A comparison of lookahead and algorithmic blocking techniques for parallel matrix factorization
In this paper, we analyse and compare the techniques of algorithmic blocking and (storage blocking with) lookahead for distributed memory LU, LLT and QR factorizations. Concepts and some useful properties of a simplified model of lookahead are explored, including the minimal degree of lookahead required for optimal performance. Issues in the implementation of lookahead are discussed, which are more involved for the cases of LLT and QR factorizations. It is also explained how hybrid algorithmic...[Show more]
|Collections||ANU Research Publications|
|TR-CS-98-07.pdf||312.93 kB||Adobe PDF|
Items in Open Research are protected by copyright, with all rights reserved, unless otherwise indicated.