Overview

Robert Cooper

Robert Cooper is currently the President, CEO, and Chairman of the Board of Atlantic Aerospace Electronics Corporation. Previously, he served simultaneously as Assistant Secretary of Defense for Research and Technology and Director of the Defense Advanced Research Projects Agency (DARPA). Under his directorship, DARPA moved into areas such as programs in advanced aeronautical systems, gallium arsenide microelectronic circuits, new-generation computing technology, and artificial intelligence concepts. Bob has also been the Director of the NASA Goddard Space Flight Center and the Assistant Director of Defense Research at MIT's Lincoln Laboratory. Bob holds a doctorate from MIT in electrical engineering and mathematics.

When I was at Goddard, we started the first massively parallel processor that was built, and it subsequently functioned at Goddard for many, many years. Interestingly enough, as I walked into this room to be on this panel, one of the folks who was on that program sat down next to me and said that he remembered those days fondly.

I'm really quite impressed by this group, and I subscribe to the comment that I heard out in the hallway just before the first session. One person was talking to another and said that he had never seen such a high concentration of computing genius in one place since 1954 at the Courant Institute, when John von Neumann dined alone. Be that as it may, I am nevertheless confident that if anything can be made to happen in the

high-end computer industry in this country, this group can play a key role in making it happen.

That comment also goes for the panel today, which is going to attack the problems of technology and perspectives for the future. We actually are starting this conference from a technical perspective by looking at the future—considering the prospects for computation—rather than looking toward the past, as we did in the first session.

Before we get started with our first speaker, I'd like to say a couple of words about what I see happening to the technology of high-end computing in the U.S. and in the world. Basically, the enabling technologies for high-end computing are the devices themselves. The physical constraints are the things that you will hear a lot about in this session: the logic devices; the memory devices; the architectural concepts, to a certain extent, which are determined by how you can fit these things together; and the interconnect technologies.

The main issue with technology developments in this area in this country is that we are somehow unable to take advantage of all of these things at the scale required to put large-scale systems together, and that is one of the reasons why we started the Strategic Computing Initiative back in 1983 at the Defense Advanced Research Projects Agency (DARPA), and that is why I think we are all hanging so much hope on the High Performance Computing Initiative that has come out of the study activity at DARPA and at the Office of Science and Technology Policy since about 1989.

I think it is the technology transition problem that we have to face. There is a role for government and a role for industry in the transition. I have been associated with some companies recently who have tried to take technology that they developed or that was somewhat common in the industry and make products out of it. I think that before we finish this particular session, we should talk about the issue of technology transition.