_n96 | Intel(R) Xeon(R) Platinum 8480C | 192 | NVIDIA H100-SXM5-80GB | 768 | MXNet NVIDIA Release 23.04 | 0.369 | details | code | |||||||||
84 | 3.0-2076 | NVIDIA | dgxh100_n96 | Intel(R) Xeon(R) Platinum 8480C | 192 | NVIDIA H100-SXM5-80GB | 768 | NVIDIA NeMo Framework | 44.816 | details | code | |||||||||
85 | 3.0-2077 | NVIDIA | dgxh100_n96 | Intel(R) Xeon(R) Platinum 8480C | 192 | NVIDIA H100-SXM5-80GB | 768 | PyTorch NVIDIA Release 23.04 | 1.511 | 0.253 | details | code | ||||||||
86 | 3.0-2078 | Quanta_Cloud_Technology | D54Q-2U | Intel(R) Xeon(R) Gold 6430 | 2 | NVIDIA H100-PCIe-80GB | 2 | NVIDIA PyTorch/MxNet | 82.589 | 67.351 | 77.394 | 43.720 | details | code | ||||||
87 | 3.0-2079 | xFusion | xFusion FusionServer 2288H V7(6x NVIDIA L4) | Intel(R)Xeon(R)Platinum 8490H | 2 | NVIDIA L4 | 6 | NGC MXNet 23.04 , NGC PyTorch 23.04 | 170.378 | 597.757 | 183.567 | 138.649 | details | code | ||||||
88 | 3.0-2080 | xFusion | xFusion FusionServer G5500 V7(10x NVIDIA A30) | Intel(R) Xeon(R) Platinum 6458Q | 2 | NVIDIA A30 | 10 | NGC MXNet 23.04 , NGC PyTorch 22.09 , NGC PyTorch 23.04 | 49.908 | 145.798 | 63.421 | 49.877 | details | code | ||||||
89 | 3.0-2081 | xFusion | xFusion FusionServer G5500 V7(10x NVIDIA L40) | Intel(R) Xeon(R) Platinum 6458Q | 2 | NVIDIA L40 | 10 | NGC MXNet 23.04 , NGC PyTorch 23.04 | 38.838 | 112.919 | 62.200 | 42.796 | 35.761 | details | code | |||||
90 | 3.0-2082 | xFusion | xFusion FusionServer G5500 V7(8x NVIDIA A30) | Intel(R) Xeon(R) Platinum 6458Q | 2 | NVIDIA A30 | 8 | NGC MXNet 23.04 , NGC PyTorch 22.09 , NGC PyTorch 23.04 | 61.254 | 82.501 | 215.357 | 76.174 | 57.239 | details | code | |||||
91 | 3.0-2083 | xFusion | xFusion FusionServer G5500 V7(8x NVIDIA L40) | Intel(R) Xeon(R) Platinum 6458Q | 2 | NVIDIA L40 | 8 | NGC MXNet 23.04 , NGC PyTorch 23.04 | 48.353 | 34.513 | 175.969 | 77.665 | 54.272 | details | code | |||||
92 | Preview | |||||||||||||||||||
93 | 3.0-2084 | Azure | ND_H100_v5 | Intel(R) Xeon(R) Platinum 8480C | 2 | NVIDIA H100-SXM-80GB | 8 | MXNet NVIDIA Release 23.04 | 13.819 | details | code | |||||||||
94 | 3.0-2085 | Azure | ND_H100_v5 | Intel(R) Xeon(R) Platinum 8480C | 2 | NVIDIA H100-SXM-80GB | 8 | PyTorch NVIDIA Release 23.04 | 37.499 | 19.835 | 17.353 | 5.430 | details | code | ||||||
95 | 3.0-2086 | Fujitsu | PRIMERGY-mxnet | Intel(R) Xeon(R) Gold 6430 | 2 | NVIDIA A100-PCIe-80GB | 10 | mxnet NVIDIA Release 23.04 | 25.831 | details | code | GPUs are installed in an external PCI box. | ||||||||
96 | 3.0-2087 | Fujitsu | PRIMERGY-pytorch | Intel(R) Xeon(R) Gold 6430 | 2 | NVIDIA A100-PCIe-80GB | 10 | pytorch NVIDIA Release 23.04 | 73.229 | details | code | GPUs are installed in an external PCI box. | ||||||||
97 | 3.0-2088 | Quanta_Cloud_Technology | D74H-7U_preview | Intel(R) Xeon(R) Platinum 8490H | 2 | NVIDIA H100-SXM5-80GB | 8 | NVIDIA PyTorch/MxNet | 13.721 | 12.102 | 37.622 | 19.520 | 16.607 | 5.704 | details | code |
1 | ID | Submitter | System | Processor | # | Accelerator | # | Software | Benchmark results (minutes) | Details | Code | Notes | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2 | Task | Image classification | Image segmentation (medical) | Object detection, light-weight | Object detection, heavy-weight | Speech recognition | LLM | NLP | Recom- mendation | |||||||||||||||
3 | Dataset | ImageNet | KiTS19 | OpenImages | COCO | LibriSpeech | Wikipedia | |||||||||||||||||
4 | Model | ResNet | 3D U-Net | RetinaNet | Mask R-CNN | RNN-T | gpt3 | BERT-large | DLRMv2 | |||||||||||||||
5 | Available Onpremise | |||||||||||||||||||||||
6 | 3.0-2089 | Intel | 16-nodes-SPR-pytorch-open | Intel(R) Xeon(R) Platinum 8480+ @ 2.00GHz | 32 | N/A | 0 | Pytorch | 31.0600125 | details | code | We perform LAMB over a fused parameter instead of individual parameters (one fused param for each datatype and weight_decay value). No change in model or any other computation compared to closed division |