Fakhar Anjam, Stephan Wong, L. Carro, G. Nazar, M. B. Rutzig
{"title":"Simultaneous reconfiguration of issue-width and instruction cache for a VLIW processor","authors":"Fakhar Anjam, Stephan Wong, L. Carro, G. Nazar, M. B. Rutzig","doi":"10.1109/SAMOS.2012.6404173","DOIUrl":null,"url":null,"abstract":"This paper presents an analysis on the impact of simultaneous instruction cache (I-cache) and issue-width reconfiguration for a very long instruction word (VLIW) processor. The issue-width of the processor can be adjusted at run-time to be 2-issue, 4-issue, or 8-issue, and the I-cache can be reconfigured in terms of associativity, cache size, and line size.We observe that, compared to reconfiguring only the I-cache for a fixed issue-width core, reconfiguring the issue-width and I-cache together can further reduce the execution time, energy consumption, and/or the energy-delay product (EDP). The results for the MiBench and the PowerStone benchmark suites show that compared to “2-issue + the best I-cache”, “4-issue + the best I-cache” can reduce execution time, energy consumption, and EDP by up to 37%, 11%, and 36%, respectively, for different applications. Similarly, compared to “2-issue + the best I-cache”, “8-issue + the best I-cache” can reduce execution time and EDP by up to 46% and 30%, respectively, for different applications.","PeriodicalId":130275,"journal":{"name":"2012 International Conference on Embedded Computer Systems (SAMOS)","volume":"125 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2012-07-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"8","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2012 International Conference on Embedded Computer Systems (SAMOS)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/SAMOS.2012.6404173","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 8
Abstract
This paper presents an analysis on the impact of simultaneous instruction cache (I-cache) and issue-width reconfiguration for a very long instruction word (VLIW) processor. The issue-width of the processor can be adjusted at run-time to be 2-issue, 4-issue, or 8-issue, and the I-cache can be reconfigured in terms of associativity, cache size, and line size.We observe that, compared to reconfiguring only the I-cache for a fixed issue-width core, reconfiguring the issue-width and I-cache together can further reduce the execution time, energy consumption, and/or the energy-delay product (EDP). The results for the MiBench and the PowerStone benchmark suites show that compared to “2-issue + the best I-cache”, “4-issue + the best I-cache” can reduce execution time, energy consumption, and EDP by up to 37%, 11%, and 36%, respectively, for different applications. Similarly, compared to “2-issue + the best I-cache”, “8-issue + the best I-cache” can reduce execution time and EDP by up to 46% and 30%, respectively, for different applications.