Applying parallel direct solver techniques to build robust high performance preconditioners

Pascal Hénon; François Pellegrini; Pierre Ramet; Jean Roman; Yousef Saad

doi:10.1007/11558958_73

Applying parallel direct solver techniques to build robust high performance preconditioners

Pascal Hénon, François Pellegrini, Pierre Ramet, Jean Roman, Yousef Saad

Computer Science and Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

The purpose of our work is to provide a method which exploits the parallel blockwise algorithmic approach used in the framework of high performance sparse direct solvers in order to develop robust preconditioned based on a parallel incomplete factorization. The idea is then to define an adaptive blockwise incomplete factorization that is much more accurate (and numerically mure robust) than the scalar incomplete factorizations commonly used to precondition iterative solvers.

Original language	English (US)
Title of host publication	Applied Parallel Computing - State of the Art in Scientific Computing - 7th International Conference, PARA 2004, Revised Selected Papers
Publisher	Springer Verlag
Pages	611-619
Number of pages	9
ISBN (Print)	3540290672, 9783540290674
DOIs	https://doi.org/10.1007/11558958_73
State	Published - 2006
Event	7th International Conference on Applied Parallel Computing, PARA 2004 - Lyngby, Denmark Duration: Jun 20 2004 → Jun 23 2004

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	3732 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Other

Other	7th International Conference on Applied Parallel Computing, PARA 2004
Country/Territory	Denmark
City	Lyngby
Period	6/20/04 → 6/23/04

Access

10.1007/11558958_73

OpenUrl availability

Full text

Cite this

Hénon, P., Pellegrini, F., Ramet, P., Roman, J., & Saad, Y. (2006). Applying parallel direct solver techniques to build robust high performance preconditioners. In Applied Parallel Computing - State of the Art in Scientific Computing - 7th International Conference, PARA 2004, Revised Selected Papers (pp. 611-619). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 3732 LNCS). Springer Verlag. https://doi.org/10.1007/11558958_73

Applying parallel direct solver techniques to build robust high performance preconditioners. / Hénon, Pascal; Pellegrini, François; Ramet, Pierre et al.
Applied Parallel Computing - State of the Art in Scientific Computing - 7th International Conference, PARA 2004, Revised Selected Papers. Springer Verlag, 2006. p. 611-619 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 3732 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Hénon, P, Pellegrini, F, Ramet, P, Roman, J & Saad, Y 2006, Applying parallel direct solver techniques to build robust high performance preconditioners. in Applied Parallel Computing - State of the Art in Scientific Computing - 7th International Conference, PARA 2004, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 3732 LNCS, Springer Verlag, pp. 611-619, 7th International Conference on Applied Parallel Computing, PARA 2004, Lyngby, Denmark, 6/20/04. https://doi.org/10.1007/11558958_73

Hénon P, Pellegrini F, Ramet P, Roman J, Saad Y. Applying parallel direct solver techniques to build robust high performance preconditioners. In Applied Parallel Computing - State of the Art in Scientific Computing - 7th International Conference, PARA 2004, Revised Selected Papers. Springer Verlag. 2006. p. 611-619. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/11558958_73

Hénon, Pascal ; Pellegrini, François ; Ramet, Pierre et al. / Applying parallel direct solver techniques to build robust high performance preconditioners. Applied Parallel Computing - State of the Art in Scientific Computing - 7th International Conference, PARA 2004, Revised Selected Papers. Springer Verlag, 2006. pp. 611-619 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{615f0a8bd0f44166b0cd07140877df41,

title = "Applying parallel direct solver techniques to build robust high performance preconditioners",

abstract = "The purpose of our work is to provide a method which exploits the parallel blockwise algorithmic approach used in the framework of high performance sparse direct solvers in order to develop robust preconditioned based on a parallel incomplete factorization. The idea is then to define an adaptive blockwise incomplete factorization that is much more accurate (and numerically mure robust) than the scalar incomplete factorizations commonly used to precondition iterative solvers.",

author = "Pascal H{\'e}non and Fran{\c c}ois Pellegrini and Pierre Ramet and Jean Roman and Yousef Saad",

year = "2006",

doi = "10.1007/11558958_73",

language = "English (US)",

isbn = "3540290672",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "611--619",

booktitle = "Applied Parallel Computing - State of the Art in Scientific Computing - 7th International Conference, PARA 2004, Revised Selected Papers",

}

TY - GEN

T1 - Applying parallel direct solver techniques to build robust high performance preconditioners

AU - Hénon, Pascal

AU - Pellegrini, François

AU - Ramet, Pierre

AU - Roman, Jean

AU - Saad, Yousef

PY - 2006

Y1 - 2006

N2 - The purpose of our work is to provide a method which exploits the parallel blockwise algorithmic approach used in the framework of high performance sparse direct solvers in order to develop robust preconditioned based on a parallel incomplete factorization. The idea is then to define an adaptive blockwise incomplete factorization that is much more accurate (and numerically mure robust) than the scalar incomplete factorizations commonly used to precondition iterative solvers.

AB - The purpose of our work is to provide a method which exploits the parallel blockwise algorithmic approach used in the framework of high performance sparse direct solvers in order to develop robust preconditioned based on a parallel incomplete factorization. The idea is then to define an adaptive blockwise incomplete factorization that is much more accurate (and numerically mure robust) than the scalar incomplete factorizations commonly used to precondition iterative solvers.

UR - http://www.scopus.com/inward/record.url?scp=33745303755&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33745303755&partnerID=8YFLogxK

U2 - 10.1007/11558958_73

DO - 10.1007/11558958_73

M3 - Conference contribution

AN - SCOPUS:33745303755

SN - 3540290672

SN - 9783540290674

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 611

EP - 619

BT - Applied Parallel Computing - State of the Art in Scientific Computing - 7th International Conference, PARA 2004, Revised Selected Papers

PB - Springer Verlag

T2 - 7th International Conference on Applied Parallel Computing, PARA 2004

Y2 - 20 June 2004 through 23 June 2004

ER -

Applying parallel direct solver techniques to build robust high performance preconditioners

Abstract

Publication series

Other

Access

OpenUrl availability

Other files and links

Fingerprint

Cite this