{"title":"An improved multiple-path deadlock-free routing algorithm in binary hypercubes","authors":"Qiang Li","doi":"10.1109/IPPS.1992.223000","DOIUrl":null,"url":null,"abstract":"This paper presents a multiple-path deadlock-free routing algorithm in direct binary hypercubes which is an improved version of a previously published algorithm by the author (1991). Between two nodes of distance k, the previous algorithm provides k disjoint paths in one direction and one path in the other. The direction with one path is a performance bottleneck. The new algorithm adds one more disjoint path to the narrow direction using buffer management technique, and preserves the deadlock-free property. Although only one path is added, simulation results presented in this paper show a significant performance improvement since the added path almost doubles the capacity of the bottleneck.<<ETX>>","PeriodicalId":340070,"journal":{"name":"Proceedings Sixth International Parallel Processing Symposium","volume":"35 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"1992-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings Sixth International Parallel Processing Symposium","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/IPPS.1992.223000","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
This paper presents a multiple-path deadlock-free routing algorithm in direct binary hypercubes which is an improved version of a previously published algorithm by the author (1991). Between two nodes of distance k, the previous algorithm provides k disjoint paths in one direction and one path in the other. The direction with one path is a performance bottleneck. The new algorithm adds one more disjoint path to the narrow direction using buffer management technique, and preserves the deadlock-free property. Although only one path is added, simulation results presented in this paper show a significant performance improvement since the added path almost doubles the capacity of the bottleneck.<>