Suppression of Intrinsic Delay Variation in FPGAs using Multiple Configurations

Author:

Matsumoto Yohei1,Hioki Masakazu1,Kawanami Takashi1,Koike Hanpei1,Tsutsumi Toshiyuki2,Nakagawa Tadashi3,Sekigawa Toshihiro3

Affiliation:

1. National Institute of Advanced Industrial Science and Technology (AIST) and CREST-Japan Science and Technology Agency

2. National Institute of Advanced Industrial Science and Technology (AIST), CREST-Japan Science and Technology Agency, and Meiji University

3. National Institute of Advanced Industrial Science and Technology (AIST)

Abstract

A new method for improving the timing yield of field-programmable gate array (FPGA) devices affected by intrinsic within-die variation is proposed. The timing variation is reduced by selecting an appropriate configuration for each chip from a set of independent configurations, the critical paths of which do not share the same circuit resources on the FPGA. In this article, the actual method used to generate independent multiple configurations by simply repeating the routing phase is shown, along with the results of Monte Carlo simulation with 10,000 samples. One simulation result showed that the standard deviations of maximum critical path delays are reduced by 28% and 49% for 10% and 30% V th variations ( σ/ μ ), respectively, with 10 independent configurations. Therefore, the proposed method is especially effective for larger V th variation and is expected to be useful for suppressing the performance variation of FPGAs due to the future increase of parameter variation. Another simulation result showed that the effectiveness of the proposed technique was saturated at the use of 10 or more configurations because of the degradation of the quality of the configurations. Therefore, the use of 10 or fewer configurations is reasonable.

Publisher

Association for Computing Machinery (ACM)

Subject

General Computer Science

Cited by 4 articles. 订阅此论文施引文献 订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3