Stochastic Computing Convolutional Neural Network Architecture Reinvented for Highly Efficient Artificial Intelligence Workload on Field-Programmable Gate Array

作     者：Yang Yang Lee Zaini Abdul Halim Mohd Nadhir Ab Wahab Tarik Adnan Almohamad 

作者机构：School of Electrical and Electronic EngineeringUniversiti Sains MalaysiaNibong Tebal 14300 Penang Malaysia School of Computer SciencesUniversiti Sains MalaysiaGelugor11800 PenangMalaysia Electrical-Electronics Engineering DepartmentFaculty of EngineeringKarabuk University78050 KarabukTurkiye 

出 版 物：《Research》 (研究（英文）)

年 卷 期：2024年第2024卷第2期

页      面：55-79页

核心收录：

学科分类：12[管理学] 1201[管理学-管理科学与工程(可授管理学、工学学位)] 081104[工学-模式识别与智能系统] 08[工学] 0835[工学-软件工程] 0811[工学-控制科学与工程] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：supported in part by the Universiti Sains Malaysia under Grant RUI 1001/PELECT/8014152 

主　　题：hardware Highly rendering 

摘      要：Stochastic computing(SC)has a substantial amount of study on application-specific integrated circuit(ASIC)design for artificial intelligence(AI)edge computing,especially the convolutional neural network(CNN)***,SC has little to no optimization on field-programmable gate array(FPGA).Scaling up the ASIC logic without FPGA-oriented designs is inefficient,while aggregating thousands of bitstreams is still challenging in the conventional *** research has reinvented several FPGA-efficient 8-bit SC CNN computing architectures,i.e.,SC multiplexer multiply-accumulate,multiply-accumulate function generator,and binary rectified linear unit,and successfully scaled and implemented a fully parallel CNN model on Kintex7 *** proposed SC hardware only compromises 0.14%accuracy compared to binary computing on the handwriting Modified National Institute of Standards and Technology classification task and achieved at least 99.72%energy saving per image feedforward and 31?more data throughput than modern *** to SC,early decision termination pushed the performance baseline exponentially with minimum accuracy loss,making SC CNN extremely lucrative for AI edge computing but limited to classification *** SC s inherent noise heavily penalizes CNN regression performance,rendering SC unsuitable for regression tasks.