Network Pruning for Bit-Serial Accelerators

doi:10.1109/TCAD.2022.3203955

	Network Pruning for Bit-Serial Accelerators
	Zhao, Xiandong 1; Wang, Ying 2; Liu, Cheng 2; Shi, Cong 3; Tu, Kaijie 1; Zhang, Lei 1
	2023-05-01
发表期刊	IEEE TRANSACTIONS ON COMPUTER-AIDED DESIGN OF INTEGRATED CIRCUITS AND SYSTEMS
ISSN	0278-0070
卷号	42 期号:5 页码:1597-1609
摘要	Bit-serial architectures (BSAs) are becoming increasingly popular in low-power neural network processor (NNP) designs for edge scenarios. However, the performance and energy efficiency of state-of-the-art BSA NNPs heavily depends on both the proportion and distribution of ineffectual weight bits in neural networks (NNs). To boost the performance of typical BSA accelerators, we present Bit-Pruner, a software approach to learn BSA-favored NNs without resorting to hardware modifications. Bit-Pruner not only progressively prunes but also restructures the nonzero bits in weights so that the number of nonzero bits in the model can be reduced and the corresponding computing can be load-balanced to suit the target BSA accelerators. On top of Bit-Pruner, we further propose a Pareto frontier optimization algorithm to adjust the bit-pruning rate across network layers and fulfill diverse NN processing requirements in terms of performance and accuracy for various edge scenarios. However, an aggressive Bit-Pruner can lead to nontrivial accuracy loss, especially for lightweight NNs and complex tasks. To this end, the alternating direction method of multipliers (ADMMs) is adapted to the retraining phase in Bit-Pruner to smooth the abrupt disturbance due to bit-pruning and enhance the resulting model accuracy. According to the experiments, Bit-Pruner increases the bit-sparsity up to 94.4% with negligible accuracy degradation and achieves an optimized tradeoff between NN accuracy and energy efficiency even under very-aggressive performance constraints. When pruned models are deployed onto typical BSA accelerators, the average performance is 2.1 x and 1.6 x higher than the baseline networks without pruning and those with classical weight pruning, respectively.
关键词	AI accelerators neural networks (NNs) NN compression
DOI	10.1109/TCAD.2022.3203955
收录类别	SCI
语种	英语
资助项目	National Natural Science Foundation of China (NSFC)[61874124] ; National Natural Science Foundation of China (NSFC)[61876173] ; National Natural Science Foundation of China (NSFC)[62222411] ; National Natural Science Foundation of China (NSFC)[62174162] ; Zhejiang Lab[2021PC0AC01] ; State Key Laboratory of Processors ; Institute of Computing Technology ; CAS ; Beijing Key Laboratory of Mobile Computing and Pervasive Device ; Institute of Computing Technology ; Chinese Academy of Sciences
WOS研究方向	Computer Science ; Engineering
WOS类目	Computer Science, Hardware & Architecture ; Computer Science, Interdisciplinary Applications ; Engineering, Electrical & Electronic
WOS记录号	WOS:000976102300018
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
引用统计
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/21442
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Wang, Ying
作者单位	1.Chinese Acad Sci, State Key Lab Processors, Beijing Key Lab Mobile Comp & Pervas Device, Inst Comp Technol,Inst Comp Technol,Univ Chinese A, Beijing 100089, Peoples R China 2.Chinese Acad Sci, Inst Comp Technol, State Key Lab Comp Architecture, Beijing 100089, Peoples R China 3.Chongqing Univ, Sch Microelect & Commun Engn, Chongqing 400044, Peoples R China
推荐引用方式 GB/T 7714	Zhao, Xiandong,Wang, Ying,Liu, Cheng,et al. Network Pruning for Bit-Serial Accelerators[J]. IEEE TRANSACTIONS ON COMPUTER-AIDED DESIGN OF INTEGRATED CIRCUITS AND SYSTEMS,2023,42(5):1597-1609.
APA	Zhao, Xiandong,Wang, Ying,Liu, Cheng,Shi, Cong,Tu, Kaijie,&Zhang, Lei.(2023).Network Pruning for Bit-Serial Accelerators.IEEE TRANSACTIONS ON COMPUTER-AIDED DESIGN OF INTEGRATED CIRCUITS AND SYSTEMS,42(5),1597-1609.
MLA	Zhao, Xiandong,et al."Network Pruning for Bit-Serial Accelerators".IEEE TRANSACTIONS ON COMPUTER-AIDED DESIGN OF INTEGRATED CIRCUITS AND SYSTEMS 42.5(2023):1597-1609.