Mohammed A. Alhartomi;Mohd Tasleem Khan;Saeed Alzahrani;Ahmed Alzahmi;Rafi Ahamed Shaik;Jinti Hazarika;Ruwaybih Alsulami;Abdulaziz Alotaibi;Meshal Al-Harthi
{"title":"短时长内存网络的低面积、低功耗 VLSI 架构","authors":"Mohammed A. Alhartomi;Mohd Tasleem Khan;Saeed Alzahrani;Ahmed Alzahmi;Rafi Ahamed Shaik;Jinti Hazarika;Ruwaybih Alsulami;Abdulaziz Alotaibi;Meshal Al-Harthi","doi":"10.1109/JETCAS.2023.3330428","DOIUrl":null,"url":null,"abstract":"Long short-term memory (LSTM) networks are extensively used in various sequential learning tasks, including speech recognition. Their significance in real-world applications has prompted the demand for cost-effective and power-efficient designs. This paper introduces LSTM architectures based on distributed arithmetic (DA), utilizing circulant and block-circulant matrix-vector multiplications (MVMs) for network compression. The quantized weights-oriented approach for training circulant and block-circulant matrices is considered. By formulating fixed-point circulant/block-circulant MVMs, we explore the impact of kernel size on accuracy. Our DA-based approach employs shared full and partial methods of add-store/store-add followed by a select unit to realize an MVM. It is then coupled with a multi-partial strategy to reduce complexity for larger kernel sizes. Further complexity reduction is achieved by optimizing decoders of multiple select units. Pipelining in add-store enhances speed at the expense of a few pipelined registers. The results of the field-programmable gate array showcase the superiority of our proposed architectures based on the partial store-add method, delivering reductions of 98.71% in DSP slices, 33.59% in slice look-up tables, 13.43% in flip-flops, and 29.76% in power compared to the state-of-the-art.","PeriodicalId":48827,"journal":{"name":"IEEE Journal on Emerging and Selected Topics in Circuits and Systems","volume":null,"pages":null},"PeriodicalIF":3.7000,"publicationDate":"2023-11-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Low-Area and Low-Power VLSI Architectures for Long Short-Term Memory Networks\",\"authors\":\"Mohammed A. Alhartomi;Mohd Tasleem Khan;Saeed Alzahrani;Ahmed Alzahmi;Rafi Ahamed Shaik;Jinti Hazarika;Ruwaybih Alsulami;Abdulaziz Alotaibi;Meshal Al-Harthi\",\"doi\":\"10.1109/JETCAS.2023.3330428\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Long short-term memory (LSTM) networks are extensively used in various sequential learning tasks, including speech recognition. Their significance in real-world applications has prompted the demand for cost-effective and power-efficient designs. This paper introduces LSTM architectures based on distributed arithmetic (DA), utilizing circulant and block-circulant matrix-vector multiplications (MVMs) for network compression. The quantized weights-oriented approach for training circulant and block-circulant matrices is considered. By formulating fixed-point circulant/block-circulant MVMs, we explore the impact of kernel size on accuracy. Our DA-based approach employs shared full and partial methods of add-store/store-add followed by a select unit to realize an MVM. It is then coupled with a multi-partial strategy to reduce complexity for larger kernel sizes. Further complexity reduction is achieved by optimizing decoders of multiple select units. Pipelining in add-store enhances speed at the expense of a few pipelined registers. The results of the field-programmable gate array showcase the superiority of our proposed architectures based on the partial store-add method, delivering reductions of 98.71% in DSP slices, 33.59% in slice look-up tables, 13.43% in flip-flops, and 29.76% in power compared to the state-of-the-art.\",\"PeriodicalId\":48827,\"journal\":{\"name\":\"IEEE Journal on Emerging and Selected Topics in Circuits and Systems\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":3.7000,\"publicationDate\":\"2023-11-06\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"IEEE Journal on Emerging and Selected Topics in Circuits and Systems\",\"FirstCategoryId\":\"5\",\"ListUrlMain\":\"https://ieeexplore.ieee.org/document/10309947/\",\"RegionNum\":2,\"RegionCategory\":\"工程技术\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q2\",\"JCRName\":\"ENGINEERING, ELECTRICAL & ELECTRONIC\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"IEEE Journal on Emerging and Selected Topics in Circuits and Systems","FirstCategoryId":"5","ListUrlMain":"https://ieeexplore.ieee.org/document/10309947/","RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"ENGINEERING, ELECTRICAL & ELECTRONIC","Score":null,"Total":0}
Low-Area and Low-Power VLSI Architectures for Long Short-Term Memory Networks
Long short-term memory (LSTM) networks are extensively used in various sequential learning tasks, including speech recognition. Their significance in real-world applications has prompted the demand for cost-effective and power-efficient designs. This paper introduces LSTM architectures based on distributed arithmetic (DA), utilizing circulant and block-circulant matrix-vector multiplications (MVMs) for network compression. The quantized weights-oriented approach for training circulant and block-circulant matrices is considered. By formulating fixed-point circulant/block-circulant MVMs, we explore the impact of kernel size on accuracy. Our DA-based approach employs shared full and partial methods of add-store/store-add followed by a select unit to realize an MVM. It is then coupled with a multi-partial strategy to reduce complexity for larger kernel sizes. Further complexity reduction is achieved by optimizing decoders of multiple select units. Pipelining in add-store enhances speed at the expense of a few pipelined registers. The results of the field-programmable gate array showcase the superiority of our proposed architectures based on the partial store-add method, delivering reductions of 98.71% in DSP slices, 33.59% in slice look-up tables, 13.43% in flip-flops, and 29.76% in power compared to the state-of-the-art.
期刊介绍:
The IEEE Journal on Emerging and Selected Topics in Circuits and Systems is published quarterly and solicits, with particular emphasis on emerging areas, special issues on topics that cover the entire scope of the IEEE Circuits and Systems (CAS) Society, namely the theory, analysis, design, tools, and implementation of circuits and systems, spanning their theoretical foundations, applications, and architectures for signal and information processing.