Resource aggregation for task-based Cholesky Factorization on top of heterogeneous machines - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2016

Resource aggregation for task-based Cholesky Factorization on top of heterogeneous machines

Abstract

Hybrid computing platforms are now commonplace, featuring a large number of CPU cores and accelerators. This trend makes balancing computations between these heterogeneous resources performance critical. In this paper we propose aggregating several CPU cores in order to execute larger parallel tasks and thus improve the load balance between CPUs and accelerators. Additionally, we present our approach to exploit internal parallelism within tasks. This is done by combining two runtime systems: one runtime system to handle the task graph and another one to manage the internal parallelism. We demonstrate the relevance of our approach in the context of the dense Cholesky factorization kernel implemented on top of the StarPU task-based runtime system. We present experimental results showing that our solution outperforms state of the art implementations.
Fichier principal
Vignette du fichier
papier (1).pdf (294.17 Ko) Télécharger le fichier
Origin Files produced by the author(s)
Loading...

Dates and versions

hal-01181135 , version 1 (29-07-2015)
hal-01181135 , version 2 (11-02-2016)
hal-01181135 , version 3 (23-08-2016)

Licence

Identifiers

  • HAL Id : hal-01181135 , version 3

Cite

Terry Cojean, Abdou Guermouche, Andra Hugo, Raymond Namyst, Pierre-André Wacrenier. Resource aggregation for task-based Cholesky Factorization on top of heterogeneous machines. HeteroPar'2016 worshop of Euro-Par, Aug 2016, Grenoble, France. ⟨hal-01181135v3⟩
562 View
626 Download

Share

Gmail Mastodon Facebook X LinkedIn More