Mr. Chairman and members of the Subcommittee, thank you for the opportunity to testify today in support of the National Science Foundation's budget request and to comment specifically on the computing research programs in the Computer and Information Science and Engineering Directorate (CISE). I am Fred W. Weingarten, Director of Government Affairs for the Computing Research Association (CRA). CRA is an association of academic departments of computer science and engineering and industrial laboratories that conduct fundamental research. In the computing field. Our membership constitutes nearly every Ph.D. granting department in the U.S. and Canada, as well as over a dozen industrial laboratories and four affiliated professional societies concerned with computing research.
The CRA commends the work of your Committee for its past efforts on behalf of research. You have helped protect the nation's basic research infrastructure in the face of excruciating budget pressures. That you have done as well as you have up to now is testimony to your leadership and to your understanding that our basic research establishment is one of the nation's critical resources.
Clearly, those budget pressures aren't getting any easier, however; and I am here representing the computing research community to urge you to continue the strong support this Congress has shown for the National Science Foundation by appropriating the full amount requested, a four and a half percent increase overall. This increase includes an 8.6% increase in funding for the Computer and Information Science and Engineering (CISE) Directorate, which supports fundamental research in computing, communications, and information. This research will be important to our nation's future economic development and international leadership well into the 21st century.
I want to use my brief time here to focus on three key points:
A. We are in a particularly critical "window of opportunity" with respect to information systems, in which continued investment in fundamental research promises to have enormous payoff.
B. The programs of the NSF CISE Directorate are a key component of the Federal Government's R&D portfolio in the computing research.
C. Reaping the social benefits of R&D requires a complex and close relationship between fundamental research (mainly at universities,) industrial research and development, and government.
This year marks the 50th anniversary of the first programmable digital computer. That breakthrough, which may well be looked back on as one of the most important technological achievements of all time, was made possible by a unique partnership between government (particularly the Navy and the Bureau of Standards, academic researchers at institutions such as the University of Pennsylvania, MIT, and Iowa State University, and industrial researchers at IBM, NCR, and other firms. These government efforts were, of course, motivated by military needs, but they were built on the shoulders of earlier basic research, some of it dating back to 19th century theoretical mathematics that had, until then, been thought to be totally lacking in utility.
What fifty years has wrought! My acquaintance with computers doesn't go back quite that far, but I do recall when, as a post-doctoral fellow at Lawrence Livermore Laboratory, I had access to one of the biggest collections of supercomputers in the world. They were housed in two buildings the size of basketball arenas. The computer power I saw at Livermore was awesome for its time, and worth many millions of dollars. But, it is absolutely dwarfed by the power of the few thousand dollars worth of desk-top computers we use now in CRA's small office. I wrote this testimony on a computer thousands of times more powerful than the biggest machine available when I was a post-doc.(And, of course, impressive as this growth has been, we have only just begun!)
Much more important than this improvement in sheer power and cost, though, is the way use of computers has moved into so many aspects of human life. In those days, computers were expensive, rare, and used for serious business-weapons design, large-scale data processing, and so on. Computers rested in sealed rooms, accessible only by highly trained experts.
But, about two decades ago, a threshold was passed. With the advent of the desktop computer, people began to apply computer power to helping with everyday tasks. At CRA, for instance, we use computers to write documents and papers, and to layout and publish a bimonthly newsletter. We design and publish brochures and special reports, maintain membership lists, and manage CRA's electronic information services on the Internet, where computing researchers can find information ranging from job announcements to statistical data on degree production.
This broadening of use rests in part, of course, on the enormous drop in price, driven by a highly competitive industry and enabled by federally funded fundamental research. Otherwise, we couldn't buy the equipment. But, the applications listed above, and countless more, also stem from an enormous base of federally funded research. Fundamental research on visualization, on data bases, on object-oriented programming, on packet-switched data communications, and on graphical user interfaces, including even the mouse, has helped make information technology useful and useable to many more people. All of this research was started when computers were still expensive and rare, but it was based on the vision that someday appropriate research investment could not only make computers cheap and plentiful, but could also make them widely useable.
