CSV spreadsheets and other material for AI accelerator survey papers
This page has a list of some of the data columns/fields that are included in the 2023 IEEE-HPEC paper:
A. Reuther, P. Michaleas, M. Jones, V. Gadepally, S. Samsi and J. Kepner, “Lincoln AI Computing Survey (LAICS) Update,” 2023 IEEE High Performance Extreme Computing Conference (HPEC), 2023, pp. 1-7, [IEEE Xplore doi: 10.1109/HPEC58863.2023.10363568] [ArXiv.org/abs/2310.09145].
For the full dataset in CSV format, please download [peak-accelerators-ieee-hpec-2023.csv].
Company | Product | Label | Peak Perf. (GOPs/ GFLOPs) | Peak Power (W) | Precision | Form Factor | References |
---|---|---|---|---|---|---|---|
Achronix | VectorPath S7t-VG6 | Achronix | 86000 | 300 | int8 | Card | [www.eetimes.com] |
Adapteva | Epiphany-V | Adaptiva | 2050 | 29.26 | fp32 | Chip | [arxiv.org] [doi.org] |
Aimotive | aiWare3 | Aimotive | 100000 | 25 | int8 | Chip | [aimotive.com] |
AIStorm | AIStorm | AIStorm | 2500 | 0.225 | int8 | Chip | [www.eetimes.com] |
Alibaba | HanGuang 800 | Alibaba | 314252 | 157.126 | int8 | Card | [medium.com] |
AlphaIC | RAP-C | AlphaIC | 60000 | 40 | int8 | System | [www.alphaics.ai] |
AlphaIC | RAP-E | AlphaIC | 30000 | 3 | int8 | Chip | [www.eenewsanalog.com] |
Amazon | Inferentia | AWSi1 | 128000 | 200 | int8 | Card | [perspectives.mvdirona.com] [www.cloudmanagementinsider.com] |
AMD | Radeon Instinct MI6 | AMD-MI8 | 8190 | 150 | fp16 | Card | [blog.exxactcorp.com] |
AMD | Radeon Instinct MI60 | AMD-MI60 | 29500 | 300 | fp16 | Card | [www.anandtech.com] |
AMD | MI100 | AMD-MI100 | 92300 | 300 | fp16 | Card | [www.anandtech.com] |
AMD | MI210 | AMD-MI210 | 181000 | 300 | fp16 | Card | [www.anandtech.com] |
AMD | MI250 | AMD-MI250 | 362000 | 550 | fp16 | Card | [www.anandtech.com] |
AMD | MI300 | AMD-MI300 | 766000 | 700 | fp16 | Card | [www.nextplatform.com] |
ARM | Ethos N77 | Ethos | 4100 | 0.8 | int8 | Chip | [fuse.wikichip.org] |
Axelera | Axelera Test Core | Axelera | 39100 | 2.79 | int8 | Chip | [www.eetimes.com] |
Baidu | Baidu Kunlun 200 | Baidu-K1 | 64000 | 160 | fp16 | Chip | [doi.org] [www.eetimes.com] [www.zdnet.com] |
Baidu | Baidu Kunlun II | Baidu-K2 | 192000 | 120 | fp16 | Chip | [www.tomshardware.com] |
Biren Technology | br100 | br100 | 1.02e+06 | 550 | fp16 | Card | [www.hpcwire.com] [www.tomshardware.com] |
Biren Technology | br104 | br104 | 512000 | 300 | fp16 | Card | [www.hpcwire.com] [www.tomshardware.com] |
Bitmain | BM1880 | Bitmain | 1000 | 2.5 | int8 | Chip | [www.linleygroup.com] |
Blaize | El Cano | Blaize | 1600 | 7 | int8 | Card | [www.blaize.com] |
Cambricon | MLU100 | Cambricon | 64000 | 80 | fp16 | Card | [www.chinamoneynetwork.com] [www.anandtech.com] |
Cambricon | MLU100 | Cambricon | 128000 | 80 | int8 | Card | [www.chinamoneynetwork.com] [www.anandtech.com] |
Canaan | Kendrite K210 | Kendryte | 230 | 0.3 | int8 | Chip | [www.linleygroup.com] |
Cerebras | CS-1 | CS-1 | 1.64e+06 | 15000 | fp16 | System | [www.cerebras.net] |
Cerebras | CS-2 | CS-2 | 3.4e+06 | 23000 | fp16 | System | [www.hpcwire.com] |
Cornami | Cornami | Cornami | 418792 | 30 | fp16 | Chip | [cornami.com] |
Enflame | Cloudblazer T10 | Enflame | 8e-08 | 225 | bf16 | Card | [www.eenewsanalog.com] |
