A 7-nm Four-Core Mixed-Precision AI Chip With 26.2-TFLOPS Hybrid-FP8 Training, 104.9-TOPS INT4 Inference, and Workload-Aware Throttling-Reference-Cited by-同舟云学术

A 7-nm Four-Core Mixed-Precision AI Chip With 26.2-TFLOPS Hybrid-FP8 Training, 104.9-TOPS INT4 Inference, and Workload-Aware Throttling

Published:2022-01 Issue:1 Volume:57 Page:182-197
ISSN:0018-9200
Container-title:IEEE Journal of Solid-State Circuits
language:
Short-container-title:IEEE J. Solid-State Circuits

Author:

Lee Sae Kyu^ORCID,Agrawal Ankur^ORCID,Silberman Joel,Ziegler Matthew^ORCID,Kang Mingu,Venkataramani Swagath^ORCID,Cao Nianzheng^ORCID,Fleischer Bruce^ORCID,Guillorn Michael,Cohen Matthew,Mueller Silvia M.,Oh Jinwook,Lutz Martin,Jung Jinwook^ORCID,Koswatta Siyu,Zhou Ching,Zalani Vidhi,Kar Monodeep,Bonanno James^ORCID,Casatuta Robert,Chen Chia-Yu^ORCID,Choi Jungwook,Haynie Howard,Herbert Alyssa,Jain Radhika^ORCID,Kim Kyu-Hyoun^ORCID,Li Yulong,Ren Zhibin,Rider Scot,Schaal Marcel,Schelm Kerstin,Scheuermann Michael R.,Sun Xiao,Tran Hung,Wang Naigang^ORCID,Wang Wei,Zhang Xin^ORCID,Shah Vinay,Curran Brian,Srinivasan Vijayalakshmi,Lu Pong-Fei,Shukla Sunil,Gopalakrishnan Kailash,Chang Leland

Publisher

Institute of Electrical and Electronics Engineers (IEEE)

Subject

Electrical and Electronic Engineering

Link

http://xplorestaging.ieee.org/ielx7/4/9664397/09610618.pdf?arnumber=9610618

Cited by 15 articles. 订阅此论文施引文献订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献

1. Fast and robust analog in-memory deep neural network training;Nature Communications;2024-08-20

2. Monotonicity of Multi-term Floating-Point Adders;IEEE Transactions on Computers;2024-06

3. Novel adaptive quantization methodology for 8-bit floating-point DNN training;Design Automation for Embedded Systems;2024-02-16

4. FreFlex: A High-Performance Processor for Convolution and Attention Computations via Sparsity-Adaptive Dynamic Frequency Boosting;IEEE Journal of Solid-State Circuits;2024

5. An Efficient Training Accelerator for Transformers With Hardware-Algorithm Co-Optimization;IEEE Transactions on Very Large Scale Integration (VLSI) Systems;2023-11