Ask the Scholar

Document scope · 1 page
doc
Scholar
Ask about this object, its catalog metadata, its source description, or the page inventory. For page-specific OCR and visual context, open one of the page chats.

Source Description

Records pertain to the Office of Science and Technology Policy.

Scholar Source Context

Document identity
localId
285790773
label
Science & Technology [4]
core
doc
dtoType
document
pageCount
1
Source metadata
id
285790773
contentType
document
title
Science & Technology [4]
description
Records pertain to the Office of Science and Technology Policy.
identifierLocal
04296-004
collections
Records of the Economic Policy Council (George H. W. Bush Administration)
Olin Lewis Wethington Subject Files
imageCount
1
hasImages
yes
source
import
hasTranscription
no
Source extras
naId
285790773
levelOfDescription
fileUnit
recordType
description
ocrSource
nara-archive
Single page context
seq
1
pageIndex
0
type
document
mediaId
1e862e7ef5a50b23
ocrText
Originally Processed With FOIA(s): FOIA Number: 2005-0336-F 2005-0336-F FOIA MARKER This is not a textual record. This is used as an administrative marker by the George Bush Presidential Library Staff. Record Group/Collection: George H.W. Bush Presidential Records Collection/Office of Origin: Economic Policy Council Series: Wethington, Olin, Files Subseries: Subject Files OA/ID Number: 04296 Folder ID Number: 04296-004 Folder Title: Science & Technology [4] Stack: Row: Section: Shelf: Position: G 13 28 4 2 A 1 CONTENTS 2 STATEMENT OF: PAGE 3 HON. ALBERT GORE, JR., U.S. SENATOR FROM TENNESSEE 2 4 HON. LARRY PRESSLER, U.S. SENATOR FROM SOUTH DAKOTA 12 5 HON. ROBERT W. KASTEN, JR., U.S. SENATOR FROM WISCONSIN 20 6 D. ALLEN BROMLEY, DIRECTOR, OFFICE OF SCIENCE 25 7 AND TECHNOLOGY POLICY 8 HON. LARRY PRESSLER, U.S. SENATOR FROM SOUTH DAKOTA 47 9 HON. CHARLES S. ROBB, U.S. SENATOR FROM VIRGINIA 54 10 DONALD LANGENBERG, CHANCELLOR OF THE UNIVERSITY OF 67 11 MARYLAND SYSTEM AT ADELPHI, MARYLAND 12 MALVIN H. KALOS, DIRECTOR, CORNELL THEORY CENTER 74 13 MR. TRACEY GRAY, VICE PRESIDENT OF MARKETING, 83 14 GOVERNMENT SYSTEMS DIVISION, US SPRINT 15 COMMUNICATIONS COMPANY LIMITED PARTNERSHIP 16 DR.. DAVID C. NAGEL, VICE PRESIDENT, ADVANCED 89 17 TECHNOLOGY, APPLE COMPUTER, INC. 18 DR. JOHN S. WOLD, EXECUTIVE DIRECTOR, LILLY RESEARCH 99 19 LABORATORIES, ELI LILLY AND COMPANY; ACCOMPANIED BY 20 DR. RIAZ ABDULLA, MANAGER, HIGH-PERFORMANCE COMPUTING 21 PROGRAM, ELI LILLY AND COMPANY 22 23 ) 24 25 ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 1 1 HEARING ON 2 S. 272 3 - HIGH PERFORMANCE COMPUTING AND COMMUNICATIONS 4 5 Tuesday, March 5, 1991 6 7 U.S. Senate 8 Subcommittee on Science, 9 Technology, and Space 10 Committee on Commerce, Science, 11 and Transportation 12 Washington, D.C. 13 The subcommittee met, pursuant to notice, at 2:04 p.m., 14 in Room SR-253, Russell Senate Office Building, Hon. Albert 15 Gore, Jr., presiding. 16 17 18 19 20 21 22 23 24 25 ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 2 1 OPENING STATEMENT OF HON. ALBERT GORE, JR., U.S. SENATOR 2 FROM TENNESSEE 3 Senator Gore: This subcommittee will come to order. 4 I would like to welcome Dr. Bromley and Dr. Wong and 5 distinguished guests, other witnesses, and ladies and 6 gentlemen. 7 I will have an opening statement, and then I will call on 8 Senator Pressler, and then we will move right into the 9 testimony of Dr. Bromley. 10 Today the Science and Technology Subcommittee is 11 considering S. 272, the High Performance Computing Act. This 12 bill is designed to ensure that the United States stays at the 13 leading edge in computer technology. It would roughly double 14 the Federal investment in research and development in new 15 supercomputers, more advanced software, and high-speed 16 computer networks. 17 Perhaps most importantly it would create a national 18 research and education network, the NREN -- or the National 19 Information Superhighway -- as I like to call it, which would 20 connect more than 1 million people at more than 1 thousand 21 colleges, universities, laboratories, and hospitals throughout 22 the country, giving them access to computing power and 23 information -- resources unavailable anywhere today -- and 24 making possible the rapid proliferation of a truly Nation- 25 wide, ubiquitous network which can do more to enhance our ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 3 1 Nation's productivity than any other single development. 2 These technologies and this network represent our 3 economic future. They are the smokestack industries of 4 today's information age. We talk a lot now about jobs and 5 economic development, pulling our country out of recession and 6 into renewal. 7 Well, our ability to meet the economic challenges of the 8 information age and beyond -- tough challenges from real 9 competitors around the world -- will rest in large measure on 10 our ability to maintain and strengthen an already threatened 11 lead in these technologies and industries. 12 We are witnessing the emergence of a much-heralded, 13 prematurely recognized -- on several occasions -- global 14 civilization. Now it is really here. And those nations best 15 able to deal with information will be the nations most 16 successful in this global civilization. 17 It is based on shared knowledge in the form of digital 18 code. And our ability to compete will depend on our ability 19 to handle knowledge in that form. It is now the lingua franca 20 of global civilization. 21 We used to think of our ability to compete in terms of 22 infrastructure -- did we have enough deep-water ports? Did we 23 have enough railroad lines or highways? Now we need to think 24 about information infrastructure. 25 I have been advocating legislation such as this for more ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 4 1 than 1 dozen years. Because I strongly believe it is critical 2 for our country to develop the best scientists, the best 3 science, the fastest, most powerful computers, and the best 4 base of knowledge -- and then to ensure access to these 5 technologies to as many people as possible, so as many people 6 as possible will benefit from them. 7 This legislation will help us do just that. 8 Every year there are new advocates. This year, finally, 9 President Bush is among them, including in his budget for 10 Fiscal 1992 $149 million in new funding to support these 11 technologies. 12 We cannot afford to wait or to put off this 13 challenge -- not if we care about jobs, economic development 14 or our ability to hold our own in world markets. 15 During the last 30 years, computer technology has 16 improved exponentially, faster than technology in any other 17 field. Computers just keep getting faster, more powerful and 18 more inexpensive. 19 According to one expert, if automobile technology had 20 improved as much as computer technology in recent years, a 21 1991 Cadillac would now cruise at 20,000 m.p.h, get 5,000 22 miles to a gallon, and cost only 3 cents. 23 [Laughter. ] 24 When my friend Jim Schlesinger heard someone deliver that 25 cliche recently, he said yes, and your Cadillac would be a ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800). FOR DEPO 5 1 few millimeters long, too. 2 [Laughter.] 3 But as a result of these amazing advances, computers have 4 gone from being expensive, esoteric, research tools, isolated 5 in the laboratory, to being an integral part of our every-day 6 life. We rely on computers at the supermarket, at the bank, 7 in the office, and in our schools. They make our life easier 8 and better in hundreds of ways. 9 And yet, the computer revolution is far from over. In 10 fact, according to some measures, the price/performance ratio 11 of computers is improving even faster now than it has in the 12 past. Anyone who has seen a supercomputer in action today, 13 has a sense of what computers can do for all of us in the 14 future. 15 Today, scientists and engineers are using supercomputers 16 to design better airplanes, understand global warming, find 17 oil fields, and discover safer, more effective medications. 18 In many cases, they can use these machines to mimic 19 experiments that would be too expensive or downright 20 impossible in real life. With the supercomputer model, 21 engineers at Ford can simulate auto crashes and test new 22 safety features for a fraction of the cost and in a fraction 23 of the time it would take to really crash an automobile. And 24 they can observe many more variables in much more detail than 25 they could with a real test. ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 6 1 The bill we are considering today is very similar to the 2 first title of S. 1067, the High Performance Computing Act of 3 1990, which passed the Senate unanimously last October. 4 Unfortunately, of course, the House was unable to act on the 5 bill before we adjourned because of differences in measures 6 passed by the two bodies. 7 It is my hope that we will be able to move this bill 8 quickly this year. There is widespread support in both the 9 House and the Senate. In the House, Congressman George Brown, 10 the new Chairman of the House Committee on Science, Space, and 11 Technology, has introduced a very similar bill -- H.R. 12 656 -- co-sponsored by Congressman Tim Valentine, Sherwood 13 Boehlert, Norm Mineta and others. 14 On Thursday, the Science Committee's Subcommittee on 15 Science and its subcommittee on Technology and Competitiveness 16 on the House side will be holding a joint hearing on the bill. 17 And I look forward to working with my House colleagues to move 18 this bill as quickly as possible. 19 This legislation provides for a multi-agency, high- 20 performance computing research and development program to be 21 coordinated by the White House Office of Science and 22 Technology policy, whose director, Dr. Alan Bromley, is our 23 first witness today. 24 The primary agencies involved are the National Science 25 Foundation, the Defense Advance Research Projects Agency, the ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 7 1 National Aeronautics and Space Administration, and the 2 Department of Energy. Each of these agencies has experience 3 in developing and using high-performance computing technology. 4 S. 272 will provide for a well-planned, well-coordinated 5 research program to effectively utilize the talents and 6 resources available throughout the Federal research agencies. 7 In addition to the agencies I just mentioned, it will involve 8 also the Department of Commerce -- in particular, the National 9 Institute of Standards and Technology, and NOAA, the 10 Department of Health and Human Services, the Department of 11 Education, the U.S. Geological Survey, the Department of 12 Agriculture, the Environmental Protection Agency, and the 13 Library of Congress as well. 14 The technology developed under this program will find 15 application throughout the Federal Government and throughout 16 the country. S. 272 will double funding for high-performance 17 computing at NSF and NASA during the next 5 years. Additional 18 funding, more than $1 billion during the next 5 years, will 19 also be needed to expand research and development programs at 20 DARPA and DOE. 21 Last year, I worked closely with Senators Johnston and 22 Domenici on the Energy Committee, to pass legislation 23 authorizing a DOE high-performance computing program. And I 24 hope to work with them and the other members of the Energy 25 Committee to see that program authorized and funded in FY ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 8 1 1991. Already, Senator Johnston and others have introduced S. 2 343, authorizing -- which would authorize DOE's part of this 3 multi-agency program. 4 To fund DOD's part of the program, last year I worked 5 with Senators Nunn and Bingaman and others on the Armed 6 Services Committee to authorize and appropriate an additional 7 $20 million for DARPA's high-performance computing program, 8 money that has been put to good use developing more powerful 9 supercomputers and faster computer networks. 10 Advanced computer technology was a key ingredient in 11 Operation Desert Storm. But we cannot simply rely on existing 12 technology. We must make the investment needed to stay at the 13 leading edge. It is important to remember that the Patriot 14 Missile and the Tomahawk Cruise missile rely on computers 15 based on technologies developed through Federal computer 16 research programs in the 1970's. 17 The High Performance Computing Act will help ensure the 18 technological lead in weaponry that has helped us win the war 19 with Iraq, and that will improve our national security in the 20 future. 