According to one view of science and technology, periods of normal progress
are punctuated by discontinuities, or revolutions. An opposing view holds
that revolution itself is the norm. While the truth probably lies between
these positions, we have come to accept a much more rapid rate of change.
Certainly, the Internet, with its three-month Web years, suggests we are in
an era where nothing is likely to remain "normal" for very long.
Yet, exciting as such change may be, corporations like IBM which create that
technology demand predictable evolution. For us in
Research, the challenge is to continually extend our core technologies to allow
not just rapid progress, but truly revolutionary advances. These revolutions
are at the heart of our commitment to exploratory programs.
In one area -- the semiconductor technology that forms the core of computers -- we have enjoyed 40 years of steady increases in processor speeds and memory density. Driving that progress has been a wide range of innovations that, collectively, have made it possible to fabricate ever-smaller devices. The trend itself has become known as Moore's Law, but the scaling laws that guide that continual shrinkage were formulated here in Research, by Bob Dennard.
While we may regard the doubling of transistors on a chip every 18 months as normal progress, the technologies enabling that evolution are requiring greater departures from "standard" techniques. And sometime in the first quarter of the next century, physical limits will very likely prevent us from making still smaller silicon transistors of the kind found in today's computers.
Does that mean the end of progress in computer technology? I believe the answer is an unqualified no. I believe our research programs will continue to provide avenues of progress through entirely new technologies. This confidence in the future is fueled by an awareness of what our researchers have done in the past and the novel exploratory programs they are working on in the present.
While we may not be able to produce revolutionary innovations like clockwork, we do know how to promote such breakthroughs. Our exploratory research efforts in semiconductors, as well as in every other field of information technology, are dedicated to extending current technologies and inventing entirely new ones. It is that pursuit of innovation that ultimately allows IBM technology to meet the needs of its customers with regularity, and will keep Research at the core of IBM's future success.
Paul M. Horn
Senior Vice President and Director,
Research
