Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Introduction RALPH E. GOMORY In the first of three papers on computers of the future, James D. Meindl discusses, or rather gives the reader a feeling for, the hardware. Because there are too many elements of hardware to permit covering each specifically, he discusses what is probably the key component, the field effect transistor (FET). In particular, Dr. Meindl discusses where FETs in ultra large scale integration are going. To do that, he considers a hierarchy of theoretical and practical limits ranging from those imposed by the laws of physics to those imposed by properties of materials, devices, circuits, and sys- tems. To project future developments, he extrapolates from past de- creases in the feature size of FETs, increases in the die or chip size, and improvements in the "cleverness" of circuit design. He also presents some thought-provoking analogies between the development of iron technology during the industrial revolution for our structural needs and the ongoing development of silicon technology for our electronic needs. In the second paper Herbert Schorr explains how hardware is put together into systems of orthodox design. "Systems" will include the hardware and the software that makes it all work together. The growth will come from both the continued improvements in technology that will reduce cycle times and the improvements in the design of processors that will reduce the number of cycles required per instruction. As Dr. Schorr indicates, we will continue to see multiple-machine environments combining uniprocessors of ever-increasing power. Micro- processors and powerful workstations offer the possibility of alternative 3
4 COMPUTERS OF THE FUTURE architectures to today's configurations where most processing is done by the central host. System interconnection and distributed processing will play an increasingly important role. And finally, Michael L. Dertouzos has the challenging task of going beyond the orthodox machines and the orthodox architectures; he ex- plains how one might put together machines of a totally different design. Computer networks provide means for autonomous, geographically dis- tributed computer systems, while tightly coupled multiprocessor systems provide for enormous, affordable, processing power. He also discusses what has been perhaps the great bottleneck of the field, which is the development of the software not its running, not its structure, but how to bring it into existence. He envisions the tran- sition of the software development process from the artisan to the mass production era. Programs will be more intelligent and flexible for ap- plications. So, in brief, this set of papers discusses the hardware, the area in which rapid advances made computers economically important and make them more important every year; the orthodox systems; and the unor- thodox systems and the creation of software. v