Publisher's Synopsis
Over the last few years the need for larger machines to handle the faster processing requirements demanded by such advanced applications as image processing, robotics, speech recognition and artificial intelligence has overtaken the capacity of the older sequential architecture machines. A new class of highly parallel architecture machines is now available, resulting in faster information processing.;This study reflects recent developments, and is the result of three years' research in parallel computer architectures. It begins with an overview of some of the parallel processors available, namely the ICL-Distributed Array Processor, the CRAY and CYBER series of super computers and such purpose-built machines as the Loughborough NEPTUNE. It describes in detail, starting from basics, how a series of numerical algorithms can be applied to parallel array processors in general, and the ICL-DAP in particular, and illustrates and explains some of the problems suffered by readers when changing from standard sequential machines to highly parallel architecture machines.