{"title":"A Systolic Array Programming Language","authors":"P. Tseng","doi":"10.1109/DMCC.1990.556326","DOIUrl":null,"url":null,"abstract":"This paper describes the AL programming lani page for the Warp systolic array, a linear array of 11 processing cells. AL has been successfully used tc program matrix computation applications on WarIi. Examples of LU decomposition, QR decompositioii , and singular value decomposition (SVD) were giv:n in this paper to illustrate the use of AL. These AI procedures were rewritten from the LINPACK Fortran programs; their algorithms and numerical prop xties were kept the same as the LINPACK implementations. More than 27 MFLOPS (out of 100 MFI ,OPS peak) on matrices of order 300 were achieved for these procedures.","PeriodicalId":204431,"journal":{"name":"Proceedings of the Fifth Distributed Memory Computing Conference, 1990.","volume":"19 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"1990-04-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"4","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of the Fifth Distributed Memory Computing Conference, 1990.","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/DMCC.1990.556326","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 describes the AL programming lani page for the Warp systolic array, a linear array of 11 processing cells. AL has been successfully used tc program matrix computation applications on WarIi. Examples of LU decomposition, QR decompositioii , and singular value decomposition (SVD) were giv:n in this paper to illustrate the use of AL. These AI procedures were rewritten from the LINPACK Fortran programs; their algorithms and numerical prop xties were kept the same as the LINPACK implementations. More than 27 MFLOPS (out of 100 MFI ,OPS peak) on matrices of order 300 were achieved for these procedures.