PARALLEL COMPUTING: ARCHITECTURES, ALGORITHMS AND APPLICATIONS
Advances in Parallel Computing This book series publishes research and development results on all aspects of parallel computing. Topics may include one or more of the following: high-speed computing architectures (Grids, clusters, Service Oriented Architectures, etc.), network technology, performance measurement, system software, middleware, algorithm design, development tools, software engineering, services and applications. Series Editor:
Professor Dr. Gerhard R. Joubert
Volume 15
Volumes 1–14 published by Elsevier Science. ISSN 0927-5452
Parallel Computing: Architectures, Algorithms and Applications Edited by
Christian Bischof RWTH Aachen University, Germany
Martin Bücker RWTH Aachen University, Germany
Paul Gibbon Forschungszentrum Jülich, Germany
Gerhard R. Joubert TU Clausthal, Germany
Thomas Lippert Forschungszentrum Jülich, Germany
Bernd Mohr Forschungszentrum Jülich, Germany
and
Frans Peters Philips Research, The Netherlands
Amsterdam • Berlin • Oxford • Tokyo • Washington, DC
Previously published as Volume 38 in the Publication Series of the John von Neumann Institute for Computing (NIC), ISBN 978-3-9810843-4-4.
© 2008 The authors and IOS Press. All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without prior written permission from the publisher. ISBN 978-1-58603-796-3 Library of Congress Control Number: 2007943213 Publisher IOS Press Nieuwe Hemweg 6B 1013 BG Amsterdam Netherlands fax: +31 20 687 0019 e-mail:
[email protected] Distributor in the UK and Ireland Gazelle Books Services Ltd. White Cross Mills Hightown Lancaster LA1 4XS United Kingdom fax: +44 1524 63232 e-mail:
[email protected]
Distributor in the USA and Canada IOS Press, Inc. 4502 Rachael Manor Drive Fairfax, VA 22032 USA fax: +1 703 323 3668 e-mail:
[email protected]
LEGAL NOTICE The publisher is not responsible for the use which might be made of the following information. PRINTED IN THE NETHERLANDS
Parallel Computing: Architectures, Algorithms and Applications, C. Bischof et al. (Eds.) © 2008 The authors and IOS Press. All rights reserved.
Preface Parallel processing technologies have become omnipresent in the majority of new processors for a wide spectrum of computing equipment from game computers and standard PC’s to workstations and supercomputers. The main reason for this trend is that parallelism theoretically enables a substantial increase in processing power using standard technologies. This results in a substantial reduction in cost compared to that of developing specialised high-performance hardware. Today the processing capacity of a desktop PC with a multicore processor supersedes the compute power of a supercomputer of two decades ago at a fraction of the cost. The utilisation of such powerful equipment requires suitable software. In practice it appears that the construction of appropriate parallel algorithms and the development of system and application software that can exploit the advantages of parallel hardware is not a simple matter. These problems have been studied for nearly five decades and, although much progress was made in the areas of parallel architectures, algorithm and software design, major problems remain to be addressed. The increasing replication of p