Esperanto | ET-SoC-1 | ES-1 | 500000 | 120 | int8 | Card | [www.computer.org] [fuse.wikichip.org] |
Flex Logix | InferX X1 | FlexLogix | 7640 | 13.5 | int8 | Chip | [flex-logix.com] |
TPU Edge | TPUedge | 4000 | 2 | int8 | System | [aiyprojects.withgoogle.com] | |
TPU1 | TPU1 | 92000 | 75 | int8 | Chip | [doi.org] [www.nextplatform.com] | |
TPU2 | TPU2 | 45000 | 280 | bf16.32 | Chip | [doi.org] [www.nextplatform.com] | |
TPU3 | TPU3 | 123000 | 450 | bf16.32 | Chip | [doi.org] [www.nextplatform.com] | |
TPU4i | TPU4i | 138000 | 175 | bf16.32 | Chip | [doi.org] | |
TPU4 | TPU4 | 275000 | 192 | bf16.32 | Chip | [www.hpcwire.com] | |
GraphCore | C2 | GraphCore | 125000 | 300 | fp16.32 | Card | [www.linleygroup.com] [www.graphcore.ai] |
GraphCore | C2 | GraphCoreNode | 1.6e+06 | 2400 | fp16.32 | System | [www.graphcore.ai] |
GraphCore | Colossus Mk2 | GraphCore2 | 250000 | 300 | bf16 | Card | [www.eetimes.com] |
GraphCore | Bow-2000 | GraphCoreBow | 350000 | 300 | bf16 | Card | [www.tomshardware.com] |
GreenWaves | GAP8 | GAP8 | 22.7 | 0.1 | int8 | Chip | [greenwaves-technologies.com] [www.eejournal.com] |
GreenWaves | GAP9 | GAP9 | 151 | 0.64 | int8 | Chip | [greenwaves-technologies.com] [www.eejournal.com] |
Groq | Groq Node | GroqNode | 1.5e+06 | 3300 | fp16.32 | System | [www.nextplatform.com] |
Groq | Groq Node | GroqNode | 6e+06 | 3300 | int8 | System | [www.nextplatform.com] |
Groq | Tensor Streaming Processor | Groq | 205000 | 300 | fp16 | Card | [groq.com] [doi.org] |
Groq | Tensor Streaming Processor | Groq | 820000 | 300 | int8 | Card | [groq.com] [doi.org] |
Gyrfalcon | Gyrfalcon | Gyrfalcon | 2800 | 0.224 | int8 | Chip | [www.eetimes.com] |
Gyrfalcon | Gyrfalcon | GyrfalconServer | 2.15e+06 | 900 | int8 | System | [www.hpcwire.com] |
Habana | Gaudi | Gaudi | 50000 | 200 | fp16 | Card | [habana.ai] [doi.org] |
Habana | Goya HL-1000 | Goya | 100000 | 150 | int8 | Card | [habana.ai] [doi.org] [www.linleygroup.com] |
Habana | Gaudi2 | Gaudi2 | NaN | 650 | Card | [www.nextplatform.com] | |
Habana | Greco | Greco | NaN | 75 | Card | [www.nextplatform.com] | |
Hailo | Hailo-8 | Hailo-8 | 2688 | 9.3 | int8 | Chip | [www.eetimes.com] |
Hailo | Hailo-15H | Hailo-15 | 20000 | 2 | int8 | Chip | [www.eetimes.com] |
Horizon Robotics | Journey2 | Journey2 | 4000 | 2 | int8 | Chip | [en.horizon.ai] |
Huawei HiSilicon | Ascend 310 | Ascend-310 | 8000 | 8 | fp16 | Chip | [e.huawei.com] |
Huawei HiSilicon | Ascend 310 | Ascend-310 | 16000 | 8 | int8 | Chip | [e.huawei.com] |
Huawei HiSilicon | Ascend 910 | Ascend-910 | 512000 | 310 | fp16 | Chip | [e.huawei.com] |
Huawei HiSilicon | Ascend 910 | Ascend-910 | 512000 | 310 | int8 | Chip | [e.huawei.com] |
IBM | TrueNorth | TrueNorth | 1890 | 0.5 | int8 | System | [www.top500.org] [doi.org] [doi.org] |
IBM | TrueNorth | TrueNorthSys | 1890 | 44 | int8 | System | [www.top500.org] [doi.org] [doi.org] |
IBM/NYU | NeuFlow | NeuFlow | 320 | 10 | int16 | Chip | [doi.org] |
Institute for Computing Technology | DaDianNao | DaDianNao | 5590 | 15.97 | int16 | Chip | [cacm.acm.org] [doi.org] |
Institute for Computing Technology | DianNao | DianNao | 452 | 0.485 | int16 | Chip | [cacm.acm.org] |
