Automatically assigned DDC number:
Manually assigned DDC number: 00435
Title: Parallel Multigrid with ADI-like Smoothers in Two Dimensions
Subject: Craig C. Douglas,Sachit Malhotra,Martin H. Schultz Parallel Multigrid with ADI-like Smoothers in Two Dimensions
Description: . Alternating direction iterative (ADI) methods do not usually work well on parallel computers due to having to do parallel rather than serial tridiagonal solves in all but one dimension. An ADI-like iteration is developed and analyzed which does not require parallel tridiagonal solves in any direction, has at least as good of a convergence rate as ADI, and has almost no communication when imbedded as a smoother inside of a multigrid solver. Numerical experiments on a network of workstations and a parallel computer are included. 1 Introduction The alternating direction iterative (ADI) method is a smoother where information moves quickly in each of the dimensions of a bounday value problem. For problems on rectangular, regular meshes, the tridiagonal solves in each direction provide robustness in excess of line SOR when used as a smoother in a multigrid solver. On parallel computers, tridiagonals solves across computer memory boundaries slow classical ADI down to the point where ADI is...
Contributor: The Pennsylvania State University CiteSeer Archives
<?xml version="1.0" encoding="UTF-8"?>
<rec ID="SELF" Type="SELF" CiteSeer_Book="SELF" CiteSeer_Volume="SELF" Title="Parallel Multigrid with ADI-like Smoothers in Two Dimensions" />