A Randomized LU-based Solver Using GPU and Intel Xeon Phi Accelerators

Abstract : We present a fast hybrid solver for dense linear systems based on LU factorization. To achieve good performance, we avoid pivoting by using random butterfly transformations for which we developed efficient implementations on heterogeneous architectures. We used both Graphics Processing Units and Intel Xeon Phi as accelerators. The performance results show that the pre-processing due to randomization is negligible and that the solver outperforms the corresponding routines based on partial pivoting.
Type de document :
Communication dans un congrès
HeteroPar'2015, Aug 2015, Vienna, Austria. 2015, 〈http://heteropar2015.ewi.tudelft.nl〉
Liste complète des métadonnées

https://hal.inria.fr/hal-01223018
Contributeur : Marc Baboulin <>
Soumis le : dimanche 1 novembre 2015 - 13:04:46
Dernière modification le : jeudi 5 avril 2018 - 12:30:23

Identifiants

  • HAL Id : hal-01223018, version 1

Citation

Marc Baboulin, Amal Khabou, Adrien Rémy. A Randomized LU-based Solver Using GPU and Intel Xeon Phi Accelerators. HeteroPar'2015, Aug 2015, Vienna, Austria. 2015, 〈http://heteropar2015.ewi.tudelft.nl〉. 〈hal-01223018〉

Partager

Métriques

Consultations de la notice

143