具有可重构数据流的卷积神经网络加速器

2018 International SoC Design Conference (ISOCC) Pub Date : 2018-11-01 DOI:10.1109/ISOCC.2018.8649988

Myungwoo Oh, Chaeeun Lee, Sanghun Lee, Youngho Seo, Sunwoo Kim, Jooho Wang, C. Park

{"title":"具有可重构数据流的卷积神经网络加速器","authors":"Myungwoo Oh, Chaeeun Lee, Sanghun Lee, Youngho Seo, Sunwoo Kim, Jooho Wang, C. Park","doi":"10.1109/ISOCC.2018.8649988","DOIUrl":null,"url":null,"abstract":"Convolutional-Neural-Network (CNN) is used in broad applications. There are dataflows for convolutional layers in CNN such as row-stationary and weight-stationary. However, these dataflows have strengths and weaknesses. This paper analyzed two representative dataflows and introduce the dataflow-reconfigurable CNN accelerator that takes advantage of both dataflows.","PeriodicalId":127156,"journal":{"name":"2018 International SoC Design Conference (ISOCC)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":"{\"title\":\"Convolutional Neural Network Accelerator with Reconfigurable Dataflow\",\"authors\":\"Myungwoo Oh, Chaeeun Lee, Sanghun Lee, Youngho Seo, Sunwoo Kim, Jooho Wang, C. Park\",\"doi\":\"10.1109/ISOCC.2018.8649988\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Convolutional-Neural-Network (CNN) is used in broad applications. There are dataflows for convolutional layers in CNN such as row-stationary and weight-stationary. However, these dataflows have strengths and weaknesses. This paper analyzed two representative dataflows and introduce the dataflow-reconfigurable CNN accelerator that takes advantage of both dataflows.\",\"PeriodicalId\":127156,\"journal\":{\"name\":\"2018 International SoC Design Conference (ISOCC)\",\"volume\":\"12 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2018-11-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"2\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2018 International SoC Design Conference (ISOCC)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/ISOCC.2018.8649988\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2018 International SoC Design Conference (ISOCC)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ISOCC.2018.8649988","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 2

摘要

卷积神经网络(CNN)有着广泛的应用。卷积层在CNN中有行平稳和权平稳等数据流。然而，这些数据流有优点也有缺点。本文分析了两种具有代表性的数据流，并介绍了利用这两种数据流的数据流可重构CNN加速器。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

查看原文本刊更多论文

Convolutional Neural Network Accelerator with Reconfigurable Dataflow

Convolutional-Neural-Network (CNN) is used in broad applications. There are dataflows for convolutional layers in CNN such as row-stationary and weight-stationary. However, these dataflows have strengths and weaknesses. This paper analyzed two representative dataflows and introduce the dataflow-reconfigurable CNN accelerator that takes advantage of both dataflows.

求助全文

通过发布文献求助，成功后即可免费获取论文全文。去求助

来源期刊

2018 International SoC Design Conference (ISOCC)

自引率

0.00%

发文量