Parallel Performance of a 3D Elliptic Solver
Computing Sciences and Computer Engineering
It was recently shown that block-circulant preconditioners applied to a conjugate gradient method used to solve structured sparse linear systems arising from 2D or 3D elliptic problems have good numerical properties and a potential for high parallel efficiency. In this note parallel performance of a circulant block-factorization based preconditioner applied to a 3D model problem is investigated. The aim of the presentation is to report on the experimental data obtained on SUN Enterprise 3000, SGI/Cray Origin 2000, Cray J-9x, Cray T3E computers and on two PC clusters.
Numerical Analysis and Its Applications
(2001). Parallel Performance of a 3D Elliptic Solver. Numerical Analysis and Its Applications, 1988, 535-543.
Available at: https://aquila.usm.edu/fac_pubs/4006