Title :
The design of scalable software libraries for distributed memory concurrent computers
Author :
Choi, Jaeyoung ; Dongarra, Jack J. ; Walker, David W.
Author_Institution :
Math. Sci. Sect., Oak Ridge Nat. Lab., TN, USA
Abstract :
Describes the design of ScaLAPACK, a scalable software library for performing dense and banded linear algebra computations on distributed memory concurrent computers. The specification of the data distribution has important consequences for interprocessor communication and load balance, and hence is a major factor in determining performance and scalability of the library routines. The block cyclic data distribution is adopted as a simple, yet general purpose, way of decomposing block-partitioned matrices. Distributed memory versions of the Level 3 BLAS provide an easy and convenient way of implementing the ScaLAPACK routines
Keywords :
distributed memory systems; linear algebra; mathematics computing; matrix algebra; parallel programming; resource allocation; subroutines; Level 3 BLAS; ScaLAPACK; banded linear algebra computations; basic linear algebra subroutines; block cyclic data distribution; block-partitioned matrix decomposition; data distribution specification; dense linear algebra computations; distributed memory concurrent computers; interprocessor communication; load balance; performance; scalable software libraries; Concurrent computing; Distributed computing; High performance computing; Laboratories; Large-scale systems; Lifting equipment; Linear algebra; Scalability; Software design; Software libraries;
Conference_Titel :
Parallel Processing Symposium, 1994. Proceedings., Eighth International
Conference_Location :
Cancun
Print_ISBN :
0-8186-5602-6
DOI :
10.1109/IPPS.1994.288214