Parallel Computing- From Multicores and GPUs to Petascale - 1607505290 - IOS Press.pdf - Programowanie - neomnemonic

PARALLEL COMPUTING: FROM MULTICORES

AND GPU’S TO PETASCALE

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 19

Recently published in this series

Vol. 18. W. Gentzsch, L. Grandinetti and G. Joubert (Eds.), High Speed and Large Scale

Scientific Computing

Vol. 17. F. Xhafa (Ed.), Parallel Programming, Models and Applications in Grid and P2P

Systems

Vol. 16. L. Grandinetti (Ed.), High Performance Computing and Grids in Action

Vol. 15. C. Bischof, M. Bücker, P. Gibbon, G.R. Joubert, T. Lippert, B. Mohr and F. Peters

(Eds.), Parallel Computing: Architectures, Algorithms and Applications

Volumes 1–14 published by Elsevier Science.

ISSN 0927-5452 (print)

ISSN 1879-808X (online)

Parallel Computing: From Multicores

and GPU’s to Petascale

Edited by

Barbara Chapman

University of Houston, USA

Frédéric Desprez

INRIA, France

Gerhard R. Joubert

TU Clausthal, Germany

Alain Lichnewsky

GENCI, France

Frans Peters

Philips Research, The Netherlands

and

Thierry Priol

INRIA, France

Amsterdam

•

Berlin

•

Tokyo

Washington, DC

•

or transmitted, in any form or by any means, without prior written permission from the publisher.

ISBN 978-1-60750-529-7 (print)

ISBN 978-1-60750-530-3 (online)

Library of Congress Control Number: 2010923473

Publisher

IOS Press BV

Nieuwe Hemweg 6B

1013 BG Amsterdam

Netherlands

fax: +31 20 687 0019

e-mail: order@iospress.nl

Distributor in the USA and Canada

IOS Press, Inc.

4502 Rachael Manor Drive

Fairfax, VA 22032

USA

fax: +1 703 323 3668

e-mail: iosbooks@iospress.com

LEGAL NOTICE

The publisher is not responsible for the use which might be made of the following information.

PRINTED IN THE NETHERLANDS

Parallel Computing: From Multicores and GPU’s to Petascale

B. Chapman et al. (Eds.)

IOS Press, 2010

Preface

Parallel Computing technologies brought dramatic changes to mainstream computing.

This trend is accelerating as the end of the development of hardware following Moore’s

law looms on the horizon. The majority of standard PC’s and even notebooks today

incorporate multiprocessor chips with up to four processors. This number is expected to

soon reach eight and more.

These standard components, COTS (Components Off The Shelf), are increasingly

combined with powerful parallel processors originally designed for high-speed graph-

ics processing, GPU’s (Graphics Processing Units), and FPGA’s (Free Programmable

Gate Arrays) to build heterogeneous parallel computers that offer a wide spectrum of

high speed processing functions. The number of processors incorporated in such sys-

tems are today of the order of up to 10 4 to 10 6 . This vast number of processors allows

the construction of high speed computers in the petascale range, and even the exascale

range, at a reasonable cost. The limiting factor for constructing more powerful hard-

ware is the energy consumption and thermal control of such systems. Many research

efforts concentrate on reducing the overall energy consumed.

In addition to the hardware design and build limitations, the practical use of peta-

scale or exascale machines is hampered by the difficulties of developing software that

efficiently and effectively run on such architectures. This holds for system as well as

application software. The ParCo conference aimed at addressing many of these issues

through contributed papers as well as the various mini-symposia presentations.

In this book, which includes selected and refereed papers presented at the interna-

tional Parallel Computing conference (ParCo2009) held from 1–4 September 2009 at

ENS (École Normale Supérieure), Lyon, France, problems associated with the devel-

opment of high speed parallel systems using new hardware concepts and the associated

software development issues are considered. The papers were presented as keynote

papers, in regular sessions, an industrial session and various mini-symposia covering

specialised topics. Overall these give a snapshot of the state-of-the-art of parallel com-

puting technologies, both in hardware as well as application and software development.

This year’s highlight is no doubt the increasing number of papers addressing the

programming of general-purpose graphics processing units. Considering the main track

of the conference, as well as the mini-symposium dedicated to GPU’s, ParCo2009

turned into one of the main scientific venues covering this important research topic in

2009.

The editors wish to express their sincere gratitude for all persons who supported

this venture and lastly made it feasible. In particular we wish to thank the many re-

viewers who, as members of the international Program Committee, not only assessed

papers, but also acted as session chairmen during the conference.

Sincere thanks is due to the members of the Organising Committee, and in particu-

lar to Laurent Lefèvre, Eddy Caron and Jean-Christophe Mignot, who spent many

hours assisting in organising a very successful event. We are also very grateful for

work done by Virginie Mahdi from Genci, Paris in attracting a considerable number of

sponsors as well as participants in the Industrial Session.

Parallel Computing- From Multicores and GPUs to Petascale - 1607505290 - IOS Press.pdf

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika: