Efficient algorithms for direct resolution of large sparse system on clusters of SMP nodes
HÉNON, Pascal
Algorithms and high performance computing for grand challenge applications [SCALAPPLIX]
Algorithms and high performance computing for grand challenge applications [SCALAPPLIX]
RAMET, Pierre
Algorithms and high performance computing for grand challenge applications [SCALAPPLIX]
Laboratoire Bordelais de Recherche en Informatique [LaBRI]
Algorithms and high performance computing for grand challenge applications [SCALAPPLIX]
Laboratoire Bordelais de Recherche en Informatique [LaBRI]
ROMAN, Jean
Algorithms and high performance computing for grand challenge applications [SCALAPPLIX]
Laboratoire Bordelais de Recherche en Informatique [LaBRI]
Algorithms and high performance computing for grand challenge applications [SCALAPPLIX]
Laboratoire Bordelais de Recherche en Informatique [LaBRI]
HÉNON, Pascal
Algorithms and high performance computing for grand challenge applications [SCALAPPLIX]
Algorithms and high performance computing for grand challenge applications [SCALAPPLIX]
RAMET, Pierre
Algorithms and high performance computing for grand challenge applications [SCALAPPLIX]
Laboratoire Bordelais de Recherche en Informatique [LaBRI]
Algorithms and high performance computing for grand challenge applications [SCALAPPLIX]
Laboratoire Bordelais de Recherche en Informatique [LaBRI]
ROMAN, Jean
Algorithms and high performance computing for grand challenge applications [SCALAPPLIX]
Laboratoire Bordelais de Recherche en Informatique [LaBRI]
< Réduire
Algorithms and high performance computing for grand challenge applications [SCALAPPLIX]
Laboratoire Bordelais de Recherche en Informatique [LaBRI]
Langue
en
Communication dans un congrès
Ce document a été publié dans
SIAM Conference on Applied Linear Algebra, 2003, Williamsburg. 2003
Résumé en anglais
In previous works, we have described an efficient static scheduling based on a mixed 1D/2D block distribution with local aggregation for a parallel supernodal version of sparse $LL^T$ factorization. In this paper, we present ...Lire la suite >
In previous works, we have described an efficient static scheduling based on a mixed 1D/2D block distribution with local aggregation for a parallel supernodal version of sparse $LL^T$ factorization. In this paper, we present new algorithms suitable for overcoming architectures based on clusters of SMP nodes and also technics to keep a good memory scalability. This algorithms are implemented in the PaStiX library that achieved high performances (resolution of a $26.10^6$ unknown system on 192 ES45 quadriprocessors with 35 percent of peak performance).< Réduire
Origine
Importé de halUnités de recherche