: Scientific Parallel Computing

University Press. All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from the publisher, except for reading and browsing via the World Wide Web. Users are not permitted to mount this file on any network servers. Introduction If you walk the footsteps of a stranger, you will learn things you never knew you never knew—Pocahantas, by Disney This chapter introduces many of the basic notions of parallel computation. In Section 1.1 we give a short overview of the book, and Section 1.2 attempts to define what we mean by parallel computing. Section 1.3 introduces the critical topic of performance, which is central to the entire subject. In Section 1.4 we describe some of the motivating factors for the development of parallel computers. This is followed by some examples of parallelizing computational problems. The first two examples (Section 1.5) are quite simple, but serve to introduce many of the most important concepts. The next examples (Section 1.6) help to introduce further concepts as well as to provide some numerical applications that will be developed more in the text. Section 1.7.1 puts into context the role of parallel computation in solving technical problems, and Section 1.7.2 (also see Section 2.7) considers parallelism in a broader context. 1.1 OVERVIEW Parallel computing enables simulation in a variety of application areas which would not be possible with sequential processing alone. To use it effectively, there are diverse subjects that must be understood. This book focuses on three main areas that contribute to overall understanding of parallel computing: algorithms, architecture, and languages. All of these are essential contributors to solving problems of interest, which we refer to as " applications " in Figure 1.1. A basic understanding of computer architecture is needed to understand and predict the behavior of programs on different machines. A variety of fundamentally different computer architectures are commercially available today. Some differ substantially in the way they are programmed and the performance that can result. An introduction to computer architecture is presented in Chapter 3 to allow us to compare and contrast existing options. Some designs can be seen to be less appropriate for certain applications based on the simple analysis presented there. Languages APPLICATIONS Figure 1.1 Knowledge of algorithms, architecture, and languages contributes to effective use of parallel …