DMCA
SPMD Parallelization of Arbitrary Doacross Loops
BibTeX
@MISC{Melin_spmdparallelization,
author = {Emmanuel Melin and Emmanuel Melin and Bruno Raffin and Bruno Raffin and Xavier Rebeuf and Xavier Rebeuf and Bernard Virot and Bernard Virot},
title = {SPMD Parallelization of Arbitrary Doacross Loops},
year = {}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
To reach high performances on massively parallel machines, optimizing compilers have to treat very efficiently all loop-level parallelism. Irregular control and irregular data accesses commonly appear in realistic scientific applications. Moreover many programs involve loops with cross-iteration data dependences, called DOACROSS loops. In this paper we propose a general technique to parallelize such crucial programs. We show that it is possible to perform a data-driven parallelization of arbitrary DOACROSS loops using non-matching send/receive and asymmetrical point to point synchronisations based on structural clocks. Keywords: Automatic parallelization; Doacross Execution; Parallel Programming Languages; Asynchronous Data-Parallel Execution; Structural Clock; Irregular Algorithm. Citation : This Research Report is the preprint version of an article currently submitted to the Europar'97 international conference. R'esum'e Pour atteindre de hautes performances sur les machines mass...
Keyphrases
spmd parallelization arbitrary doacross loop structural clock research report asynchronous data-parallel execution irregular control pour atteindre loop-level parallelism irregular algorithm doacross loop parallel programming language automatic parallelization doacross execution parallel machine high performance hautes performance preprint version cross-iteration data dependence asymmetrical point machine mass irregular data access data-driven parallelization non-matching send receive arbitrary doacross general technique many program international conference crucial program realistic scientific application
