How to Build a Beowulf: A Guide to the Implementation and Application of PC Clusters (Scientific and Engineering Computation) - page 2

cover.jpg

 

Contents
Series Foreword
xiii
Foreword
xv
Preface
xix
1 Introduction
1
1.1 A Brief History
2
1.2 The Beowulf Book
4
2 Overview  of Beowulf Systems
9
2.1 What Is a Beowulf?
9
2.2 A Taxonomy of Parallel Computing
12
2.3 Benefits of Beowulf
13
2.4 A Critical Technology Convergence
14
2.5 The Beowulf System Node
16
2.6 The Beowulf Network
18
2.7 Linux
19
2.8 Message Passing for Interprocessor Communication
20
2.9 Beowulf System Management
21
2.10 The Beowulf Challenge
21
3 Node Hardware
25
3.1 Overview of a Beowulf Node
26
3.1.1 Principal Specifications
28
3.1.2 Basic Elements
29
3.2   Processors
32
3.2.1 Intel PentiumII
33
3.2.2 DEC Alpha
34
3.2.3 AMD K6
34
3.2.4 Future Processors
34

 

Series Foreword
The world of modern computing potentially offers many helpful methods and tools toscientists and engineers, but the fast pace of change in computer hardware, software, and algorithms often makes practical use of the newest computing technology difficult. The Scientific and Engineering Computation series focuses on rapid advances in computing technologies and attempts to facilitate transferring these technologies to applications in science and engineering. It will include books on theories, methods, and original applications in such areas as parallelism, large-scale simulations, time-critical computing, computer-aided design and engineering, use of computers in manufacturing, visualization of scientific data, and human-machine interface technology.
The series will help scientists and engineers to understand the current world of advanced computation and to anticipate future developments that will impact their computing environments and open up new capabilities and modes of computation.
This volume is about the increasingly successful distributed/parallel system called Beowulf. Beowulf is a cluster of PCs interconnected by network technology running Linux operating system and employing the message-passing method for parallel computation. Key advantages of this approach to parallel computation are (a) high performance for low price, (b) system scalability, and (c) rapid adjustment to new technological advances.
The book provides a complete description of the Beowulf hardware, software, programming, and selected applications. It can be used as an academic textbook, as well as a practical guide for designing and implementing Beowulf for those in science and industry who need a powerful system but are reluctant to purchase an expensive system such as MPP or vector machine.
As any other alternative parallel system the Beowulf system has technical challenges created by relatively large communication latencies. To perform well Beowulf must run suitable algorithms characterized by infrequent communication and medium to large granularity of parallel tasks. Fortunately there are many practical large-scale problems that satisfy this requirement. The book provides detailed examples illuminating the algorithmic performance and scalability of Beowulf.
JANUSZ S. KOWALIK