ems.
Processing is the name given to what computers do to information placed into them (Senn, 1990). There are different ways in which computers go about this task. "Modern computing, as well as the historical development of computing, has been dominated by sequential monoprocessing" (Zakharov, 1984, p. 45). Sequential processing means that all data and instruc tions "must be passed back and forth . . . through a single link to a central memory" (Douglas, 1983, p. 71). The potential to exploit contemporary developments in com puter science to maximum advantage is threatened by the sequen tial processing requirements of existing computer architecture. The development of very large scale integrated (VSLI) circui try, means that it will be possible to "put thousands of pro cessors on a single silicon chip . . . If central storage of data and partial results is maintained, the programmer may be hardpressed to keep lots of processors busy" (Douglas, 1983, p. 71). The processing bottleneck characterizing the architecture of contemporary computers is a major problem, because even "with separate streams of data routed to multiple processors, conventional computers are limited by the need for central con trol and for communication based on addresses within a large memory" (Douglas, 1983, p. 72). One means of minimizing the bottleneck problem is the use of a data flow computer with a
circular pipeline to carry a continuous stream of data and instructions, which "can be executed without central control once they have sufficient data" (Douglas, 1983, p. 72).
A second significant problem involving emerging computer technology and sequential processing is that concerning the growing demands for local area networks (LAN). These networks tie together large numbers of electronic data processing devices for the purposes of sharing data and peripherals (Schindier, 1984,
...