21 This same technology is improving our economic security 22 by helping American scientists and engineers develop new 23 products and processes to keep the U.S. competitive in world 24 markets. Supercomputers can dramatically reduce the time it 25 takes to design and test a new product -- whether an airplane, ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 9 1 a new drug or an aluminum can. That means more energy- 2 efficient, cheaper products. It means higher profits and more 3 jobs for Americans. 4 But perhaps the most important contribution this bill 5 will make to our economic security is the National Research 6 and Education Network, the cornerstone of the program, funded 7 by the bill. 8 In 1996, this fiber optic computer network will connect 9 more than 1 million people at more than 1 thousand colleges 10 and universities in all 50 states, allowing them to not only 11 send electronic mail and shared data, but access 12 supercomputers and use research facilities such as radio 13 telescopes; log-on to databases containing trillions of bytes 14 of information on all sorts of topics. 15 This network will speed research and accelerate 16 technology transfer so that the discoveries made in our 17 university laboratories can be quickly and effectively turned 18 into profits for American companies. 19 Today the National Science Foundation runs NSF NSFNET, 20 allowing researches and educators to exchange up to 1.5 bytes 21 of data per second. The NREN will be at least 1,000 times 22 faster, allowing researchers to transfer all the information 23 in the entire Encyclopedia Britannica from coast to coast in 24 seconds. 25 With today's networks, it is easy to send documents and ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 10 1 data, but images and pictures require much faster speeds. 2 That will require the NREN, which can carry billions of bytes 3 per second. 4 That is important. Because one of the only ways we can 5 successfully deal with the mountains of excess data we now 6 have is by organizing it into coherent, mosaic patterns and 7 images which can be comprehended in gulps or chunks, instead 8 of byte-by-byte, one byte at a time. It is impossible to deal 9 with this much information in any other way. 10 With access to computer graphics, researchers throughout 11 the country will be able to work together far more effectively 12 than today. It will be much easier for teams of researchers 13 at colleges throughout America to work together. They will be 14 able to see the results of their experiments as the data comes 15 in. They will be able to share the results of their computer 16 models in real time, and brainstorm by tele-conference. 17 William Wulf, formerly Assistant Director for Computer 18 and Information Science at NSF likes to talk about the 19 "National Collaboratory" -- a laboratory without walls, which 20 this network will make possible. 21 Researchers throughout the country at colleges and labs, 22 large and small, will be able to stay on top of the latest 23 advances in their fields. The NREN and the other technology 24 funded by S. 272 will also provide enormous benefits to 25 American education. And I will include in the record a ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 11 1 lengthy description of how education will be improved by this 2 network. 3 In conclusion, let me say that with this bill, I believe 4 we can help shape the future and shape it for the better. 5 This is an investment in our national security and our 6 economic security which we cannot afford not to make. For 7 that reason, I was very glad to see the administration propose 8 a high-performance computing and communications initiative, a 9 program very similar to the one outlined in S. 272. 10 In intent to work closely with Dr. Bromley and others 11 within the administration, as well as my colleagues in 12 Congress, to secure the funding needed to implement this 13 critically important program. 14 Before turning to Dr. Bromley, I would recognize now, 15 Senator Pressler. I apologize for the length of my opening 16 statement. But after 12 years, I have a lot to say about this 17 topic. And I will now recognize my friend, the Senator from 18 South Dakota. 19 [The prepared statement of Senator Gore follows:] 20 21 22 23 24 25 ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 12 1 STATEMENT OF HON. LARRY PRESSLER, U.S. SENATOR FROM 2 SOUTH DAKOTA 3 Senator Pressler: Thank you, Mr. Chairman. 4 And I think it is very appropriate that we have Dr. 5 Bromley here at this opening hearing. I want to thank you, 6 Mr. Chairman, for holding this hearing on S. 272, and for your 7 early and continued leadership in support of a national 8 supercomputer network. 9 Let me say that I am glad that the administration has 10 come with the proposal this year. I challenged the 11 administration to come with a proposal last year. I am a 12 cosponsor of this legislation. But I look forward also to 13 working out the differences and listening very closely to the 14 administration, here today. 15 Let me -- rather than repeat everything that Al has said 16 in terms of what supercomputers can do -- to cite a special 17 interest that I have. And that is to be sure that small 18 businesses, smaller cities and towns, smaller universities, 19 Indian reservations, and others are taken care of this 20 communications transformation we are going through. 21 I am very interested, for example, in finding a way that 22 fiber optics cable can be laid to every household and business 23 in the United States, and not leave some out. Generally 24 speaking, in the Communications Subcommittee and this 25 Subcommittee, when people talk about supercemputers or fiber ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 13 1 optics cable, or new electronic services, they are talking 2 about serviceAthe to wealthy suburbs. They are talking about 3 serving densely populated areas where there is a market that 4 yields greatest profits. 5 I always cite as an example that in cable T.V., for 6 example -- my wife and I have just obtained cable T.V. at one 7 of our homes for the first time of last year, we live in 8 Washington, D.C. where we have a home, and Humbolt, South 9 Dakota. In neither place could we get cable T.V. until this 10 past year. We finally have it in Washington, D.C. And we can 11 get it by microwave now, in Humbolt, South Dakota. 12 But the point is that the rural and smaller town areas SMALL SCHO 13 have something in common with inner cities, in that all the they are not TALKING about ^ A merica 14 great talk about the communications revolution, people they are 15 talking about the largest universities, the wealthiest 16 suburbs, and so forth. And We cannot become a Nation of two 17 or three communication systems. 18 x I am working on legislation in the Communications 19 Subcommittee related to this that would provide that the 20 telephone companies, or whoever, or the cable companies should 21 be laying fiber optics cable rather than copper cable. That 22 we should get fiber optics to everybody. And from there we 23 can rent space on the fiber optics cable for the different determine how 24 users or figure it out somehow 30 these services can be POLICY. 25 available on a national basis sort a of universal service so- ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202)289-2260 (800) FOR DEPO 14 1 to speak. 2 And the same is true for supercomputers. Of course, 3 every small business does not need a supercomputer. But many 4 of our universities do -- many small businesses will benefit 5 from being hooked-up by fiber optics cable. In my own State 6 there are certain institutions. For example, at South Dakota 7 State University, the bio-stress lab where they are studying 8 new kinds of plants, resistance to drought, very much wants to 9 be connected to a supercomputer. Archives 10 The EROS Data Center, which has pictures the LANDSAT researchers 11 pictures, very much will benefit, and the country will the EROS DATABASE 12 benefit from being hooked to a supercomputer. 13 South Dakota School of Mines, the project such as the 14 deep-drilling project, certain EPSCoR-type projects, where 15 professors and individuals are doing research, and they are 16 not associated with a huge university will benefit a great 17 deal. 18 Indeed, you can build the argument that small hospitals 19 and small universities need this need access to a 20 supercomputer more than a big one. Because the big ones 21 probably have their own. So I think this is a very important 22 initiative. 23 Let me also say small business -- now not every small 24 business is going to want to be hooked to a supercomputer. 25 But if they have the option, as we move into the 1990's, there ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202)289-2260 (800) FOR DEPO 15 1 are jobs to be created in areas -- everybody does not have to 2 be in the Boston beltway or a place like that to do technical 3 work. And it will enrich our country. Now I am not saying 4 there is anything wrong with being in the Boston beltway area. 5 But we also need a few in the Sioux Falls beltway area, who 6 have access to a supercomputer. 7 [Laughter.] 8 I am also concerned, Mr. Chairman, with -- in addition to 9 this, and in this subcommittee -- I hope we look at this year 10 what is happening in our schools in math and science training. 11 I think we are falling behind in science and mathematics. 12 Someone who is 15 or 18 years old and has not gotten the 13 basics is out of the system forever. Where our educational 14 system works, I am told that over half of our graduate 15 students in engineering are from abroad. Part of that is due 16 to our pay structure. The thing to be is an investment banker 17 or a lawyer and move papers that result in no productivity but 18 the pay is pretty good. We have to think very hard about 19 the impact, the long-term impact of losing the math and 20 science edge. 21 There are a whole series of other things. I was going to 22 go into some of the benefits of supercomputers. The Chairman 23 has done that very aptly in his statement. I have mention 24 South Dakota State University's bio-tress laboratories is 25 developing strains of crops and livestock. With the ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 16 1 supercomputer network, the results of the researched produced 2 in the bio-stress labs could be placed on-line for ready 3 access_ for users at other universities that are doing similar 4 research. 5 And let me say that I think many of our smaller 6 universities very much want to be connected to some of the 7 work at larger universities -- certainly some of our smaller 8 hospitals very much are in need of this type of service. 9 I have mentioned the EROS data center archives -- the 10 pictures that are sent back from our LANDSAT 11 satellite -- since 1971, near Sioux Falls, it has stored over 12 1 million satellite images on magnetic tape. Government and 13 university researchers and commercial enterprises use the 14 LANDSAT images for oil and mineral exploration, land use 15 planning, map making, climate change monitoring, crop 16 assessments, and many other applications. 17 With the proposed network, any researcher could use the 18 LANDSAT pictures that are stored in South Dakota without 19 having to leave his or her desk or laboratory. This 20 capability will become even more important when the EROS data 21 center is used later in this decade to achieve the massive 22 amounts of data for the Earth Observing System, EOA, NASA's 23 contribution to our multi-agency global change research 24 program. And I was proud to be on hand last summer when we 25 raised the NASA flag at the EROS data center in Sioux Falls. ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 17 1 The EOS satellites will be transmitting the data 2 equivalent of the entire Library of Congress every 5 days. I 3 think that is really an amazing thing. To use this data 4 effectively, scientists will need ready access to the EROS 5 Data Center database. 6 Mr. Chairman, I also want to take this time to commend 7 President Bush for his support for our multi-agency 8 supercomputing initiative in its Fiscal Year 1992 fiscal year 9 budget request. President Bush has allocated $638 million to 10 a supercomputer network and research effort. The Bush plan, 11 which has the same purposes as S. 272, is patterned after a 12 program set forth in a 1989 report by the White House Office 13 of Science and Technology Policy. As I have mentioned, I am 14 pleased that Dr. Alan Bromley, the Director of that office is 15 testifying today. I look forward to hearing his details about 16 the administration's initiative. 