Evaluation of several variants of explicitly restarted lanczos eigensolvers and their parallel implementations

Authors:
V. Hernandez;J. E. Roman;A. Tomas
Affiliations:
D. Sistemas Informáticos y Computación, Universidad Politécnica de Valencia, Valencia, Spain;D. Sistemas Informáticos y Computación, Universidad Politécnica de Valencia, Valencia, Spain;D. Sistemas Informáticos y Computación, Universidad Politécnica de Valencia, Valencia, Spain
Venue:
VECPAR'06 Proceedings of the 7th international conference on High performance computing for computational science
Year:
2006

Citing 8
Cited 1

Sparse matrix test problems

ACM Transactions on Mathematical Software (TOMS)
Iterative algorithms for Gram-Schmidt orthogonalization

Computing
Implicit application of polynomial filters in a k-step Arnoldi method

SIAM Journal on Matrix Analysis and Applications
A Shifted Block Lanczos Algorithm for Solving Sparse Symmetric Generalized Eigenproblems

SIAM Journal on Matrix Analysis and Applications
The symmetric eigenvalue problem

The symmetric eigenvalue problem
External selective orthogonalization for the Lanczos in distributed memory

Parallel Computing - Linear systems and associated problems
Lanczos Algorithms for Large Symmetric Eigenvalue Computations, Vol. 1

Lanczos Algorithms for Large Symmetric Eigenvalue Computations, Vol. 1
SLEPc: A scalable and flexible toolkit for the solution of eigenvalue problems

ACM Transactions on Mathematical Software (TOMS) - Special issue on the Advanced CompuTational Software (ACTS) Collection

Parallel Arnoldi eigensolvers with enhanced scalability via global communications rearrangement

Parallel Computing

Quantified Score

Hi-index	0.00

Visualization

Abstract

It is well known that the Lanczos process suffers from loss of orthogonality in the case of finite-precision arithmetic. Several approaches have been proposed in order to address this issue, thus enabling the successful computation of approximate eigensolutions. However, these techniques have been studied mainly in the context of long Lanczos runs, but not for restarted Lanczos eigensolvers. Several variants of the explicitly restarted Lanczos algorithm employing different reorthogonalization strategies have been implemented in SLEPc, the Scalable Library for Eigenvalue Computations. The aim of this work is to assess the numerical robustness of the proposed implementations as well as to study the impact of reorthogonalization in parallel efficiency.