Parallel implementation of eigenvalue algorithms on distributed memory machines

137. B. B. Zhou and R. P. Brent, Parallel implementation of eigenvalue algorithms on distributed memory machines, Australian Computer Science Communications 15 (1993), 19-25.

Abstract: dvi (2K), pdf (73K), ps (25K).

Paper: pdf (1182K).

Abstract

This paper considers the parallel solution of large eigenvalue problem on a mesh-connected processor array with distributed memories.  New parallel Jacobi algorithms are introduced for solving the problem.  The algorithms require a small number of data communications between processing elements on our computing model.  They have been implemented on the Fujitsu AP 1000.  The paper also reports our analytical and experimental results.

Comments

For background material see [84].

Go to next publication

Return to Richard Brent's index page