17 I am hopeful that during this Congress the administration 18 and congressional supporters of supercomputing will join 19 forces to implement a national, multi-agency supercomputing 20 initiative -- boy that is a mouthful of Washington 21 words -- whether done by statute or otherwise. 22 Mr. Chairman, I am sure we will all agree that the main 23 goal here is to begin in this Congress a national commitment 24 to supercomputing and a national supercomputing network 25 linking more than 1 million of our country's computers by ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 18 1 1996. 2 Mr. Chairman, I look forward to hearing from our 3 distinguished panel of witnesses on both S. 272 and 4 supercomputing generally. 5 [The prepared statement of Senator Pressler follows:] 6 LEOMMITTEE INSERT] 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 19 1 Senator Gore: Thank you, Senator Pressler. 2 Senator Kasten. 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 20 1 STATEMENT OF HON. ROBERT W. KASTEN, JR., U.S. SENATOR 2 FROM WISCONSIN 3 Senator Kasten: Mr. Chairman, I am not going to try to 4 compete with you and Senator Pressler in terms of length of 5 opening statements. 6 So what I would like to do is ask unanimous consent that 7 the entire statement that I have appear in the record and that 8 I be allowed to excerpt a couple of quick points from it. 9 I first of all want to thank you for the hearing. And I 10 am pleased to be an original co-sponsor of this piece of 11 legislation. I agree with both of you that it is necessary 12 for the continued ability to innovate and compete in the 13 marketplace. 14 I am also pleased that though there are differences in 15 emphasis, the administration recognizes the needs for the 16 kinds of programs that we are supporting in S. 272. I believe 17 that the legislative initiatives, particularly S. 1067, in the 18 last Congress, Mr. Chairman, some of these initiatives and 19 other pressures from Congress is, in fact, what has helped the 20 administration -- helped assure the Administratif's attention 21 to this important area. And for this I commend you. 22 And I also commend you for you statement -- in your 23 opening statement saying that we want to work with the 24 administration. Because I believe that we are close on the 25 dollars and not SO far apart on the programs. And SO we have ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 21 1 the opportunity here now to come together. But I think it was 2 Congressional initiative, and particularly your leadership, 3 Mr. Chairman. 4 The future developments in the area of supercomputing, 5 hardware, software, and communications, are critical to the 6 country's security -- economically, environmentally, and 7 militarily. It is important that an American company -- it 8 also happens to be a Wisconsin company, Cray Research, is the 9 current world leader in the field. And I believe that our 10 bill will help to keep American supercomputer companies 11 healthy. And it will also assure that American companies, 12 researchers, and others who use supercomputers have the 13 software, the training and the access so that we are all using 14 our capabilities and using them to the fullest. 15 Both S. 272 and the administration's proposal look to 16 develop the networks that will be necessary if there is going 17 to be broader access and broader utilization of the Nation's 18 supercomputers. Through the computing capability of our 19 supercomputers -- although it is incredible -- our 20 communications systems are not up to the task of transmitting 21 the-data between computers or to -- as Senator Pressler 22 discussed. -- some remote locations. 23 So I think we are on the right track here. I will work 24 to assure that America realizes the full promise that 25 supercomputers can offer in SO many different areas of ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 22 1 scientific, educational, and commercial endeavor. And I think 2 that we, Mr. Chairman, have not only a hearing on a very 3 important subject today, but we have the beginnings now of a 4 mark-up of a bill that is going to be passed-out by the 5 Committee, going to be passed by the Senate, passed by the 6 House of Representatives, and, in fact, signed into law. 7 And that is, I know, your goal and your purpose. And I 8 just want to say that I am an anxious lieutenant on this team 9 or in this area that we are going to do something here. And 10 it is more than just simply a hearing on a very important 11 subject. And we all will work together. And I look forward, 12 Mr. Chairman, to working with you. 13 Unfortunately, I have another hearing that also started 14 at 2:00. And I would ask that in addition to my entire 15 statement appearing in the record, that I be permitted to 16 submit some questions for the record of Dr. Bromley. 17 Senator Gore: Absolutely, without objection. 18 Senator Kasten: Thank you, Mr. Chairman. 19 [The prepared statement of Senator Kasten follows:] 20 [COMMITTEE INSERT] 21 22 23 24 25 ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 23 1 Senator Gore: And may I say, before you must depart, on 2 a personal note, I deeply appreciate the key role that you 3 have played in pushing this bi-partisan effort forward. And I 4 want to say the same to Senator Pressler. 5 Indeed, this has been bi-partisan from the very 6 beginning. And you mentioned the cooperative relationship 7 between the committee and the administration. That is a 8 spirit we want to continue. 9 In fact, the OSTP plan which Senator Pressler mentioned 10 coming out in 1989, really came about because of legislation 11 this committee passed which both of you supported, the 12 Supercomputer Network Study Act of 1986 which required that 13 study, but more than requiring it, invited the administration 14 to join in a dialogue with the committee which resulted in the 15 OSTP plan 2 years ago. 16 And it is not accidental, by any means, that the 17 administration plan and the legislation which we here are 18 supporting, are so similar. Because they have both resulted 19 from a meeting of the minds on what is the best interests of 20 our country. 21 Dr. Bromley, again, let me apologize for the length of 22 time it has taken us to get to your statement. We are very 23 interested in it. 24 I want to acknowledge the presence of Dr. Eugene Wong, 25 Associated Director for Physical Sciences at OSTP. And among ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 24 1 the many things for which you are due congratulations, is the 2 excellent choices you have made of people to help you. 3 I-remember when you first took the post, you said you 4 were going to do it. I agreed with your assessment that this 5 was one of the keys to making this office a more integral part 6 of the Government. And congratulations on getting the 7 President on board on this initiative. 8 I know you would probably want me to choose different 9 words in saying that, but that is what I feel about it. I 10 think this is you speaking for the administration now. And we 11 welcome you. So please proceed. 12 13 14 15 16 17 18 19 20 21 22 23 24 25 ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 25 1 STATEMENT OF D. ALLEN BROMLEY, DIRECTOR, OFFICE OF 2 SCIENCE AND TECHNOLOGY POLICY 3 Dr.-Bromley: Thank you, Mr. Chairman. 4 I welcome this opportunity to appear before you and the 5 members of the committee to discuss this initiative. And I 6 would, in fact, begin by acknowledging the leadership that you 7 and this group have shown in this area that I agree is one of 8 the most important that we could address in terms of its long- 9 term and short-term impact on this Nation. 10 The hearing, of course, today addresses this whole 11 question of high-performance computing and communication. And 12 I would like to, with your permission, sir, have the formal 13 testimony that I have provided to the committee included in 14 the record and simply abstract it. 15 Senator Gore: Without objection, that would be fine. 16 Dr. Bromley: Thank you, sir. 17 Now, the President's initiative included in the 1992 18 budget is described in detail in the document that accompanies 19 that, precisely, the report that is called the Grand 20 Challenges: High Performance Computing in Communication. 21 This is a report that was prepared by a working group on high- 22 performance computing and communications that operates under 23 the Committee on Physical Sciences and Engineering, one of the 24 seven umbrella committees that operates under the Federal 25 Coordinating Council for Science, Engineering and Technology. ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202)289-2260 (800) FOR DEPO 26 1 And L I see, sir, that you already have a copy of the report. 2 The goals of the whole initiative high performance 3 computing and communication I think, sir, are ones that we 4 have shared for a long time. And my staff and I look forward 5 to working with you to realize these goals for the American 6 people. 7 In the document that we have provided to you, the Grand e 8 Challenges document, we have attempted to illustrate and 9 symbolize the importance of this activity by setting forth a 10 series of very important scientific and societal problems 11 whose solutions simply elude us at the moment, but would 12 become within reach with the kind of program that both of us 13 are talking about. the questions 14 These include matters of global climate change, mapping 15 the human genome, understanding the nature and helping in the 16 fabrication of tailored materials -- an entirely new and 17 frontier problems that are directly applicable to our 18 19 would agree with L us sir -- both military strength and economic national security, and that includes /ing -- and I believe you 20 strength. It also, of course, includes the design of ever 21 more sophisticated computers. And Essentially, I think that the list 22 essentially I. is limited only by our imagination at the moment e 23 A great many of these topics of course, I think are 24 already partially underway. But there is a tremendous promising 25 frontier out there, and qualitative as well as quantitative ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202)289-2260 (800) FOR DEPO 27 1 changes in our capabilities and in the problems that we can 2 address effectively. 3 The initiative that we present to you represents a full 4 integration of component programs I in a whole number series of the l 5 Federal agencies These that have agencies been involved for some time I sir and 6 in high Fe performance computing and in the design, development, 7 utilization or { their own agency computer networks. And as was 8 noted earlier, the initiative that the President has brought 9 forward proposes a 30 percent increase in the funding in to 10 Fiscal Year 1992 in n support of these activities. 11 And ₦ think that I should la at the outset S SIT, state that 12 it is our goal in OSTP that between the next 4 and 5 years we will be doubled But it is 13 double the support for this activity but much more important 14 that we increase the speed, the memory capacity, of and the data 15 transmission capacity of our systems by factors between 100 16 and 1000. And this is certainly within technical reach. now 17 Now, I would like, if I might, sir, to take a moment just 18 to trace the history of this initiative, in which you have 19 been, yourself, very heavily involved. I think that it traces 20 its formative years back to the early 1980's and before, and 21 stems also & & in part from the recognition in a great many of our 22 agencies I that to satisfy their mission needs I they required 23 computing capacity far beyond anything that was then chenf 24 available. 25 And as the science and the technology in these agencies ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202)289-2260 (800) FOR DEPO 28 and 1 improved, it became rapidly obvious that the quantity of data 6" 2 the number of databases that were being developed made 3 absolutely mandatory the need and for more sophisticated data- 4 archiving, data-processing,, data retrieval. And one of the 5 points I would make in passing sir, I is that there are now 6 between 6,000 and 7,000 data bases in use around the world. 7 And & almost without exception, these are not inter I comparable 8 or readable, one by any other. And this is the beginning of a 9 Tower of Babel that we can ill afford. And that matter is one 10 that I think that we should work together to correct. 11 In 1982 I a FCCSET committee examined the status of 12 supercomputing in the United States in response, as you 13 suggested, to a request from the Congress and reviewed the 14 role of the Federal Government in this area of technology. 