Block Wiedemann algorithm on multicore architectures
Résumé
Solving a linear system with large sparse matrices is a com-putational kernel used in a wide range of applications. The block version of Wiedemann’s algorithm proposed in [Wiedemann 1986] takes advantage of the sparsity to achieve better performance.
Our objectives in this poster is to provide an efficient implementation of block Wiedemann algorithm on NUMA multicores architectures.
Domaines
Calcul formel [cs.SC]
Fichier principal
Block_Wiedemann_on_Multicore_Architectures_ISSAC2013.pdf (217.94 Ko)
Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...