Background to the Supercomputer Project

A number of factors converged in the late 1970s to produce the Super-computer Project. Not necessarily in this order of significance, the first was the reaching of the limits of some of the core computer technologies and the desire to develop newer, potentially efficient, yet still highly uncertain, technologies. More specifically, increasing processing speeds had begun to be limited by the sequential structure of the conventional von Neumann computer architecture thus aiding the search for alternative parallel processing architectures. Although VLSI technology had considerably increased the performance of silicon memory and logic devices, new materials such as gallium arsenide held out the hope of substantially faster electron mobility and therefore processing speed. Devices operating at cryogenic temperatures, such as Josephson junctions, if they could be developed in a cost-effective way, similarly appeared to offer a potentially viable way of greatly increasing electron mobility. However, as will be documented in more detail shortly, research on these technologies was impeded by the significant degree of uncertainty that existed regarding longer-run outcomes. A government-funded project in this area appeared to offer a solution to the problem of uncertainty and underinvestment in Japan in these technologies and provide a way of further developing Japanese technological capabilities in the fields of computing and electronic devices.

The second factor influencing the establishment of the Supercomputer Project, as in the case of the earlier national computer projects analysed in this book, was the ‘IBM factor’.