Skip to main content

Practical Parallel Processing: An introduction to problem solving in parallel

Alan Chalmers, Jonathan Tidmus, Practical Parallel Processing: An introduction to problem solving in parallel. International Thomson Computer Press. ISBN 1-85032-135-3. March 1996. No electronic version available. External information

Abstract

The problems facing scientists and engineers today are computationally complex, limiting their solution on computers using traditional approaches. The evolutionary development of the serial computer may finally be reaching its zenith due to fundamental physical limitations. As computational expectations continue to rise, more users are viewing parallel processing as a means to circumvent these restrictions. In this way, a number of processors are able to co-operate in the solution of a single problem, offering almost unlimited possibilities for solving complex problems. Parallel processing offers the potential of solving real problems in acceptable times. This text is a comprehensive description of the techniques needed to solve complex problems on multiprocessor systems with special emphasis on those with distributed memory. The aim of this book is to provide an understanding of the necessary techniques, present a methodology as to how they may be applied, and demonstrate how their correct usage will improve system performance. Many of the ideas associated with parallel processing have been developed in the last ten years, and whilst textbooks exist, these have mainly dealt with descriptions of the architectures, been introductions to the field, or dealt with theoretical concepts. These texts provide a useful insight into the intricacies of parallel processing, but provide few guidelines as to how complex problems may be implemented successfully on a given multiprocessor system. Similarly, although the solutions of a number of science and engineering problems on large multiprocessor systems have been reported in the research literature, the details associated with the implementation are often omitted. People with complex problems are not necessarily skilled parallel programmers and will often need to collaborate with computer scientists who may have no understanding of the problems' underlying concepts. These observations point to the need for a comprehensive text which will enable any complex problem to be analysed to determine the criteria necessary for an efficient parallel implementation. This is such a text. As

Bibtex entry.

Contact details

Publication Admin