M. Stojcev, E. Milovanovic, S. Marković, I. Milovanovic
{"title":"Synthesis of orthogonal systolic arrays for fault-tolerant matrix multiplication","authors":"M. Stojcev, E. Milovanovic, S. Marković, I. Milovanovic","doi":"10.1109/MIEL.2010.5490472","DOIUrl":null,"url":null,"abstract":"This paper presents a procedure for designing fault-tolerant systolic array with orthogonal interconnects and bidirectional data flow (2DBOSA) for matrix multiplication. The method employs space-time redundancy to achieve fault-tolerance. The obtained array has Ω = n(n+2) processing elements, and total execution time of Ttot = 6n -5. The array can tolerate single transient errors and the majority of multiple error patterns with high probability. Compared to hexagonal array of same dimensions, the number of I/O pins is reduced for approximately 30%.","PeriodicalId":271286,"journal":{"name":"2010 27th International Conference on Microelectronics Proceedings","volume":"11 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2010-05-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"4","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2010 27th International Conference on Microelectronics Proceedings","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/MIEL.2010.5490472","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 4
Abstract
This paper presents a procedure for designing fault-tolerant systolic array with orthogonal interconnects and bidirectional data flow (2DBOSA) for matrix multiplication. The method employs space-time redundancy to achieve fault-tolerance. The obtained array has Ω = n(n+2) processing elements, and total execution time of Ttot = 6n -5. The array can tolerate single transient errors and the majority of multiple error patterns with high probability. Compared to hexagonal array of same dimensions, the number of I/O pins is reduced for approximately 30%.