Institute for Computing Technology | PuDianNao | PuDianNao | 1060 | 0.596 | int16 | Chip | [cacm.acm.org] [dl.acm.org] |
Institute for Computing Technology | ShiDianNao | ShiDianNao | 194 | 0.32 | int16 | Chip | [cacm.acm.org] [doi.org] |
Intel | Arria 10 1150 | Arria | 283000 | 85 | fp16.32 | Chip | [arxiv.org] [www.nextplatform.com] |
Intel | Mobileye EyeQ5 | EyeQ5 | 12000 | 5 | int8 | Chip | [www.blaize.com] |
Intel | Movidius Myriad X | MovidiusX | 1000 | 2 | int16 | Chip | [www.extremetech.com] |
Intel | Xe-HPC | Intel-Xe-HPC | 360000 | 600 | fp16 | Card | [www.tomshardware.com] [www.hpcwire.com] [www.tomshardware.com] |
Intel | Flex140 | Flex140 | 105000 | 75 | int8 | Card | [www.nextplatform.com] |
Intel | Flex170 | Flex170 | 256000 | 150 | int8 | Card | [www.nextplatform.com] |
Intel | Nervana Lake Crest | Nervana1 | 38000 | 210 | fp32 | Card | [newsroom.intel.com] |
Intel | Nervana Spring Crest | Nervana2 | 120000 | 210 | fp16.32 | Card | [newsroom.intel.com] |
Intel | Xeon Platinum 8180 | 2xXeon8180 | 4480 | 205 | fp32 | Chip | [www.anandtech.com] [www.cpu-world.com] |
Intel | Xeon Platinum 8180 | 2xXeon8280 | 38700 | 205 | int8 | Chip | [www.anandtech.com] [www.cpu-world.com] |
Kalray | Coolidge | Kalray | 3000 | 20 | fp16.32 | Chip | [www.european-processor-initiative.eu] [www.eenewsanalog.com] |
Kalray | Coolidge | Kalray | 1000 | 20 | fp32 | Chip | [www.european-processor-initiative.eu] [www.eenewsanalog.com] |
Kalray | Coolidge | Kalray | 24000 | 20 | int8.32 | Chip | [www.european-processor-initiative.eu] [www.eenewsanalog.com] |
Kneron | KL520 Neural Processing Unit | KL520 | 300 | 0.5 | int8 | Chip | [www.eetasia.com] |
Kneron | KL720 | KL720 | 1400 | 1.556 | int8 | Chip | [www.eetimes.com] |
Maxim | Max 78000 | Maxim | 56 | 0.028 | int8 | Chip | [www.eetimes.com] [www.linleygroup.com] [doi.org] |
MemryX | MX3 | MX3 | 5000 | 1 | fp16 | Chip | [www.eetimes.com] |
Microsoft | Brainwave | Brainwave | 2000 | 150 | int8 | Chip | [www.nextplatform.com] |
MIT | Eyeriss | Eyeriss | 67.2 | 0.278 | int16 | Chip | [doi.org] [doi.org] [doi.org] |
MIT | Netcast | Netcast | 10000 | 0.001 | int8 | Chip | [arxiv.org] |
Mythic | M1076 | Mythic76 | 25000 | 3 | analog | Chip | [www.eetimes.com] [www.nextplatform.com] [medium.com] |
Mythic | M1108 | Mythic108 | 35000 | 4 | analog | Chip | [www.eetimes.com] [www.nextplatform.com] [medium.com] |
NovuMind | NovuTensor | NovuMind | 15000 | 15 | int8 | Chip | [moorinsightsstrategy.com] [www.eetimes.com] |
NVIDIA | Ampere A10 | A10 | 125000 | 150 | fp16.32 | Card | [www.nextplatform.com] |
NVIDIA | Ampere A100 | A100 | 312000 | 400 | fp16.32 | Card | [devblogs.nvidia.com] |
NVIDIA | Ampere A800 | A800 | 218000 | 300 | fp16.32 | Card | [www.tomshardware.com] |
NVIDIA | Ampere A30 | A30 | 165000 | 165 | fp16.32 | Card | [www.nextplatform.com] |
NVIDIA | Ampere A40 | A40 | 150000 | 300 | fp16.32 | Card | [www.nextplatform.com] |
NVIDIA | DGX Station | DGX-Station | 480000 | 1500 | fp16.32 | System | [www.tomshardware.com] |
NVIDIA | DGX-1 | DGX-1 | 900000 | 3500 | fp16.32 | System | [www.tomshardware.com] [www.anandtech.com] |
NVIDIA | DGX-2 | DGX-2 | 1.92e+06 | 10000 | fp16.32 | System | [www.anandtech.com] |
NVIDIA | DGX-A100 | DGX-A100 | 5e+06 | 6500 | fp16.32 | System | [devblogs.nvidia.com] |
