The past decade has been one of the most exciting periods in computer development that the world has ever experienced. Performance improvements, in particular, have been dramatic; and that trend promises to continue for the next several years.
In particular, microprocessor technology has changed rapidly. Microprocessors have become smaller, denser, and more powerful. Indeed, microprocessors have made such progress that, if cars had made equal progress since the day they were invented, we would now be able to buy a car for a few dollars, drive it across the country in a few minutes, and not worry about parking because the car would fit into one's pocket. The result is that microprocessor-based supercomputing is rapidly becoming the technology of preference in attacking some of the most important problems of science and engineering.
These processors are now the main stay of the workstation market.
The vendors of high-performance computing have turned to RISC microprocessors for performance.
Collections of these processors are interconnected by hardware and software to attack various applications. The physical interconnection of these processors may be contained in one or more cabinets as part of a multiprocessor, or the processors may be standalone workstations dispersed across a building or campus interconnected by a local area network. The effectiveness of using a collection of processors to solve a particular application is constrained by the amount of parallelism in the application, compiler technology, message passing software, amount of memory, and by the speed of the processors and of the interconnecting network.