Parallelization of a Recursive Decoupling Method for Solving Tridiagonal Linear Systems on Distributed Memory Computer.
ABSTRACT . This work presents a parallelization of a recursive decouplingmethod for solving tridiagonal linear system on distributed memory computer.We study the fill-in in the algorithm to optimize the execution ofthe scalar algorithm and to perform the communications. Finally, weevaluate the algorithm through specific test on the Fujitsu AP3000.1 IntroductionIn recent years considerable effort has been devoted to solve tridiagonal systems(TS), a very important class of linear systems which...
[Show abstract] [Hide abstract]
ABSTRACT: In this work we apply a methodology for the parallelization of algorithms for tridiagonal solvers. We classify tridiagonal solvers as a function of their data flows and present a unified version of the projection of these algorithms onto computers with mesh topology and distributed memory. Finally, we evaluate the algorithms and compare them through specific tests on the Fujitsu AP1000.International Journal of High Speed Computing 06/1997; 9(2):101-126. DOI:10.1142/S012905339700009X
International Journal of High Speed Computing 09/1994; 6:375-386. DOI:10.1142/S0129053394000196