15 In 1985 the committee recommended Government action 16 necessary to sustain the technological superiority that we had 17 at that time, and to further the development and use of 18 supercomputers in this country. Subsequent planning resulted 19 in a who 1e series of workshops that were held in 1987 la and a a 20 set of reports that set forth the outlines of, research and 21 development strategy. The synthesis of all of this activity 22 appeared in the report entitled the "Federal High Performance 23 Computing Program" that was issued by my office in September 24 of of & 1989, as you noted. And the initiative that we bring 25 forward now in the 1992 budget is, in my view, a realization ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO and 29 1 of the goals that were spelled-out by you, by us, by a number of people back in those days as being thengoals, but now & they they were seen as 2 are 3 ^ targets that we fully believe The we can achieve. 4 What we have here is / an program that involves 8 partners S 5 And this is the Defense Advanced Research Project Agency, the 6 National Science Foundation, the Department of Energy, the Administration, the) 7 A National Aeronautics and Space He National Library of Medicine, 8 the Environmental Protection Agency, the National Institute of 9 Standards and Technology, and NOAA, the National Oceanic and 10 Atmospheric Administration, within Commerce. 11 And And the planning and implementation of the HPCC program, 12 I would submit to you, sin have been the result of a rather e 13 remarkable degree of cooperation and I must say Q 14 enthusiastic cooperation wherein I these and agencies have been 15 quite prepared to readjust, to I realign, redesign their in Y 16 programs and high-performance computing so that together these 17 made parts of a coordinated Federal whole. 18 And & I would like to pay tribute to the level of 19 cooperation that we have enjoyed in putting this program 20 together. There is a level of mutual trust, cooperation, and 21 synergism It that I believe & is remarkable / both inside and 22 outside of Government. And I would also have to say that the 23 success of this activity has depended in no small measure on 24 the input that we have received from the private sector. It 25 has been crucially important for us in developing the program ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202)289-2260 (800) FOR DEPO 30 and 1 that we have brought before you to have it calibrated I to have 2 it A tested against the real-world environment in which our 3 industries operate, because they, in the long-run, are the 4 ones that we look to to implement this program. 5 The program itself has four major components -- as you concerns 6 know very well, sir. The first has to do with high- a 7 performance computing systems. And we use that phrase 8 advisedly because we wanted to include not just 9 supercomputers of but also high-performance computers on all 10 levels because the synergism there is important and should 11 not be forgotten. 12 Secondly, we talk about advanced software and algorithms, 13 because without that, all the hardware in the word is 14 essentially useless. 15 Thirdly, we have the National Research and Education 16 Network, the information superhighway to which you referred. 17 Because without that, very few people have access to either of 18 the first two components. 19 And finally, we have a basic research and human resource 20 component. Because without that, we do not maintain our 21 frontier status, nor do we have the trained personnel who will 22 not only move these frontiers forward. But I believe, 23 equally important an d often forgotten and are the technical people 24 who, in fact, operate these systems A make sure that they do 25 what they were designed to do. ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 31 1 We have, in developing this program, set what I believe 2 are ambitious goals -- ambitious but realistic goals. As I 3 just mentioned, we seek a thousand fold improvement in useful 4 computing capability. That takes us to a trillion operations 5 per second. And & the focus here will be on developing the 6 generic technologies that will prove valuable not just in this 7 sector I but in many different industrial sectors. 8 Where appropriate, we feel it very important that the d 9 development be performed on a cost-share A basis with industry. 10 Because we want to involve industry as deeply as we can from 11 the very beginning of this activity. 12 In software development, I think we clearly have a major 13 challenge. Because our software here in this country -- as 14 indeed I in every other country -- is now lagging behind the 15 development of hardware. And & most important, as we have 16 discussed many times before, if we are to have the rapid 17 expansion of the use of our new capacities, it is essential 18 that we develop software that is user friendly. And And of 19 course, as we both know, the high performance software of 20 today is not user friendly by the wildest stretch of the 21 imagination. 22 The National Research and Education Network is going to 23 dramatically expand and enhance the capability of the existing inter connected computer network that L we refer to as INTERNET. red 24 25 The overall goal here is to achieve a 100-fold increase in ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 32 1 communication speed. We want to take this up to the level of 2 gigabytes per second. 3 In addition, as you have noted many times, one of our 4 major goals here is to develop a vastly greater number of on to 5 and off ramps on this information superhighway. There are too 6 many isolated institutions and areas in our country for a e 7 modern civilization e 8 And I hope that in the not-too-distant future our public 9 will look on this network as commonplace and as little-to-be- That 10 feared as the telephone system. This is a goal that I think e 11 is quite within our reach. it could have 12 If we have such a network, I think we have the 13 possibility of a catalytic effect on just about every 14 component of our society Lo on our industries, both small and : 15 large and I appreciate the reference to small industries, 16 because I believe that perhaps there is one of the vory & major this could be 17 potential impacts; universities and research organizations; and 18 bus perhaps even more important the elementary and secondary 19 schools of the Nation where the real deficiencies in our 20 educational system are most apparent. 21 Finally, I think that no plan is better than its 22 execution. And & the execution of this initiative is going to 23 depend very critically on the synergy that has been developed 24 among the agencies that are participating in it. What we have 25 tried to do is to develop the plans SO that each agency does ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 33 1 what it does best I and does it in a coherent way to amplify 2 the effects of its sister agencies in this overall natural program. 3 4 constituencies l and has historical strengths. I think that / as And / Each of the agencies, as you well know, has national 5 has been noted before, that this particular I initiative and has the 6 promise of very high payback in economic terms in social 7 terms and & as I. indicated at the outset I find it very 8 difficult to think of any other initiative that has the 9 potential of a higher payback to the American taxpayer. 10 The high-performance end of the computer market l as we we 11 all know, is relatively is small But its influence far 12 transcends it size. That is where the leading- edge technology the leading-edge applications are developed, and and ies and 13 1110 a 14 Federal investment in leading- edge computer technology will 15 speed the growth of the overall computer market and can 16 catalyze I I believe, investments on the part of U.S. industry. 17 But I would again come back to the matter of synergy 18 mentioned earlier. I would not want this to be thought of as 19 only a supercomputer initiative. It is much more important 20 than that. Supercomputers play a role, and a very important 21 one, but by no means the only role. 22 I think also that the initiative that we bring before you also 23 today, has the potential to be a major contributor to meeting a 24 number of other very important national needs -- national 25 security, health, transportation, education, energy, ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202)289-2260 (800) FOR DEPO 34 1 environment -- all of these are areas where the availability and 2 of a new generation of high-performance information is handling f 3 information transfer will be essential. And the dependence, I 4 believe will, in fact, only grow. 5 If we are to realize the full potential of this 6 initiative, it is not enough that it reach its technology e 7 goals. It is equally important I in my view, that that 8 technology be deployed by the private sector in a timely 9 fashion. And the continued development and use of Government- 10 funded, high-performance computing and communication 11 prototypes can certainly have an important, positive impact on 12 the commercialization of these technologies. And & in fact, we 13 can make available to a great many institutions in this 14 country that which cannot themselves justify through the hardware investment 15 the power that will be available an the proposed networks. however, l 16 This diffusion, I would submit, Mr. Chairman, is not 17 possible by Federal action alone. The administration's 18 initiative will serve the Nation best as a catalyst for 19 private action. 20 Nott Some analysts have suggested that the initiative can spur 21 several hundred billion dollars of GNP growth. If so, it will 22 be because American companies, both small and large, have been 23 able to deploy these technologies in the production of high- 24 quality goods and services. 25 I think also that as come have predicted, that the ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 35 1 National Research and Education Network will lead to the 2 establishment -- and I hope quickly -- of a truly national, 3 high-speed network that really does connect essentially every and 4 home A office I in the Nation. And if that happens it will be 5 because the private investments that make it possible have 6 been stimulated by the initiatives taken here. 7 Now the legislative proposals pending before the 8 Congress, I would suggest, Senator, perhaps do not fully 9 recognize the comprehensive inter I agency effort that has been 10 achieved through the years of collaboration that have led to 11 this particular activity. NOH 12 When I testified last year T before the corresponding now I 13 hearing, I noted one concern and I would simply only reiterate 14 that concerno si six And that is that in a field of technology 15 that is moving as rapidly as is the case here, as you 16 illustrated with the automotive analogy, I am somewhat 17 concerned that by freezing the program in legislation we may 18 have given up some flexibility and that we may want in order to be 19 able to adjust this program,A the agency I participation, on an 20 ongoing basis. 21 I would emphasize that the FCCSET activity should still 22 be viewed as partly experimental. This is the first year that 23 we have really had it working as I had hoped it might. And we 24 will certainly move forward next year to do a better job in 25 these areas than we have been able to this year. ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 36 1 That was one concern. The other concern I have, sir, 2 reflects a fact that you mentioned. And that is e e 3 that we have worked long and hard to bring about this total 4 integration of the agency programs. And my concern, which I 5 believe is one that you I share, is that in the 6 subsequent actions hero in of the Congress I it would be a great 7 pity of that coordination and integration were not carried 8 forward as the various pieces and players in the program 9 present their programs to your sister subcommittees. 10 And I would very much look forward to working with you, 11 sir, and ask for your assistance in making sure that we do 12 retain the coordination that is now a hallmark of this 13 activity. 