The Sourcebook of Parallel Computing

With the widespread introduction of small inexpensive shared-memory processor systems and clusters of PCs, more and more scientists and engineers and educational institutions can realistically consider buying and using parallel systems to solve problems They all need to learn how to program these systems efficiently. With over 60 contributors, The Sourcebook of Parallel Computing provides a broad overview to parallel computers and parallel computing by first examining the architecture of modern parallel computer systems and identifying key considerations for programming them. Through specific application studies, the authors demonstrate how to identify appropriate software and algorithms and the significant implementation issues.