Publisher's Synopsis
This monograph outlines the development of the DAP (Distributed Array of Processors) system which provides massively parallel, fine grain processing. The selection of papers covers the history, programming and some of the applications of the DAP.;DAP used to mean "Distributed Array Processor" because processing power was distributed throughout a memory module; recently the meaning of the acronym has been changed to "Distributed Array of Processors" in order to explicitly distinguish DAP from other architectures designed to process arrays, but which use techniques such as vector pipelining rather than arrays of processors. Another view of DAP is to describe it as "Active Memory".;It was designed to exploit LSI technology, which put the emphasis on simple units repeated many times, rather than a smaller number of more complex units. DAP is not seen as a self-contained computer system, but rather acts under the control of a conventional computer acting as a host. This host system, which initially was a mainframe computer, provided facilites such as filestore, compilation and editing in a familiar environment, as well as controlling the DAP.
