A distributed and parallel unite and conquer ...
Document type :
Compte-rendu et recension critique d'ouvrage
Title :
A distributed and parallel unite and conquer method to solve sequences of non-Hermitian linear systems
Author(s) :
Wu, Xinzhe [Auteur]
Maison de la Simulation [MDLS]
Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189 [CRIStAL]
Université de Lille
Petiton, Serge [Auteur]
Maison de la Simulation [MDLS]
Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189 [CRIStAL]
Université de Lille
Maison de la Simulation [MDLS]
Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189 [CRIStAL]
Université de Lille
Petiton, Serge [Auteur]
Maison de la Simulation [MDLS]
Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189 [CRIStAL]
Université de Lille
Journal title :
Japan Journal of Industrial and Applied Mathematics
Publisher :
Kinokuniya Company
Publication date :
2019
ISSN :
0916-7005
English keyword(s) :
Eigenvalues
Iterative methods
Unite and Conquer
Krylov methods
Linear solvers
Sequence of linear systems
Iterative methods
Unite and Conquer
Krylov methods
Linear solvers
Sequence of linear systems
HAL domain(s) :
Informatique [cs]/Calcul parallèle, distribué et partagé [cs.DC]
Informatique [cs]/Analyse numérique [cs.NA]
Informatique [cs]/Analyse numérique [cs.NA]
English abstract : [en]
Many problems in science and engineering often require to solve a long sequence of large-scale non-Hermitian linear systems with different Right-hand sides (RHSs) but a unique operator. Efficiently solving such problems ...
Show more >Many problems in science and engineering often require to solve a long sequence of large-scale non-Hermitian linear systems with different Right-hand sides (RHSs) but a unique operator. Efficiently solving such problems on extreme-scale platforms requires the minimization of global communications, reduction of synchronization and promotion of asynchronous communications. Unite and Conquer GMRES/LS-ERAM (UCGLE) method [30] is a suitable candidate with the reduction of global communications and the synchronization points of all computing units. It consists of three computing algorithms with asyn-chronous communications that allow the use of approximate eigenvalues to accelerate the convergence of solving linear systems and to improve the fault tolerance. In this paper, we extend both the mathematical model and the implementation of UCGLE method to adapt to solve sequences of linear systems. The eigenvalues obtained in solving previous linear systems by UCGLE can be recycled, improved on the fly and applied to construct a new initial guess vector for subsequent linear systems, which can achieve a continuous acceleration to solve linear systems in sequence. Numerical experiments using different test matrices to solve sequences of linear systems on supercomputer Tianhe-2 indicate a substantial decrease in both computation time and iteration steps when the approximate eigenvalues are recycled to generate the initial guess vectors.Show less >
Show more >Many problems in science and engineering often require to solve a long sequence of large-scale non-Hermitian linear systems with different Right-hand sides (RHSs) but a unique operator. Efficiently solving such problems on extreme-scale platforms requires the minimization of global communications, reduction of synchronization and promotion of asynchronous communications. Unite and Conquer GMRES/LS-ERAM (UCGLE) method [30] is a suitable candidate with the reduction of global communications and the synchronization points of all computing units. It consists of three computing algorithms with asyn-chronous communications that allow the use of approximate eigenvalues to accelerate the convergence of solving linear systems and to improve the fault tolerance. In this paper, we extend both the mathematical model and the implementation of UCGLE method to adapt to solve sequences of linear systems. The eigenvalues obtained in solving previous linear systems by UCGLE can be recycled, improved on the fly and applied to construct a new initial guess vector for subsequent linear systems, which can achieve a continuous acceleration to solve linear systems in sequence. Numerical experiments using different test matrices to solve sequences of linear systems on supercomputer Tianhe-2 indicate a substantial decrease in both computation time and iteration steps when the approximate eigenvalues are recycled to generate the initial guess vectors.Show less >
Language :
Anglais
Popular science :
Non
ANR Project :
Collections :
Source :
Files
- https://hal.archives-ouvertes.fr/hal-01918738/document
- Open access
- Access the document
- https://hal.archives-ouvertes.fr/hal-01918738/document
- Open access
- Access the document
- https://hal.archives-ouvertes.fr/hal-01918738/document
- Open access
- Access the document
- document
- Open access
- Access the document
- xinzhewu.pdf
- Open access
- Access the document
- xinzhewu.pdf
- Open access
- Access the document