中国科学院机构知识库网格
Chinese Academy of Sciences Institutional Repositories Grid
首页 机构 成果 学者
  
  1. CAS IR Grid
  2. 计算技术研究所
  3. 中国科学院计算技术研究所
  4. 中国科学院计算技术研究所期刊论文
  5. 英文
Statistical Modeling of Soft Error Influence on Neural Networks

文献类型：期刊论文

作者Huang, Haitong3,4; Xue, Xinghua3,4; Liu, Cheng3,4; Wang, Ying3,4; Luo, Tao2; Cheng, Long1; Li, Huawei4; Li, Xiaowei3,4
刊名IEEE TRANSACTIONS ON COMPUTER-AIDED DESIGN OF INTEGRATED CIRCUITS AND SYSTEMS
出版日期2023-11-01
卷号42期号:11页码:4152-4163
关键词Fault analysis fault simulation neural network (NN) reliability statistical fault modeling
ISSN号0278-0070
DOI10.1109/TCAD.2023.3266405
英文摘要Soft errors in large VLSI circuits have a significant impact on computing- and memory-intensive neural network (NN) processing. Understanding the influence of soft errors on NNs is critical to protect against soft errors for reliable NN processing. Prior work mainly relies on fault simulation to analyze the influence of soft errors on NN processing. They are accurate but usually specific to limited configurations of errors and NN models due to the prohibitively slow simulation speed especially for large NN models and datasets. With the observation that the influence of soft errors propagates across a large number of neurons and accumulates as well, we propose to characterize the soft error-induced data disturbance on each neuron with a normal distribution model using the central limit theorem and develop a series of statistical models to analyze the behavior of NN models under soft errors in general. The statistical models reveal not only the correlation between soft errors and the accuracy of NN models but also how NN parameters, such as quantization and architecture affect the reliability of NNs. The proposed models are compared with fault simulations and verified comprehensively. In addition, we observe that the statistical models that characterize the soft error influence can also be utilized to predict fault simulation results in many cases and we explore the use of the proposed statistical models to accelerate fault simulations of NNs. Our experiments show that the proposed accelerated fault simulation provides almost two orders of magnitude speedup with negligible loss of simulation accuracy compared to the baseline fault simulations.
资助项目National Natural Science Foundation of China[62174162] ; National Key Research and Development Program of China[2022YFB4500405] ; Singapore Government's Research, Innovation and Enterprise 2020 Plan (Advanced Manufacturing and Engineering Domain)[A1687b0033]
WOS研究方向Computer Science ; Engineering
语种英语
WOS记录号WOS:001098114300051
出版者IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
源URL[http://119.78.100.204/handle/2XEOYT63/38085]  
专题中国科学院计算技术研究所期刊论文_英文
通讯作者Liu, Cheng
作者单位1.North China Elect Power Univ, Sch Control & Comp Engn, Beijing 102206, Peoples R China
2.ASTAR, Inst High Performance Comp, Singapore 138632, Singapore
3.Univ Chinese Acad Sci, Dept Comp Sci, Beijing 100190, Peoples R China
4.Chinese Acad Sci, Inst Comp Technol, State Key Lab Processors, Beijing 100190, Peoples R China
推荐引用方式
GB/T 7714		 Huang, Haitong,Xue, Xinghua,Liu, Cheng,et al. Statistical Modeling of Soft Error Influence on Neural Networks[J]. IEEE TRANSACTIONS ON COMPUTER-AIDED DESIGN OF INTEGRATED CIRCUITS AND SYSTEMS,2023,42(11):4152-4163.
APA Huang, Haitong.,Xue, Xinghua.,Liu, Cheng.,Wang, Ying.,Luo, Tao.,...&Li, Xiaowei.(2023).Statistical Modeling of Soft Error Influence on Neural Networks.IEEE TRANSACTIONS ON COMPUTER-AIDED DESIGN OF INTEGRATED CIRCUITS AND SYSTEMS,42(11),4152-4163.
MLA Huang, Haitong,et al."Statistical Modeling of Soft Error Influence on Neural Networks".IEEE TRANSACTIONS ON COMPUTER-AIDED DESIGN OF INTEGRATED CIRCUITS AND SYSTEMS 42.11(2023):4152-4163.

入库方式： OAI收割

来源：计算技术研究所

浏览0
下载0
收藏0
其他版本

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。