A framework for exploiting task and data parallelism on distributed memory multicomputers

Ramaswamy, S.; Sapatnekar, Sachin S.; Banerjee, P.

doi:10.1109/71.642945

Cited by 91 publications

(79 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Ce sont les algorithmes d'ordonnancement en une étape (Boudet et al, 2003). La seconde catégorie regroupe les algorithmes d'ordonnancement en deux étapes (Radulescu et al, 2001a;Radulescu et al, 2001b;Ramaswamy et al, 1997;Rauber et al, 1998) où l'allocation et le placement sont séparés en deux procédures distinctes.…”

Section: Exemple D'ordonnancement En Deux éTapes D'un Graphe De Tâcunclassified

“…De nombreuses études ont été réalisées pour l'ordonnancement de graphes de tâches parallèles dans le cas de plates-formes homogènes (Lepère et al, 2002;Radulescu et al, 2001a;Radulescu et al, 2001b;Ramaswamy et al, 1997;Rauber et al, 1998). Or les plates-formes hétérogènes sont de plus en plus répandues et très attrayantes car elles peuvent permettre de déployer des applications parallèles à des échelles sans précédents.…”

Section: Exemple D'ordonnancement En Deux éTapes D'un Graphe De Tâcunclassified

“…Si certaines études théoriques existent (Lepère et al, 2002), la plupart des algorithmes d'ordonnancement de tâches parallèles (Radulescu et al, 2001a;Radulescu et al, 2001b;Ramaswamy et al, 1997;Rauber et al, 1998) sont des algorithmes en deux étapes et ont été conçus pour des milieux homogènes. La première étape vise à déterminer un nombre de processeurs adéquat pour chaque tâche.…”

Section: Travaux Précédentsunclassified

See 2 more Smart Citations

Heuristiques d'ordonnancement en deux étapes de graphes de tâches parallèles

N'Takpé

2009

Techniques et sciences informatiques

View full text Add to dashboard Cite

show abstract

Section: Exemple D'ordonnancement En Deux éTapes D'un Graphe De Tâcunclassified

Section: Travaux Précédentsunclassified

See 1 more Smart Citation

Heuristiques d'ordonnancement en deux étapes de graphes de tâches parallèles

N'Takpé

2009

Techniques et sciences informatiques

View full text Add to dashboard Cite

show abstract

“…In other words, a task is assigned a single core. The techniques presented in [2]- [5] take into account data parallelism within tasks as well as task parallelism. Their methods take a task graph as input and perform task scheduling and mapping simultaneously, aiming at minimization of schedule length or maximization of pipeline throughput.…”

Section: Related Workmentioning

confidence: 99%

Static Mapping of Multiple Data-Parallel Applications on Embedded Many-Core SoCs

Kaida

Hara-Azumi

Hieda

et al. 2013

IEICE Trans. Inf. & Syst.

View full text Add to dashboard Cite

SUMMARYThis paper studies the static mapping of multiple applications on embedded many-core SoCs. The mapping techniques proposed in this paper take into account both inter-application and intra-application parallelism in order to fully utilize the potential parallelism of the manycore architecture. Two approaches are proposed for static mapping: one approach is based on integer linear programming and the other is based on a greedy algorithm. Experiments show the effectiveness of the proposed techniques.

show abstract

“…Compiler-based approaches seek to identify task-parallel structures automatically, within data-parallel speci cations 11,14,21], while language-based approaches provide new language constructs for specifying task parallelism explicitly 3,6,19,24]. Both approaches have shown promise in certain application areas, but each also has disadvantages.…”

Section: Introductionmentioning

confidence: 99%

A Library-Based Approach to Task Parallelism in a Data-Parallel Language

Foster

Kohr

Krishnaiyer

et al. 1997

Journal of Parallel and Distributed Computing

View full text Add to dashboard Cite

The data-parallel language High Performance Fortran (HPF) does not allow e cient expression of mixed task/data-parallel computations or the coupling of separately compiled data-parallel modules. In this paper, we show how these common parallel program structures can be represented, with only minor extensions to the HPF model, by using a coordination library based on the Message Passing Interface (MPI). This library allows data-parallel tasks to exchange distributed data structures using calls to simple communication functions. We present microbenchmark results that characterize the performance of this library and that quantify the impact of optimizations that allow reuse of communication schedules in common situations. In addition, results from two-dimensional FFT, convolution, and multiblock programs demonstrate that the HPF/MPI library can provide performance superior to that of pure HPF. We conclude that this synergistic combination of two parallel programming standards represents a useful approach to task parallelism in a data-parallel framework, increasing the range of problems addressable in HPF without requiring complex compiler technology.

show abstract

A framework for exploiting task and data parallelism on distributed memory multicomputers

Cited by 91 publications

References 23 publications

Heuristiques d'ordonnancement en deux étapes de graphes de tâches parallèles

Heuristiques d'ordonnancement en deux étapes de graphes de tâches parallèles

Static Mapping of Multiple Data-Parallel Applications on Embedded Many-Core SoCs

A Library-Based Approach to Task Parallelism in a Data-Parallel Language

Contact Info

Product

Resources

About