正在进行的工作:EAST-DNN:使用深度神经网络加速架构模拟

2019 International Conference on Hardware/Software Codesign and System Synthesis (CODES+ISSS) Pub Date : 2019-10-01 DOI:10.1145/3349567.3351728

Arko Dutt, G. Narasimman, Lin Jie, V. Chandrasekhar, M. Sabry

{"title":"正在进行的工作:EAST-DNN:使用深度神经网络加速架构模拟","authors":"Arko Dutt, G. Narasimman, Lin Jie, V. Chandrasekhar, M. Sabry","doi":"10.1145/3349567.3351728","DOIUrl":null,"url":null,"abstract":"A rapid and accurate architectural simulator is a cornerstone for an efficient design-space exploration of computing systems. In this paper, we introduce EAST-DNN, a feed-forward deep neural network, to accelerate architectural simulations. EAST-DNN achieves $> 10^{6}\\times$ speedup with an average prediction error of 4.3% over the baseline simulator. It also achieves an average of $2\\times$ better accuracy with at least $2.3\\times$ speedup compared to state-of-the-art.","PeriodicalId":194982,"journal":{"name":"2019 International Conference on Hardware/Software Codesign and System Synthesis (CODES+ISSS)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":"{\"title\":\"Work-in-Progress: EAST-DNN: Expediting Architectural SimulaTions Using Deep Neural Networks\",\"authors\":\"Arko Dutt, G. Narasimman, Lin Jie, V. Chandrasekhar, M. Sabry\",\"doi\":\"10.1145/3349567.3351728\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"A rapid and accurate architectural simulator is a cornerstone for an efficient design-space exploration of computing systems. In this paper, we introduce EAST-DNN, a feed-forward deep neural network, to accelerate architectural simulations. EAST-DNN achieves $> 10^{6}\\\\times$ speedup with an average prediction error of 4.3% over the baseline simulator. It also achieves an average of $2\\\\times$ better accuracy with at least $2.3\\\\times$ speedup compared to state-of-the-art.\",\"PeriodicalId\":194982,\"journal\":{\"name\":\"2019 International Conference on Hardware/Software Codesign and System Synthesis (CODES+ISSS)\",\"volume\":\"12 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2019-10-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"1\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2019 International Conference on Hardware/Software Codesign and System Synthesis (CODES+ISSS)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1145/3349567.3351728\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2019 International Conference on Hardware/Software Codesign and System Synthesis (CODES+ISSS)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/3349567.3351728","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 1

摘要

快速、准确的建筑模拟器是高效探索计算系统设计空间的基石。在本文中，我们引入了一种前馈深度神经网络EAST-DNN来加速建筑仿真。与基线模拟器相比，EAST-DNN实现了$> 10^{6}\倍$加速，平均预测误差为4.3%。与最先进的技术相比，它的平均精度提高了2倍，加速速度至少提高了2.3倍。

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

Work-in-Progress: EAST-DNN: Expediting Architectural SimulaTions Using Deep Neural Networks

A rapid and accurate architectural simulator is a cornerstone for an efficient design-space exploration of computing systems. In this paper, we introduce EAST-DNN, a feed-forward deep neural network, to accelerate architectural simulations. EAST-DNN achieves $> 10^{6}\times$ speedup with an average prediction error of 4.3% over the baseline simulator. It also achieves an average of $2\times$ better accuracy with at least $2.3\times$ speedup compared to state-of-the-art.

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

2019 International Conference on Hardware/Software Codesign and System Synthesis (CODES+ISSS)

自引率

0.00%

发文量