BEAUMONT, Olivier; MARCHAL, Loris

doi:10.1145/2600212.2600223

Métadonnées

Licence d’utilisation du document

BEAUMONT, Olivier
Laboratoire Bordelais de Recherche en Informatique [LaBRI]
Reformulations based algorithms for Combinatorial Optimization [Realopt]

MARCHAL, Loris
Laboratoire de l'Informatique du Parallélisme [LIP]
Optimisation des ressources : modèles, algorithmes et ordonnancement [ROMA]

Langue

Communication dans un congrès

Ce document a été publié dans

ACM Symposium on High-Performance Parallel and Distributed Computing, 2014-06-23, Vancouver. 2014

Résumé en anglais

The tremendous increase in the size and heterogeneity of supercomputers makes it very difficult to predict the perfor-mance of a scheduling algorithm. Therefore, dynamic solu-tions, where scheduling decisions are made at runtime have overpassed static allocation strategies. The simplicity and efficiency of dynamic schedulers such as Hadoop are a key of the success of the MapReduce framework. Dynamic sched-ulers such as StarPU, PaRSEC or StarSs are also developed for more constrained computations, e.g. task graphs coming from linear algebra. To make their decisions, these runtime systems make use of some static information, such as the distance of tasks to the critical path or the affinity between tasks and computing resources (CPU, GPU,. . .) and of dy-namic information, such as where input data are actually located. In this paper, we concentrate on two elementary linear algebra kernels, namely the outer product and the matrix multiplication. For each problem, we propose sev-eral dynamic strategies that can be used at runtime and we provide an analytic study of their theoretical performance. We prove that the theoretical analysis provides very good estimate of the amount of communications induced by a dy-namic strategy and can be used in order to efficiently deter-mine thresholds used in dynamic scheduler, thus enabling to choose among them for a given problem and architecture.< Réduire

Mots clés en anglais

Dynamic scheduling

data-aware algorithms

randomized al-gorithms

performance evaluation

matrix multiplication

URI

https://oskar-bordeaux.fr/handle/20.500.12278/194548

DOI

http://dx.doi.org/10.1145/2600212.2600223

Project ANR

Solveurs pour architectures hétérogènes utilisant des supports d'exécution - ANR-13-MONU-0007
Résilience des applications scientifiques sur machines exascales - ANR-10-BLAN-0301

Origine

Importé de hal

Unités de recherche

Institut de Mathématiques de Bordeaux (IMB) - UMR 5251