14 Now I think, sir, that I woul & d, at that point I simply 15 say that I would conclude my prepared remarks. The full 16 testimony will be in the record and I would P welcome your 17 questions sirl sir 18 [The prepared statement of Dr. Bromley follows:] 19 20 21 22 23 24 25 ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 37 1 Senator Gore: Well, thank you very much. I know my 2 colleagues will have questions also. 3 Let me begin by thanking you very much for your 4 statement. I thought it was very well put. And I, too, look 5 forward to the private investment which will be stimulated by 6 this initiative and the passage of this bill. In fact, I know 7 you are aware of the Merit proposal, with IBM and MCI and 8 MErit, announcing a massive not-for-profit venture designed 9 solely to quickly expand and proliferate the network as soon 10 as the Federal backbone network is in place. 11 I want to see a day when a school child in Tennessee can 12 come home after class and sit down, and instead of playing 13 Nintendo, use something that looks like a Nintendo apparatus 14 and plug into the Library of Congress; and read just not 15 words, but look at pictures and moving graphics presented 16 artfully and imaginatively in a way that captures and holds 17 that child's attention; responds to the child's curiosity so 18 the child can navigate through an ocean of information 19 according to what he or she wishes to explore at the moment. 20 We know how to do that. The technologies are here, 21 available, today. We ought to be empowering -- to use the 22 buzz-word of the day -- not just one or two or a few 23 individuals, but we ought to be empowering the whole country 24 to make better use of the information that is out there. 25 I also agreed with your emphasis on the level of ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 38 1 cooperation that is so necessary to making this successful. 2 And, I could not agree more when you say do not freeze it. We 3 are interested in passing this and getting it moving, 4 authorizing it for the number of years necessary to complete 5 it -- sending those clear signals to the private sector that 6 it is there, it is going to be there, and it will be 7 completed, not just started, and then reevaluating every year 8 the ways to improve it and make it even better. 9 I also would like to identify with your statements about 10 the coordination of these efforts. 11 Let me start with my questions on that one. We are 12 agreed that the goals of the administration's initiative and 13 S. 272 are very, very similar, and that S. 272 is consistent 14 with what the administration is proposing. 15 Last year, as I had mentioned, the precursor, S. 1067, 16 passed the Senate unanimously. And then its passage was 17 delayed by a difference of opinion with the House, and among 18 different committees in the Senate. 19 In particular, the Department of Energy was pushing to 20 have DOE coordinate the whole program, and run the National 21 Research and Education Network. That was at-odds with the 22 administration's plan, and at-odds with the legislation. 23 I know that DOE labs like Oak Ridge and Los Alamos have a 24 critical role to play in this initiative. But I also know 25 that the best program is one that uses the strengths and ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 39 1 resources of all relevant Federal agencies. 2 Luckily, many busy people within the administration have 3 spent the last couple of years designing just such a balanced 4 program as we have. However, last year, personnel from DOE 5 and elsewhere actively lobbied against the OSTP approach and 6 bad-mouthed some of the other agencies in the program, saying 7 that they were unable to accomplish the missions that you laid 8 out for them. 9 Should we expect to see more of such bureaucratic in- 10 fighting? Or is everybody on your team on-board this year? 11 Dr. Bromley: I would devoutly hope not, Mr. Chairman. I 12 have spoken with -- 13 Senator Gore: You would hope we would not see more in- 14 fighting. 15 Dr. Bromley: No more in-fighting. 16 [Laughter. ] 17 Senator Gore: Okay, I just wanted to clarify that. 18 Dr. Bromley: That is important. 19 I have spoken with the senior officials of both the 20 Department of Energy and the Department of Defense during the 21 formulation of this plan I and as we went through the FCCSET 22 process. And I have been assured by them that they are full 23 players and full participants in the program that we have 24 presented to you. We recognize that many of the agencies will Q 25 have need for their own independent, mission-related ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 40 1 activities. But we agree completely that in order for this to 2 be a truly national program, it requires management as a 3 national program And we have identified NSF as the l 4 appropriate agency for that. I believe that to the best of my 5 knowledge that everyone who is represented in the report that 6 you have before you is committed to moving forward with the that report 7 program and plan laid out in there. 8 Senator Gore: Well, that is good news. And I welcome 9 that assessment. And, along with you, I devoutly hope that it 10 is the case. And I believe that it is. 11 Now, I appreciate your kind words and your statement of 12 support for the principles involved in the bill. And I fully 13 understand why any administration, given its druthers, would 14 like to have the money but no requirements on how the money is 15 spent. 16 And I take it that you -- your basic position is you like 17 the legislation. But your basic position is that you would 18 like to recommend that the Congress appropriate all of the 19 money that you have requested, but to give you no formal 20 instructions on how to spend it. Is that basically -- 21 Dr. Bromley: I would phrase it somewhat differently, 22 sir. 23 [Laughter.] 24 Dr. Bromley: I would perhaps put it that we look forward 25 to working with you. ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 41 1 [Laughter. ] 2 Senator Gore: Well, we look forward to working with you 3 too, Doctor. 4 Does the administration's initiative represent more than 5 just a 1-year commitment? Do you have a commitment from OMB 6 for the entire 5-year program or just for the first year? 7 Dr. Bromley: We have held, as you understand, detailed 8 discussions with not only the agencies but with OMB. And the 9 program that is before you in the 1992 budget is the one that 10 we agreed on with OMB based on a 5-year plan. But you also 11 understand, sir, that the OMB, in its normal activities, so so 12 focuses on the particular year And we will go back and we 13 will have to make our case again in the following year. But I 14 am convinced from our activities this year that there is P a 15 full recognition that what we are talking about here is an 16 important initiative It Li is a presidential initiative. And 17 that we are talking about a 5-year program to achieve what is 18 involved. 19 So, although I cannot guarantee to you anything about the 20 years beyond this, my best judgment, and that of my 21 colleagues, is that this administration is fully committed to & 22 this as a very important and a presidential initiative. 23 Senator Gore: Of course, things change, and budgets 24 change, pressures build, lips get red, circumstances alter. 25 But formally, you have a commitment for 1 year? ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 42 1 Dr. Bromley: That is correct, sir. 2 Senator Gore: And you look forward to working with OMB 3 for the other 4 years -- just as you look forward to working 4 with us for the 5 years. 5 Dr. Bromley: I look forward to working with anyone who 6 will push this initiative forward, sir. 7 Senator Gore: All right. 8 Dr. Bromley: But let me say that I have a very 9 reasonable degree of confidence that OMB understands fully 10 that this objective is a very important one, and that it is 11 certainly the intent of everyone involved at the present time 12 to move this forward expeditiously. 13 Senator Gore: Well, I think that is a very important 14 signal to send out. I am making the point, of course, that 15 the legislation is needed, even though any administration 16 would like to have all the money for everything, each year, 17 simply appropriated and not authorized. 18 But I do not want my efforts to make that point obscure 19 the very clear signal that you are sending to the private 20 sector, to all of the agencies involved. This is going to 21 happen. This is going to happen. And everybody needs to get 22 with the program and make certain that it does happen. 23 I just have a couple more questions before deferring to 24 my colleagues. I want to explore the relationship between 25 this and other OSTP initiatives. ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. . SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 43 1 Last month when the budget was released, you presented 2 two other multi-agency initiatives -- a new math and science 3 education initiative, and -- and, as it has for the past 2 4 years, the budget also included a coordinated and integrated 5 U.S. Global Change Research program involving nine different 6 agencies. 7 Is it fair to say, as I regularly do, that this program 8 will contribute greatly to these other initiatives, enhancing 9 our Nation's ability to pursue them productively. I do not 10 want to lead you on. I suspect you agree. But since these 11 other two initiatives are yours, I would like you to flesh 12 that out just a little. 13 Dr. Bromley: Well, first of all, let me say that it is I 14 imminently e fair to make that statement. That, in fact, not a 15 only will the high-performance computing initiatives initiative have a not only 16 very important impact on ^ the other two ^ that you have just 17 mentioned, but, indeed, on a great many other of the 18 activities we have under way. 19 The most obvious connection has to do with the global 20 change arena where, l as you noted earlier, Bir, L we have a 21 Mission to Planet Earth that I consider very important under 22 consideration. And as was noted earlier this afternoon, once 23 we have the EOS A platform up, which will be a unique part of will 24 that program, because a it would allow observation of individual 25 points on the earth's surface through 14 to 15 instruments e ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 44 1 simultaneously. And & that is critically important Because 2 from those simultaneous data one can then I extract vastly more 3 than one can from just 14 or 15 individual sensors flying 4 independently. Because there is no air column correction, no 5 cloud correction, nothing of that sort. 6 However, But as was indicated, the data flow from that particular 7 remote-sensing complement alone w will send us the equivalent 8 of the Library of Congress in less than 5 days in fact. And e 9 So, unless we have the kind of speed, capacity and information 10 transfer capability that we are talking about in the high computing and communications 11 performance ^ initiative, we simply cannot cope with the flood 12 of information that will be coming to us from the sensors that that & 13 we we have in action. 14 And perhaps even more importantly I think l is a point 15 that you touched on earlier. And that is that the human brain 16 is substantially limited with what it can do with wi in whacking through 17 great stacks of computer print-out. But on the other hand, it is 18 almost miraculously able to form hypotheses and sense patterns 19 in those same data if presented in a ^ graphical fashion. That the 20 think, is probably the largest qualitative difference that 21 high-performance computing can make in any area. Because Luss it 22 will allow us to take this flood of data and actually do 23 something with it -- make decisions, understand phenomena & that 24 would simply be beyond us otherwise) so in that area the 25 question has a very obvious answer. ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202)289-2260 (800) FOR DEPO 45 1 In education and mathematics, I think that the impact 2 probably is going to be greater in the long run But than even on 3 global change. But This it is not yet as obvious. ^ The fact that 4 impresses me enormously in this area is that with a single 5 fiber optic going into a classroom, every student in that 6 classroom can have self-paced, individualized instruction in 7 any subject -- repetition where repetition is necessary, 8 positive reinforcement where that has been earned and I 9 cannot think of anything that the will improve the quality of our 10 education, particularly at elementary and secondary (more school levels, where our greatest weakness now lies / X than the 11 12 introduction of this kind of technology into the education 13 field. solt V in the long-run, I think the impact may be even 14 15 greater there, although it will take a little longer to bring 16- it into place. 