Paper Info
Title
Evaluating the Performance of Basic Linear Algebra Subroutines on a Torus Array Processor
Abstract
The basic linear algebra subroutines (BLAS) are stan- dard operations to efficiently solve the linear algebra prob- lems on high performance and parallel systems. In this pa- per, we study the implementation of some important BLAS operations on a N脳N torus array processor. We show that the performance of the Level-3 BLAS represented by the n脳n matrix multiply-add operation, nN, approaches the theoretical peak as n increases since the degree of data reusing is high. While the performance of Level-1 and Level-2 BLAS operations is low as a result of low data reusing. Fortunately, many applications are based on inten- sive use of Level-3 BLAS with small percentage of Level-1 and Level-2 BLAS.
Year
DOI
Venue
2007
10.1109/CIT.2007.100
CIT
Keywords
Field
DocType
level-2 blas operation,level-3 blas,data reusing,n torus array processor,torus array processor,basic linear algebra subroutine,n increase,basic linear algebra subroutines,high performance,level-2 blas,n matrix multiply-add operation,important blas operation,linear algebra,parallel processing,bias,parallel systems
Linear algebra,Computer science,Matrix (mathematics),Parallel processing,Parallel computing,Torus,Basic linear algebra subroutines,Vector processor,Basic Linear Algebra Subprograms
Conference
ISBN
Citations 
PageRank 
0-7695-2983-6
0
0.34
References 
Authors
7
2
Name
Order
Citations
PageRank
Ahmed S. Zekri1378.91
Stanislav G. Sedukhin210114.84