Logo image
On using object oriented parallel programming to build distributed algebraic abstractions
Conference paper   Peer reviewed

On using object oriented parallel programming to build distributed algebraic abstractions

Dennis Gannon, Jenq Kuen Lee and Srinivas Narayana
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Vol.634 LNCS, pp.769-774
1992

Abstract

designing numerical linear algebra computations;parallel object oriented programming language;C++;Concurrent Aggregate;hierarchy of distributed structures;BLAS3 style blocked computation Theoretical Computer Science,Computer Science (all)
This paper considers the problem of designing numerical linear algebra computations with a parallel object oriented programming language. An extension to C++ based on a Concurrent Aggregate concept is used to describe a hierarchy of distributed structures including dense matrices, sparse matrices, banded matrices, vectors and the associated linear algebra. We show that these abstractions are ideal for describing BLAS3 style blocked computation. As an example we show how to build a distributed conjugate gradient algorithms for sparse matrix (based on the NAS sparse benchmark and we show results for four different parallel machines.

Metrics

1 Record Views

Details

Logo image