17 Senator Gore: I welcome that response. Just briefly, I 18 heard a presentation in one of the early hearings on this, I 19 do not know how many years ago, where someone said, and I have 20 repeated it often since then, that if one analyzes the human 21 brain in computer terms, you could say that we have a low bit 22 rate, but very high resolution. 23 The telephone company decided years ago that seven 24 numbers were the most we could remember, then they added 25 three. And yet, when we see a trillion bits of data arrayed ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 46 1 in a mosaic pattern where each has a meaning in context 2 related to all of the others, we can comprehend them all 3 almost instantaneously. 4 What the new supercomputers allow us to do is to 5 configure data in shapes and patterns over time, which enable 6 us to absorb very large quantities of it conveniently and 7 quickly. Secondly, they allow us to search through vast 8 oceans of data and instantly retrieve those particular bits 9 which are necessary to make up the particular pattern that we 10 are looking for in order to understand the problem that we are 11 trying to solve. 12 In any event, I will come to questions later. Let me 13 recognize Senator Pressler. 14 15 16 17 18 19 20 21 22 23 24 25 ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202)289-2260 (800) FOR DEPO 47 1 STATEMENT OF HON. LARRY PRESSLER, U.S. SENATOR FROM SOUTH 2 DAKOTA 3 Senator Pressler: Thank you very much. 4 Dr. Bromley, I am fascinated with page 26 and 27 of the 5 report which show all the agencies that have to be coordinated 6 that are involved in the program, and the agency 7 responsibilities. It is amazing to me how many of the 8 agencies -- and I am sure they are all filled with highly 9 trained scientists and highly trained people. 10 When I was in the Army in Vietnam, I was at one point on 11 detail to -- we called it ARPA then; it is now DARPA. But I 12 know the difficulty of getting highly trained people to work 13 together. You would probably have the classic job in public 14 administration or an administration in coordinating all these 15 people and getting them to work together. What is the number 16 one problem in keeping all these agencies working together on 17 this program? 18 Dr. Bromley: I would have to say, sir, that probably the problem 19 number one program is information transfer. 20 Senator Pressler: The number one problem? 21 Dr. Bromley: Yes, because there are, as you say, a very 22 large number of people involved here, and they will work to 23 maximum effectiveness if we can be sure that everyone 24 understands what everyone else is doing, and that they are 25 really part of a coherent program. Jue & MM are devoting a very ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 48 1 large amount of our activity to trying to make sure that this 2 is the case. 3 And And I must say that I want to pay tribute here Le to the 4 people who have been involved, the people who prepared this 5 document, because they have managed what is really a They have brought 6 remarkable feat in in bringing about a fusion of what started out 7 as separate programs in each of the agencies you see listed on 8 these pages, sir. And Ld these folk have spent many long hours 9 sitting and looking at each program and asking how does & it fit⁵ 10 as part of a national program, and then adjusting wherever the 11 overlap, the duplication, the gaps were to make it actually 12 fit. And l so we already have developed a level of personal 13 communication among the members of the community involved in all these agencies that I think l will service extraordinarily serve vs 14 15 well in the years ahead. 16 Senator Pressler: Now, as I understand the general 17 difference -- and I am not advocating either one here 18 necessarily, because last year in one of my statements I 19 called on the administration to come forward with a plan. 20 As I understand, the basic difference is that these 21 various agencies, we would depend on you to coordinate them as 22 you saw fit; whereas, you feel the Gore-Pressler-Kasten, et 23 cetera, bill would codify too much the relationships on the 24 HPCC program regarding all these agencies. Is that a general 25 statement? ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 49 1 Dr. Bromley: I would have to modify that just a little 2 if I might, sir, because we in OSTP do not force anybody in 3 any agency to do anything but we & do when we are successful, 4 we persuade them that they have the opportunity to become part of A 5 a much larger entity, the a national program. And so to that 6 extent, everyone here already is involved. The responsibility 7 is the agency's, but if the coordination fails, then the blame 8 is ultimately mine. 9 Senator Pressler: But basically, these are all 10 administration-appointed -- well, they are all -- the heads of 11 all these agencies are appointed by the President, usually 12 with the advice and consent of the Senate. So therefore, the 13 White House could order them to do something. 14 Dr. Bromley: In principle, yes. 15 [Laughter.] 16 Dr. Bromley: There is a famous quote that springs to 17 mind, sir. I can summon spirits from the vast deep, but will 18 they come. 19 [Laughter.] 20 Senator Pressler: Well, I know back from my days as a 21 second lieutenant in the Army listening in on meetings, to get 22 the attention of an assistant secretary was hard on some of 23 these things. If you could do that, you could accomplish 24 something, but there is great competition. 25 Dr. Bromley: May I add one word of clarification, sir? ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 50 1 One of the important things that perhaps I should have 2 emphasized is that this has been a concern within the and 3 administration within my office. And in order III to address that 4 concern, during this past year I have restructured the FCCSET 5 committee so that its members now comprise the cabinet 6 secretaries, or deputy secretaries, and the heads of the 7 independent agencies that are responsible for all these As arcsult, Q 8 activities so that that ^ once the FCCSET group makes the decision, 9 that then is 9 decision that will stick from there on in 10 because it has been made by the senior officer of the agency 11 involved, and that is a key part. FCCSET 12 Although that group is not involved in the actual 13 construction of the initiative, and it receives the initiative, 14 approves the initiative, gives the initiative its blessing 15 before it moves forward as part of the budget process. 16 Senator Pressler: I am glad to hear that, because that 17 is really key. As a point of view of public administration, 18 many of these activities require the input of highly trained 19 people, and frequently the very top people administratively in 20 these agencies are so preoccupied with the war in the Gulf or 21 with other matters that what happens, as a practical matter, 22 is that you start having these decisions made -- you have 23 assistant secretaries battling it out, so to speak, for turf. 24 And I am glad to hear that the administration has gotten 25 the very top people involved on this. I think that is key, ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202)289-2260 (800) FOR DEPO 51 1 and I hope you are able to continue that level of interest. 2 And I just point that out because I am fascinated with all 3 these agencies here. This chart is an amazing Washington 4 chart, and to make all this work together with these highly 5 skilled people who perhaps do not submit to traditional 6 administrative discipline quite as much as some others, you 7 have a big job. 8 Now, let me ask a couple of questions here. Do you 9 foresee the development, at some point, of a system of user 10 fees so the supercomputer network will pay for itself 11 partially? And how would that system be structured SO that 12 user fees do not deny access to users with limited resources, 13 such as small schools and individual researchers? 14 Dr. Bromley: Wel of in the long term, sir, as I touched 15 on just briefly earlier, I look on the National Research and 16 Education Network as a pilot if you like, a precursor 1 for 17 what, as fast as we can, will become a national service that 18 is provided by the private sector as a utility service just 19 like the telephone. 20 And under those circumstances it would seem to me that we 21 would function much as we do at the present time namely that e 22 small schools, for example, or people who wanted access to 23 this utility would make application as part of their normal 24 process of receiving support for their research activities to 25 an appropriate agency and that just as we now provide part of ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 52 the 1 grants and contracts to support ^ use of telephones, copying 2 machines, and whatnot, I think L we would also be more than 3 happy to include in that list of necessary tools if you will, e 4 the charges that might be levied by those public utilities for 5 access to the computer net. 6 Senator Pressler: Now, as you know, we have EPSCOR 7 legislation to ensure that the smaller institutions are not 8 unfairly left out when Federal research grants are made. What 9 steps can be taken to ensure that the computer research and 10 development called for here in S. 272 and your proposal will 11 include small institutions? Or to put it more specifically, 12 how can we be sure an EPSCOR professor working on deep 13 drilling at the South Dakota School of Mines will be included? 14 Dr. Bromley: Well & I think perhaps the first and most 15 important answer that I can give to that is simply to tell you 16 that this is the President's wish transmitted directly to me 17 that nat if this is going to be his initiative he wants it to be 18 broadly available to institutions large and small, both in the 19 educational and in the industrial sector. And so that I look that can 20 on as an instruction, and we will do everything we possibly A to 21 make sure it happens. Frankly, I do not think that we will 22 achieve anything like the potential of this system unless we 23 do just what you suggest. 24 Senator Pressler: Thank you. 25 I have some additional questions for the record, Mr. ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 53 1 Chairman. 2 Senator Gore: Thank you very much, Senator Pressler. 3 Senator Robb. 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 54 1 STATEMENT OF HON. CHARLES S. ROBB, U.S. SENATOR FROM 2 VIRGINIA 3 Senator Robb: Thank you, Mr. Chairman. I am pleased to 4 join you and Senator Pressler and other of our colleagues on 5 this committee again as a cosponsor of this legislation, and I 6 am delighted to have a chance to be with Dr. Bromley again. 7 We have coconspired in previous incarnations on other 8 scientific projects, and it is always a pleasure to work with 9 him. 10 I did have a very brief statement that you may have 11 already included in the record. In any event, I would ask 12 unanimous consent to include it without delivering it just on 13 this particular topic. 14 Senator Gore: Without objection. 15 [The prepared statement of Senator Robb follows:] 16 [COMMITTEE INSERT] 17 18 19 20 21 22 23 24 25 ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 55 1 Senator Robb: I just tell you, Mr. Chairman, I was 2 struck when you made mention of the fact that you would like 3 to find a way to be able to let a child at home tune in to the 4 Library of Congress. I could not help but think last night 5 when I came home, I have a child who was trying desperately to 6 do just that. She had just lost about an hour's worth of work 7 that she had done on her own PC. It had gobbled it up, and 8 she was looking for some recourse to higher authority, which I 9 was unable to provide at the moment. 10 And then when I heard Dr. Bromley say something to the 11 effect that the entire Library of Congress could be 12 transmitted in 5 days, it occurred to me that if any of these 13 children had a printer and some sort of an arrangement to 14 simply put it on automatic pilot, that we could incur costs 15 that are beyond the ability of even Congress to take care of. 16 [Laughter. ] 17 Senator Robb: I have a couple of questions that are more 18 technical in nature that I would like to submit for the 19 record, but I would like to ask Dr. Bromley just in a very 20 sense, since this is presented again in a cost-conscious 21 environment, and several of the lighter remarks and exchanges 22 took place with respect to future funding and the coincidence 23 of intent, and what have you, with respect to others here and 24 at OMB and elsewhere. 