Three observations stem from this example:
1. The goal for computing research is not just to make faster and cheaper computers, but to make them more useful and useable. Fundamental research helps us understand how to use computers effectively for more tasks, to make them more easily used, to make them more reliable and secure, and so on. We have a large list of problems to work on. For instance, we need better fundamental concepts and tools for designing complex, large systems. There is renewed interest in understanding the future use of information technology in education, and we need to understand better how to organize very large collections of information so that it can be easily searched and used. We need to understand how to better build "intelligence" into the systems we design. We need to begin research and planning for the next generation of high speed data communication technology, as current Internet technologies seem to be reaching some natural limits. These are just a few higlights from a much bigger and evolving agenda.
2. Those goals are difficult and will take time to meet. Ten years from now, the desktop computer will have the power of today's supercomputer and will be connected to a high speed network of other machines around the world. The fundamental computing research of today will determine whether we will tap the full potential of such future technology.
3. The firms that will earn money on making these applications available in the marketplace may not even exist yet. The output of today's fundamental research will create the basis for future proprietary innovations. Just as companies like Apple, Microsoft, Sun, Silicon Graphics, and hundreds more arose to refine, package and market the results of the early research, so even now new firms are arising to commercialize Internet-type services and applications of World Wide Web technology, for example.
However, these agencies support research in accordance with their own missions. Those missions can change. Even more frequently, the agency perception of the technological needs for their mission can change. Programs can move rapidly from one technological area to another. They can move from short-term focus to a long-term focus, and then shift back again depending on budgetary and programmatic pressures. The health of the fundamental research community depends on a strong National Science Foundation core support for the broad field.
The term, "balance wheel" has often been used in the past, but that term is too diminutive. Rather, NSF is the heart of federal funding of long-term academic research in computing and communications, and in our growing information based society. With strong NSF programs, mission agencies will have a strong, existing research base they can turn to in addressing their own future needs. Furthermore, research that advances broad, long term societal interests but that may not fit any specific agency mission has a chance for funding on its merits. Finally, NSF has always exercised a primary responsibility for nurturing this young field, of helping to establish and sustain computer science and engineering in the nation's research universities as a field of research and graduate education.
A little over a year ago, the Computer Science and Telecommunication Board of the National Academy of Sciences released a report on the HPCC program called Evolving the High Performance Computing and Communications Initiative to Support the Nation's Information Infrastructure.. The report contained a diagram on page 2 that was, in our view, compelling. It showed the way ideas flowed between academic research laboratories, government labs, and industry, and it demonstrated that simple linear models of this relationship were wrong. That is, researchers don't simply come up with a new idea and throw it over the transom to the industrial developer, never top see it again. Rather, there is a two-way flow of ideas, back and forth.
Sometimes, new fundamental research ideas come from attempts to create new products. Many years ago, for example, it was only after many attempts by users to create programming languages (in order to make their own lives easier) that academic computer scientists were able to develop a theoretical basis for creating and translating programming languages. Similarly, today, the development of the Java language by Sun Microsystems has stimulated renewed interest and directions in distributed processing environments.
More commonly, ideas flow from academic research into industry or application laboratories, and there, they spawn new questions that, in turn, flow back to the fundamental research community.
We don't know exactly where technology is headed, but we are certain that things will be very different in the new information world. The key question is the degree to which this country will sustain its historical leadership in shaping that future by supporting the fundamental research necessary now to tap fully that potential. It would be a shame to leave it to others to capitalize on a technology base that was to a large extent created by past U.S. government funded research.
We are confident that you will not let this happen, but we also know the task will not be easy. The Computing Research Association, and the community we represent appreciates the past support of this Subcommittee and are ready to assist you in what is surely a difficult job. I would be happy to answer now or later any questions your committee might have, or provide any new information.
Copyright © 2004 Computing Research Association. All Rights Reserved. Questions? E-mail: webmaster@cra.org.