A 1.625 TOPS/W SOC for Deep CNN Training and Inference in 28nm CMOS

Autor:	Chun-Pin Lin, Chi-Shi Chen, Chao-Yang Yu, Ming-Hang Hsieh, Chuking Kung, Tzi-Dar Chiueh, Yu-Tung Liu, Hsiu-Wen Wang
Rok vydání:	2021
Předmět:	CMOS Computational complexity theory Computer engineering Computer science Computer number format System on a chip External Data Representation Convolutional neural network Efficient energy use Convolution
Zdroj:	ESSCIRC
Popis:	This work presents a FloatSD8-based system on chip (SOC) for the inference as well as the training of a convolutional neural networks (CNNs). A novel number format (FloatSD8) is employed to reduce the computational complexity of the convolution circuit. By co-designing data representation and circuit, we demonstrate that the AISOC can achieve high convolution performance and optimal energy efficiency without sacrificing the quality of training. At its normal operating condition (200MHz), the AISOC prototype is capable of 0.69 TFLOPS peak performance and 1.625 TOPS/W in 28nm CMOS.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_dedup___::53fce8283fce3769d5bca09815728fc1 https://doi.org/10.1109/esscirc53450.2021.9567756 Zobrazit plný text záznamu