Compile-time and run-time issues in an auto-parallelisation system for the cell BE processor

Alastair Donaldson, Paul Keir, Anton Lokhmotov

Research output: Chapter in Book/Report/Conference proceedingChapter

5 Citations (Scopus)

Abstract

We describe compiler and run-time optimisations for effective auto-parallelisation of C++ programs on the Cell BE architecture. Auto-parallelisation is made easier by annotating sieve scopes, which abstract the "read in, compute in parallel, write out" processing paradigm. We show that the semantics of sieve scopes enables data movement optimisations, such as re-organising global memory reads to minimise DMA transfers and streaming reads from uniformly accessed arrays. We also describe run-time optimisations for committing side-effects to main memory. We provide experimental results showing the benefits of our optimisations, and compare the Sieve-Cell system with IBM’s OpenMP implementation for Cell.
Original languageEnglish
Title of host publicationEuro-Par 2008
Subtitle of host publicationEuro-Par 2008 Workshops - Parallel Processing
EditorsEduardo César, Michael Alexander, Achim Streit, Jesper Larsson Träff, Christophe Cérin, Andreas Knüpfer, Dieter Kranzlmüller, Shantenu Jha
PublisherSpringer-Verlag
Pages163-173
Number of pages11
Volume5415
ISBN (Print)978-3-642-00955-6
DOIs
Publication statusPublished - 2009
Externally publishedYes

Publication series

NameLecture Notes in Computer Science
PublisherSpringer
ISSN (Print)0302-9743

    Fingerprint

Cite this

Donaldson, A., Keir, P., & Lokhmotov, A. (2009). Compile-time and run-time issues in an auto-parallelisation system for the cell BE processor. In E. César, M. Alexander, A. Streit, J. L. Träff, C. Cérin, A. Knüpfer, D. Kranzlmüller, ... S. Jha (Eds.), Euro-Par 2008: Euro-Par 2008 Workshops - Parallel Processing (Vol. 5415, pp. 163-173). (Lecture Notes in Computer Science). Springer-Verlag. https://doi.org/10.1007/978-3-642-00955-6_21