Fast solution of large N × N matrix equations in an MIMD-SIMD hybrid system
Parallel Computing - Special issue: Parallel and distributed scientific and engineering computing
A class of Lanczos-like algorithms implemented on parallel computers
Parallel Computing
Hi-index | 0.00 |
Many important scientific and engineering problems require the computation of a small number of eigenvalues of large nonsymmetric matrices. The biorthogonal Lanczos method is one of the methods to solve that problem. In this paper, we introduce the s-step biorthogonal Lanczos method generating reduction matrices which are similar to reduction matrices generated by the standard biorthogonal Lanczos method. The s-step generalization of biorthogonal Lanczos method enhances parallel properties by forming s simultaneous search direction vectors. The s-step biorthogonal Lanczos method has the minimized synchronization points, which resulted in the minimized global communication compared to the standard biorthogonal Lanczos method.