An Estimation of Complexity and Computational Costs for Vertical Block-Cyclic Distributed Parallel LU Factorization

摘要

The Vertical Block-cyclic Distributed Parallel LU Factorization Method (VBPLU) is effec- tively processed on a distributed memory parallel computer. VBPLU is base on the two techniques, the block algorithm and the aggregation of communications. Since startup time dominates the data com- munication and the aggregation reduces communication issues, the total performance has been much improved. Furthermore this method uses long vectors so that it is also advantageous on vector processors.