Parallel spectral division using the matrix sign function for the generalized eigenproblem

被引:6
|
作者
Huss-Lederman, S
Quintana-Ortí, ES
Sun, XB
Wu, YJY
机构
[1] Beloit Coll, Dept Math & Comp Sci, Beloit, WI 53511 USA
[2] Univ Jaume 1, Dept Informat, Castellon de La Plana 12080, Spain
[3] Duke Univ, Dept Comp Sci, Durham, NC 27708 USA
[4] Boeing Co, Phantom Works, Math & Engn Anal, Seattle, WA 98124 USA
来源
关键词
generalized nonsymmetric eigenproblem; spectral division; matrix sign function; generalized Newton iteration; parallel distributed memory computers;
D O I
10.1142/S0129053300000084
中图分类号
TP301 [理论、方法];
学科分类号
081202 ;
摘要
In this paper we demonstrate the parallelism of the spectral division using the matrix sign function for the generalized nonsymmetric eigenproblem. We employ the so-called generalized Newton iterative scheme in order to compute the sign function of the matrix pair. A recent study showed a considerable reduction (by 75%) in the computational cost of this iteration, making this approach competitive when compared to the traditional QZ algorithm. The experimental results on an IBM SP3 multicomputer report the parallel performance (efficiency around 60-80%) and scalability of this approach.
引用
收藏
页码:1 / 14
页数:14
相关论文
共 50 条