Harnessing Aspect Oriented Programming on GPU: Application to Warp-Level Parallelism (WLP)

Abstract : Stochastic simulations involve multiple replications in order to build confidence intervals for their results, and Designs Of Experiments (DOEs) to explore their parameters set. In this paper, we propose Warp-Level Parallelism (WLP), a GPU-enabled solution to compute Multiple Replications In Parallel (MRIP) on GPUs (Graphics Processing Units). GPUs are intrinsically tuned to process efficiently the same operation on several data, which is not suited to parallelize MRIP or DOEs. Our approach proposes to rely on small thread groups, called warps, to perform independent computations such as replications. This approach has proved to be efficient on three classical simulation models, but originally lacked the transparency users might expect. In this work, we enhance WLP using Aspect Oriented Programming (AOP). Our work describes the way to combine CUDA and AOP, and brings forward the techniques available to exploit AOP in a CUDA-enabled development.
Type de document :
Article dans une revue
International Journal of Computer Aided Engineering and Technology, 2015, 7 (2), pp.158-175
Liste complète des métadonnées

Littérature citée [15 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/hal-01099208
Contributeur : Jonathan Passerat-Palmbach <>
Soumis le : jeudi 1 janvier 2015 - 15:38:24
Dernière modification le : jeudi 11 janvier 2018 - 06:16:31
Document(s) archivé(s) le : jeudi 10 septembre 2015 - 23:57:11

Fichiers

ijcaet2013_frree.pdf
Fichiers produits par l'(les) auteur(s)

Licence


Distributed under a Creative Commons Paternité 4.0 International License

Identifiants

  • HAL Id : hal-01099208, version 1

Citation

Jonathan Passerat-Palmbach, Pierre Schweitzer, Jonathan Caux, Pridi Siregar, Claude Mazel, et al.. Harnessing Aspect Oriented Programming on GPU: Application to Warp-Level Parallelism (WLP). International Journal of Computer Aided Engineering and Technology, 2015, 7 (2), pp.158-175. 〈hal-01099208〉

Partager

Métriques

Consultations de la notice

348

Téléchargements de fichiers

214