Advances in digital computer technology since the middle decades of the twentieth century have transformed many subject areas in astrophysics. Topics which had previously been dealt with by analytic approximations (usually to a very limited number of special cases) suddenly became amenable to detailed numerical modelling for all cases. Frequently, insights derived from this modelling ran ahead of other techniques in predicting physical phenomena before they were either observed or discerned in purely analytic treatments.

It has sometimes been said that the second half of this century has seen the advent of a totally new modus operandi in scientific research, which ranks alongside the two traditional approaches–experimentation and theory. The most powerful computers now available have greatly accelerated these developments. They employ simultaneous computational techniques (either vector processing or parallel processing, or both), and their throughput is so large that, for most problems, the only way in which the human mind can fully appreciate the scientific content of the numerical results being calculated is by transforming those results into pictorial representations.

This paper draws on my experience as Academic Director of the ANU Supercomputer Facility during the first eighteen months of its operation, to describe the place which I believe super-computers will occupy in the development of astrophysics during the 1990s and into the next century.