NVIDIA | DGX-H100 | DGX-H100 | 1.6e+07 | 10200 | fp16.32 | System | [wccftech.com] |
NVIDIA | H100 | H100 | 1e+06 | 700 | fp16.32 | Card | [www.anandtech.com] |
NVIDIA | Jetson AGX Xavier | XavierAGX | 11000 | 30 | fp16 | System | [www.anandtech.com] |
NVIDIA | Jetson AGX Xavier | XavierAGX | 32000 | 30 | int8 | System | [www.anandtech.com] |
NVIDIA | Jetson NX Orin | OrinNX | 50000 | 25 | int8 | System | [www.hothardware.com] [www.nvidia.com] |
NVIDIA | Jetson AGX Orin | OrinAGX | 138000 | 60 | int8 | System | [www.hothardware.com] [www.nvidia.com] |
NVIDIA | Jetson TX1 | Jetson1 | 408 | 11.7 | fp16.32 | System | [devblogs.nvidia.com] |
NVIDIA | Jetson TX2 | Jetson2 | 580 | 12.8 | fp16.32 | System | [devblogs.nvidia.com] |
NVIDIA | Jetson Xavier NX | XavierNX | 6000 | 15 | fp16 | System | [www.anandtech.com] |
NVIDIA | Jetson Xavier NX | XavierNX | 21000 | 15 | int8 | System | [www.anandtech.com] |
NVIDIA | DRIVE AGX L2 | AGX-L2 | 200000 | 45 | int8 | System | [hothardware.com] |
NVIDIA | DRIVE AGX L5 | AGX-L5 | 2e+06 | 800 | int8 | System | [hothardware.com] |
NVIDIA | L40 | L40 | 90500 | 300 | fp16 | Card | [www.techpowerup.com] |
NVIDIA | Pascal P100 | P100 | 21200 | 300 | fp16.32 | Card | [www.nvidia.com] [www.anandtech.com] |
NVIDIA | T4 | T4 | 131000 | 150 | int8 | Card | [devblogs.nvidia.com] |
NVIDIA | Volta V100 | V100 | 125000 | 300 | fp16.32 | Card | [www.nvidia.com] [www.anandtech.com] |
Perceive | Ergo | Perceive | 4000 | 0.0727 | int8 | Chip | [www.forbes.com] |
PEZY Computing | PEZY-SC2 | PEZY-SC2 | 8190 | 546.13 | fp32 | System | [fuse.wikichip.org] |
Preferred Networks | MN-3 | PN-3 | 524000 | 500 | fp16 | Card | [projects.preferred.jp] [www.anandtech.com] |
Quadric | q1-64 | Quadric | 6240 | 12 | int8 | Chip | [quadric.io] |
Qualcomm | Cloud AI 100 | Qcomm | 400000 | 75 | int8 | Card | [www.eetimes.com] [www.eetimes.com] |
Qualcomm | QRB5165 | RB5 | 15000 | 30 | int8 | System | [www.therobotreport.com] |
Qualcomm | QRB5165N | RB6 | 200000 | 30 | int8 | System | [www.qualcomm.com] |
Rockchip | RK3399Pro | RK3399Pro | 2400 | 3 | int8 | Chip | [www.rock-chips.com] |
SiMa.ai | SiMa.ai | SiMa.ai | 9120 | 4 | int8 | Chip | [www.linleygroup.com] |
Stanford | EIE | EIE | 102 | 0.6 | int16 | Chip | [ieeexplore.ieee.org] |
Stanford | TETRIS | Tetris | 128 | 8.42 | int16 | Chip | [dl.acm.org] |
Syntiant | NDP101 | Syntiant | 200 | 0.01 | int4.8 | Chip | [www.eetimes.com] [www.eetimes.com] |
Tachyum | Prodigy | Tachyum | 1.2e+07 | 950 | fp16.32 | Chip | [www.tomshardware.com] |
Tenstorrent | Tenstorrent | Tenstorrent | 368000 | 75 | int8 | Card | [www.tenstorrent.com] |
Tesla | Tesla Full Self-Driving Computer | Tesla | 72000 | 72 | int8 | System | [doi.org] [en.wikichip.org] |
Texas Instruments | TDA4VM | TexInst | 8000 | 20 | int8 | Chip | [www.eetimes.com] [www.ti.com] [www.linleygroup.com] |
Toshiba | 2015 | Toshiba | 20000 | 10 | int8 | System | [www.eetimes.com] |
Tsinghua | Tianjic | Tianjic | 1210 | 0.95 | int8 | Chip | [www.nature.com] |
Untether | TsunAImi | TsunAImi | 2e+06 | 400 | int8 | Card | [www.linleygroup.com] |
XMOS | xcore.ai | xcore.ai | 51.2 | 1 | int8 | Chip | [www.eetimes.com] |
Copyright 2023 MIT, Albert I. Reuther