{"title":"Concentration and Its Impact on Mesh and Torus-Based NoC Performance","authors":"S. Loucif","doi":"10.1109/PDP.2015.35","DOIUrl":null,"url":null,"abstract":"This paper investigates the effects of concentration on the performance of k-ary n-cubes. Simulation results indicate that only large ratios of packet length-to-average hop-count are in favor of concentrated mesh and torus. The Cmesh takes full advantage of its high channel bandwidth to outperform Ctorus. Moreover, non-local traffic suffers more from performance bottleneck than local traffic at routers. Providing dedicated input ports, one for each IP, at routers, reduces the average packet latency compared to a configuration with a single input port shared by all IP cores of the cluster.","PeriodicalId":285111,"journal":{"name":"2015 23rd Euromicro International Conference on Parallel, Distributed, and Network-Based Processing","volume":"55 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2015-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2015 23rd Euromicro International Conference on Parallel, Distributed, and Network-Based Processing","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/PDP.2015.35","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 2
Abstract
This paper investigates the effects of concentration on the performance of k-ary n-cubes. Simulation results indicate that only large ratios of packet length-to-average hop-count are in favor of concentrated mesh and torus. The Cmesh takes full advantage of its high channel bandwidth to outperform Ctorus. Moreover, non-local traffic suffers more from performance bottleneck than local traffic at routers. Providing dedicated input ports, one for each IP, at routers, reduces the average packet latency compared to a configuration with a single input port shared by all IP cores of the cluster.