This paper reviews areas where a perspective from social research could throw light upon the development of parallel computers. Loosely put, the latter are computers in which more than one processor operates simultaneously on different parts of a task. They differ from the more conventional sequential, or von Neumann, computer, which has a single processor which operates on a task one step after another. The largest single question to be researched is why, despite important earlier projects, it has only been in the 1980s that we have seen a real upsurge of interest in parallel computing. More particular questions include policy issues to do with the promotion of parallel computing; the history of the Distributed Array Processor (a British parallel computer); the development of the transputer (a British microchip specially designed for parallel computing; national differences in approach to parallelism; whether attitudes to the organisation of society affect attitudes to analogous questions in the internal technical organisation of the computer; the social construction of measures of computer performance; and the social bases of credibility of technical arguments, particularly those pertaining to the claimed 'trajectory' of microchip development.
David MacKenzie (Fri,) studied this question.