25 But I wonder if you could just talk a little bit about ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 56 1 ways that those who believe that this would be a very good 2 investment for our country, for our individual States, for our 3 industrial sector, for others, how we could attempt to provide 4 what would be traditional cost benefit analysis. How could be 5 grapple with the basic question of justifying the kinds of 6 expenditures that we are proposing here, or that are implicit 7 here, and in the long term to realize the full potential that 8 I believe and others believe this has in terms of the costs, 9 so that we could make some comparison to the costs that other 10 competing interests for the finite Government dollar might be 11 justified or evaluated. 12 Dr. Bromley: That, Senator, is an excellent question, 13 and I wish I had a better answer for you than I am going to we 14 give you. We have back in 1989, for example do entered into a from 15 contract with a group an Los Alamos to try to, just what you 16 have asked namely to try to come up with an economic cost 17 benefit analysis of this initiative 1 and it has been under way 18 since that time. 19 Frankly, I would have to say that I do not put too much 20 credence in the numbers that we have at the moment because e 21 the sort of range that people are talking about is that if we 22 were to implement the initiative that we have presented to 23 you, that the payback would be somewhere between $170 billion 24 and $500 billion over a period of the next 10 years. The 25 range in itself tells you a lot about how good the actual ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 57 1 calculation is; and I frankly, sir, I think that it is a little 2 premature. 3 Senator Gore: Even so, Dr. Bromley, it might be good to 4 have that on the record. 5 Dr. Bromley: It is not bad. 6 [Laughter.] 7 Senator Gore: It might be good to look at. I do not 8 want to interrupt Senator Robb's question, but those are 9 intriguing numbers. I mean, we can deal with a range. 10 [Laughter. ] 11 Senator Robb: That is right. Regrettably, very little 12 justification is frequently required around this institution. 13 If you want to believe, you do believe. 14 Dr. Bromley: Let me, then, sirs, tell you that the only 15 numbers Life concrete numbers that I have available to me at the 16 moment are those that came from a Gardner study that, in fact, in the 17 was requested of back & a number of years ago/i) and and it is that 18 range $170 billion to $500 billion over 10 years. And that 19 is a very impressive payback. But I also would caution you, since are 20 sir, that like you being familiar with a great many cost 1 21 benefit analyses, that there are many potential pitfalls Not 22 And I give it to you only as an indication in support of 23 my personal belief that the payoff here is probably one of the 24 best in terms of an investment viewpoint of anything that I H 25 can conceive of us doing. We just recently, for example, had ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 58 1 the study of Professor Edwin Mansfield of the University of 2 Pennsylvania, which focused on what I. is the rate of return on 3 Federal investment in academic research and he came out with 4 a figure of 28 percent. 5 Not Now that is a marvelous figure, because it was created 6 by a first-rank economist. And we scientists love figures of 7 that kind because our economist friends cannot argue with us 8 about them. But I would submit, sir, that if that is the 9 return across the board on Federal investment S in fundamental 10 research, then I would be prepared to wager rather heavily 11 that the return on this initiative would be higher by 12 substantial factors. 13 Senator Robb: Dr. Bromley, I cannot tell you that that 14 is encouraging news. I would only warn you that having given 15 a figure like that, you may find that in an off-set within the 16 current caps on budget, somebody will pencil in this program 17 and then spend those savings for some other program which may 18 be equally worthy. 19 [Laughter.] 20 Senator Robb: But I do appreciate it, and I think that 21 the fact that there is, at the very least, a very substantial 22 benefit in hard dollars that could saved ought to be cranked 23 into the equation sometimes. And there are occasions when we 24 spend money to get even more money back. It does not happen 25 often in government, but there are occasions when it does, and ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 59 1 in the private sector as well. 2 As I said, the other questions are really more technical, 3 and I will submit those in writing. But again, I thank you. 4 And Mr. Chairman, I thank you for not only again 5 sponsoring the legislation, but for another hearing, and 6 hopefully more understanding of the importance of this 7 particular possibility on the horizon. 8 Senator Gore: Well, Senator Robb, thank you, and thank 9 you for your early and vigorous support of the initiative in 10 the last Congress as well. And as one of the former Governors 11 who took a leading role in stimulating high-tech research and 12 development in your State, I do not know if that is the reason 13 why you were such an early cosponsor of this, but I appreciate 14 your -- 15 Senator Robb: Mr. Chairman, if I could, let me just say 16 that had the economy not soured a little bit just south of the 17 Potomac, I had some proposals that we were going to use our 18 Center of Innovative Technology to house a supercomputer, but 19 got into some of the costs involved and recognized the fact 20 that the economy, in least in terms of available expendable 21 dollars, was going sour too rapidly to push it very hard. 22 But I hope when we recover, that -- I actually made a 23 presentation on this bill and the purposes of it to an 24 appropriations subcommittee that was taking a look at it, and 25 I hope at some point that we may be able to get back in as a ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 60 1 part of this overall highway that you hope to create. 2 Senator Gore: Well, thank you again for your early and 3 vigorous support. I wanted to follow up on the figures that 4 you had Dr. Bromley comment on, and I accept the caveats on 5 why models like these are not reliable. But the numbers are 6 quite significant, and they include estimates not only related 7 to GNP, but also reductions in the deficit. So your thing 8 about the offset there I know was in jest, partly in jest 9 anyway, but this -- 10 Senator Robb: No, Mr. Chairman, I have learned in this 11 institution you do not jest about things like that. 12 [Laughter.] 13 Senator Gore: But this does project, as unreliable as 14 such figures are, very significant reductions in the Federal 15 budget deficit because of this. To use a more reliable way of 16 estimating its benefits, you could say that the total 17 expenditure on this program represents about 1 percent of the 18 Federal R&D budget. 19 If, therefore, the improvements in the productivity of 20 even a tiny fraction of the other 99 percent of the Federal 21 R&D budget results from this, and you know it will, then we 22 are ahead right there in terms of value saved for the 23 taxpayers before you even consider the benefits for the 24 economy. 25 It took only a tiny leap of faith to embolden those who ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 61 1 created the interstate highway program to allow them to assume 2 that it would be used when it was built. They could really 3 see that it would be used, and it was; it has been. And it 4 has vastly improved our economy. It takes an even smaller 5 leap of faith to assume that when this network is built it 6 will be used. 7 The utilization rate for the network which now exists, 8 one one-thousandth of this capacity, is growing by -- I cannot 9 remember the exact figures, but they are measured not 10 annually, but monthly. The increase is just phenomenal. And 11 so it may be hard to put reliance in specific numbers, but it 12 is easy indeed to assume that it is going to make a tremendous 13 difference for our economy. 14 I have just a few other questions, and then we will let 15 you move on. Your initiative, Dr. Bromley, places a great 16 deal of emphasis on massively parallel supercomputers. 17 Clearly massive parallelism is the only way we are going to 18 soon achieve the 1,000-fold improvements in computing power 19 needed to solve many of the grand challenges. Yet, there is 20 still a good deal of good science being done on so-called 21 conventional supercomputers. 22 And I am wondering, will the administration's initiative 23 focus solely on massively paralleled supercomputers, or will 24 funding also be provided for purchasing supercomputers that 25 are now on the market and developing new and improved software ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 62 1 for them. 2 Dr. Bromley: It is certainly, Mr. Chairman, not our 3 intention to in any way ^ suggest that the only way to go is the 4 massively parallel route. The important thing is that given a 5 problem in either science or technology, it usually turns out 6 that either the parallel or the mainframe has major advantages 7 in the solution of that particular problem,) and so & that the 8 goal here is to have both facilities on the network and so 9 that you have essentially a transparent system. NOH An 10 So the individual working in Senator Pressler's would & ve 11 laboratory, simply uses his work station and has access through 12 the network to whichever is most suitable to the problem that 13 happens to be addressed at the time. We in no way are 14 suggesting that we want to eliminate one of these approaches in favor of the other. We do, however, simply note here l that l 15 16 if we are to reach the increase in speed in the time that we 17 project here, that l the only way to do it is through the 18 scaléable massively parallel architectures. 19 Senator Gore: Well, I am a big fan of massive 20 parallelism as you know, because you and I have talked about 21 it. But I also recognize the kind of balance in the program 22 that you have just indicated with your response and how 23 important that is. 24 One other question, and it involves education. In the 25 administration's proposal, the primary justification for the ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 63 1 initiative is research and development. Frankly, I was a 2 little bit surprised that more attention was not given to the 3 educational applications of this technology. There are 4 hundreds if not thousands of ways that a national network in 5 supercomputing can help students in colleges, high schools, 6 junior highs, and even elementary schools. 7 For instance, in January I attended the annual meeting of 8 the American Library Association in Chicago, and saw a 9 demonstration of how librarians are using the NSFNET to 10 provide students with information from databases all over the 11 United States. Yet, in the administration's proposal, there 12 is almost no mention of the role that libraries will play in 13 providing information resources to other users of the NREN. 14 Was this an unintentional oversight, or does the 15 administration intend to focus almost exclusively on research? 16 And how would you personally like to see existing networks and 17 the NREN used to improve the American education system? 18 Dr. Bromley: I am in complete agreement with you, 19 Senator. And & the fact that this does not appear in this 20 particular report ssue reflects and back l on my earlier comment that this 21 is the first year -71 the first attempt built was done under immense 22 time pressure, and it was essentially building on activities 23 in which the agencies are currently involved. 24 No one in the group that developed this initiative 25 questions for a moment the tremendous importance that it will ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. . SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 64 1 have for education. And I think I can promise you that when 2 you see the report of this group next year, you will see a 3 much greater expansion of areas like are not educations are 4 included this year because, as I say, we were working to get 5 this report to you, and there was a natural tendency to build 6 on those familiar areas in which the agencies are currently 7 involved. 8 Senator Gore: So you would not object if in our 9 authorization this year we included the educational 10 components? 11 Dr. Bromley: I believe that education is going to be a 12 very important part of this initiative, sir. 13 Senator Robb: The tough questions have got to stop. 14 Senator Gore: Yes, I know, I know. I am going to relent 15 soon, Doctor. 16 [Laughter.] 