Libraries, Numerical
- PDF / 68,625,276 Bytes
- 2,211 Pages / 565.087 x 755.008 pts Page_size
- 28 Downloads / 218 Views
David Padua (Ed.)
Encyclopedia of Parallel Computing With Figures and Tables
123
Editor-in-Chief David Padua
University of Illinois at Urbana-Champaign Urbana, IL USA
ISBN ---- e-ISBN ---- DOI ./---- Print and electronic bundle ISBN: ---- Springer New York Dordrecht Heidelberg London Library of Congress Control Number: © Springer Science+Business Media, LLC All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Springer Science+Business Media, LLC, Spring Street, New York, NY , USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com)
Preface Parallelism, the capability of a computer to execute operations concurrently, has been a constant throughout the history of computing. It impacts hardware, software, theory, and applications. The fastest machines of the past few decades, the supercomputers, owe their performance advantage to parallelism. Today, physical limitations have forced the adoption of parallelism as the preeminent strategy of computer manufacturers for continued performance gains of all classes of machines, from embedded and mobile systems to the most powerful servers. Parallelism has been used to simplify the programming of certain applications which react to or simulate the parallelism of the natural world. At the same time, parallelism complicates programming when the objective is to take advantage of the existence of multiple hardware components to improve performance. Formal methods are necessary to study the correctness of parallel algorithms and implementations and to analyze their performance on different classes of real and theoretical systems. Finally, parallelism is crucial for many applications in the sciences, engineering, and interactive services such as search engines. Because of its importance and the challenging problems it has engendered, there have been numerous research and development projects during the past half century. This Encyclopedia is our attempt to collect accurate and clear descriptions of the most important of those projects. Although not exhaustive, with over entries the Encyclopedia covers most of the topics that we identified at the outset as important for a work of this nature. Entries include many of the best known projects and span all the important dimensions of parallel computing including machine design, software, programming languages
