We were requested to predict the future specifications of the electronic materials, electronic packages and micro processors for high performance computer systems going as far out as the year 2000 and beyond. These were to be based on the emerging requirements of the end users. Literature and conferences yielded very little quantitative data. As a result, we understood to design and develop a deterministic and macroscopic model based on physics, math, VLSI design, architecture and integration level considerations. The resulting model transcended the electronic world from system performance to microprocessor description to packaging specifications and required material's specifications. Based on this model, it is likely that the world will see an integration of 100 million transistors on a single chip providing clock frequencies approaching 213 MHz and CPU power near about 238 MIPS. Transportation is a major business for Alcoa. We have used our model to examine what architectural directions the Mobile 2000 would take for satisfying its command and control, navigational, entertainment, communication, engine, transmission and body/cockpit control and management systems.
A scenario is developed based on the current and emerging requirements for the above functions leading to the automobile of the year 2000. This is then translated to yield control and system requirements which then are converted to the total computational requirements. The requirements are used to define the future system specifications based on our model's predictions. Alternative architectures are then presented. Cautions and caveats are discussed based on our experience on several large scale Computer Integrated Manufacturing (CIM) networks', design development and Implementation. It is imperative to bring the large scale systems thinking to Mobile 2000's automation and support systems so that the classical pitfalls of cost, performance, lack of planning, ad hoc and proliferation of incompatible systems does not take place in this exciting arena.