The purpose of this one-year feasibility study was to design a high-speed scientific computer based on the following requirements:
the fastest scalar machine available in the world at the time of delivery, with execution speed competitive with the leading vector supercomputers for all but the most vectorised applications
strict adherence to market standards (such as UNIX, X-Open, OSI, RISC)
price/performance ratio competitive with minisupercomputers for all but the most vectorised applications
a product which it will be a pride to own: reliable, easy to operate and maintain, and with excellent add-on products
rapid entry into the high-end scientific and technical computing market
Design options were evaluated and a global architecture defined. The work continues in the AMUS project (2716).