A preliminary nested-parallel framework to efficiently implement scientific applications

Conference paper (2005)

Authors

A Gonzáles-Escribano External organisation
A.J.C. van Gemund Software Engineering - EEMCS
V Cardeñoso-Payo External organisation
R Portales-Fernandez External organisation
JA Caminero-Granja External organisation
Research Group
Software Engineering (EEMCS) (TU Delft)
More Info
expand_more

Published Date

2005

Language

Undefined/Unknown

Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Faculty

Electrical Engineering, Mathematics and Computer Science

Department

Software Technology

Research Group

Software Engineering

Abstract

Nested-parallel programming models, where the task graph associated to a computation is series-parallel, present good analysis properties that can be exploited for scheduling, cost estimation or automatic mapping to different architectures.
In this work we present a preliminary framework approach to exploit some of these advantages. In our framework we reconstruct an application task graph from a high-level specification, where no scheduling or communication details are yet expressed. The obtained synchronization structure determines which mapping modules or back-ends are used to port the application to an specific platform.
The first results obtained with our prototype show that even simple balancing techniques for irregular scientific applications may be easily integrated in this nested-parallel framework, to obtain efficient implementations from high-level and portable specifications.
Topic: Parallel and Distributed Computing.