17 Senator Gore: All right. Well, I have other questions 18 like Senator Robb and Senator Pressler which I will supply for 19 the record. Is it okay if we move on now? We have had you 20 here a long time, Dr. Bromley, Dr. Wong. You have our 21 admiration for the job you are doing, and you may get tired of 22 me bragging on the good work you do in so many areas. Maybe 23 it makes up for the few -- 24 Dr. Bromley: I never tire of that, Senator. 25 [Laughter.] ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. . SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 65 1 Senator Gore: Maybe it will make up for the few where we 2 disagree. But thank you so much for your leadership in this 3 area in particular. It has been a joy to work with you 4 publicly and privately, to get to know you better and, in the 5 process, help mutually to move this matter along. And we look 6 forward to continuing that working relationship. 7 Dr. Bromley: I would welcome the opportunity to work 8 with you, sir. 9 Senator Gore: Thank you very much, and thank you, Dr. 10 Wong. We will now call our panel. We have five more 11 witnesses in a panel, and then we will conclude the hearing. 12 First, Dr. Donald Langenberg, Chancellor of the 13 University of Maryland System at Adelphi, Maryland; Dr. Melvin 14 Kalos, Director of the Cornell National Supercomputer Facility 15 in Ithaca; Mr. Tracey Gray, Vice President of Marketing for 16 Government Systems with U.S. Sprint, based in Herndon, 17 Virginia; Dr. David Nagel, Vice President of Advance 18 Technology with Apple Computer, Inc.; Dr. John Wold, Executive 19 Director of the Lilly Research Laboratory, Eli Lilly & Company 20 in Indianapolis, who is accompanied by Dr. Riaz Abdulla, Head 21 of Supercomputer Applications and Molecular Design with Eli 22 Lilly. 23 Without objection, the full prepared statements of all 24 our witnesses will be included in the record. We invite you 25 to summarize what you have to present to the committee today. ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202)289-2260 (800) FOR DEPO 66 1 Dr. Langenberg, we will begin with you. I would like to 2 thank all of you for coming. I know that some of you have 3 traveled great distances. We really appreciate the time and 4 effort you have put into making the hearing today a useful and 5 productive one, and we will hear all of you before going to 6 questions. 7 Dr. Langenberg, please begin. 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. . SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 67 1 STATEMENT OF DONALD LANGENBERG, CHANCELLOR OF THE 2 UNIVERSITY OF MARYLAND SYSTEM AT ADELPHI, MARYLAND 3 Dr -Langenberg: Thank you, Mr. Chairman. Most of my 4 biases on the issues before you stem from my service as chair 5 of a national research council panel that 2 years ago wrote a 6 report entitled Information Technology and the Conduct of 7 Research: The User's View. I will come back to that subtitle 8 in a bit. 9 I would like to make just a few points related to the 10 work of that panel and the issues before you in S. 272. The 11 panel found that there exists significant technical, 12 financial, behavioral, and infrastructural impediments to the 13 widespread use of information technology in research. And 14 though the panel's charge was confined to research, I believe 15 the same impediments exist with respect to education. 16 We made three main recommendations and a host of 17 subrecommendations. S. 272 responds to most of them, and 18 responds very well. One of the panel's principal 19 recommendations was that, and I quote, "The institutions 20 supporting the Nation's researchers led by the Federal 21 Government should develop an interconnected national 22 information technology network for use by all qualified 23 researches." 24 The National Research and Education Network responds 25 directly to the need reflected in this recommendation, and ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 68 1 also, I believe, to the very important collateral need of the 2 education sector. In my judgment, NREN, if that is the 3 correct pronunciation, will revolutionize both research and 4 education, though, of course, in a evolutionary way. 5 My third point is that when one thinks of what NREN might 6 do for education, one thinks first of the education of 7 scientists and engineers, and then perhaps of the incredible 8 potential inherent in linking NREN to every elementary school, 9 every secondary school, every public library, and every museum 10 in the country. 11 There is another educational need of utmost importance. 12 I believe that part of the challenge we face is the creation 13 of an entirely new kind of institutional infrastructure for 14 managing the new information technology, led and supported by 15 a new breed of information professionals. These may bear some 16 resemblance to librarians or to computer scientists or to 17 publishers. And whatever they might be, we need to create 18 schools for training them and institutions within which they 19 can function. And that means educational and institutional 20 innovation of a kind that S. 272 appears well designed to 21 foster 22 My fourth point is that the most important words in the 23 title of our panel report reflect the panel's most important 24 observation. And those words are "the user's view. In 25 simple terms, the panel concluded that the development of ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 69 1 information technology and its applications in the conduct of 2 research -- and here I would say education, as well -- are far 3 too important to be left to the experts. The panel cautioned 4 that planning and development should be guided by users of 5 information technology, both current and prospective, not by 6 information specialists, information scientists, information 7 technologists, or local national and international 8 policymakers. 9 It may not invariably be true that the customer is always 10 right, but institutions that create technology or make policy 11 without a clear understanding and appreciation of the real 12 needs of their clients and constituents risk making serious 13 and expensive blunders. S. 272 calls for the advice of users 14 in the development of a national research and education 15 network, and I especially applaud this provision. 16 My fifth point is a very strongly held view. In my 17 preface to our panel's report I wrote, and I quote, "I share 18 with many researchers the strong belief that much of the power 19 of science whether practice by scientists, engineers, or 20 clinical researchers derives from the steadfast commitment to 21 free and unfettered communication of information and 22 knowledge," This principle has been part of the ethos of the 23 global research community for centuries, and has served it and 24 the rest humanity well. 25 If asked to distill one key insight from my service on ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 70 1 this panel I would respond with the assertion that information 2 technology is of truly enormous importance to the research 3 community, and hence to all humanity, precisely because it has 4 the potential to enhance communication of information and 5 knowledge within that community by orders of magnitude. We 6 can only now dimly perceive what the consequences of that fact 7 may be. 8 That there is a revolution occurring in the creation and 9 dissemination of information, knowledge, and ultimately, 10 understanding, is clear to me. It is also clear to me that it 11 is critically important to maintain our commitment to free and 12 unfettered communication as we explore the uses of information 13 technology in the conduct of research. 14 Now, what I asserted there about research, I would assert 15 now about education. And if I am right, and by far the most 16 profoundly important consequence of the creation of NREN will 17 not be the expedition of research or the improvement of next 18 year's balance of trade. It will be the fundamental 19 democratization of all the world's knowledge. And this means 20 placing the accumulated intellectual wealth of centuries at 21 the beck and call of every man, woman, and child. 22 What that might mean ultimately can only be guessed, but 23 let me reminisce for a moment. I grew up in a small town on 24 the Great Plains, and in that town was a Carnegie library; one 25 of hundreds Andrew Carnegie endowed across the Nation. In ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 71 1 that modest building in the modest collection of books that it 2 housed opened the world to me, and I have been grateful to 3 that Pittsburgh steel maker ever since. 4 What if I had had direct personal access to the Library 5 of Congress, the British museum, the Louvre, and the Deutsches 6 Museum all in the course of a summer afternoon in North 7 Dakota? Just imagine. Now, my point here is that there is an 8 overriding public interest in NREN and in the rest of the 9 provisions of S. 272, an interest that transcends research and 10 its industrial applications or issues of governance and the 11 timetable for commercialization. I truly believe we have an 12 opportunity here for an American achievement of truly 13 Jeffersonian proportions. Let us not blow it. 14 Finally, for my sixth point, I note with approval that 15 S. 272 identifies the National Science Foundation as the lead 16 agency for the development of NREN. The choice is wise, I 17 think. NSF has a demonstrated capacity to manage large, 18 complex, technical operations. Unlike other S&T agencies, 19 NSF's focus is not on some mission, but on its users, that is 20 to say, its client science and engineering communities. 21 And perhaps most important, alone among Federal agencies 22 NSF bears responsibility for the support of research across 23 the full spectrum of science and engineering disciplines and 24 for the training of those who perform the research and for the 25 general education in science and technology of everybody else. ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 72 1 Now, Mr. Chairman, you will have gathered that I have a 2 considerable enthusiasm for S. 272; I do. I urge you and your 3 colleagues- to enact it into law. 4 Thank you. 5 [The prepared statement of Dr. Langenberg follows:] 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800.) FOR DEPO 73 1 Senator Gore: Thank you very much. We certainly 2 appreciate your forceful statement and the way you delivered 3 it. 4 Dr. Kalos, from Cornell, welcome. Swing that microphone 5 around there so we can hear you. Thank you very much for 6 coming. 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 74 1 STATEMENT OF MALVIN H. KALOS, DIRECTOR, CORNELL THEORY 2 CENTER 3 Dr. Kalos: I am sitting on the end to give me ready 4 access to the computer terminal, because I am going to 5 demonstrate some. 6 Senator Gore: Very good. 7 Dr. Kalos: Mr. Chairman, it is a privilege to be invited 8 to comment on S. 272, the High Performance Computing Act of 9 1991; however, being asked to follow Dr. Bromley and others 10 makes me feel like the man who survived the Johnstown Flood. 11 The Cornell Theory Center, which is dedicated to the 12 advancement and exploitation of high-performance computing and 13 networking for science, engineering, and industrial 14 productivity, is one of the National Science Foundation 15 Supercomputer Centers. We are part of the transformation of 16 our science and engineering culture brought about by the 17 advent and adoption of high-performance computing and 18 communications in our technological society. 19 Senator Gore, Senator Pressler, Senator Robb, and the 20 other cosponsors of this bill, and the President, understand 21 the -deep and positive implications for our future. 22 Dr. Bromley has done essential work in translating these ideas 23 into effective policy. FCCSET for the first time has unified 24 the Federal approach to high-performance computing. Theirs is 25 a well-designed, well-integrated program that shows good ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 75 1 balance between the need to exploit advancing supercomputing 2 technology, the need for very high-performance networking, and 3 the need to bring these new tools to the widest possible 4 community through research and education. 5 The aim of fundamental science is to connect all our 6 knowledge in a seamless webb of quantitative understanding. 7 This is now harder to do, because we probe into more and more 8 complex phenomena that defy analysis by the mathematical tools 9 we have. Computational modeling is essential to fill this 10 need. Many areas of science involve systematic connection 11 among different phenomena at different scales of length or 12 energy. Chemistry, biology, medicine, science of materials, 13 astrophysics, are very good examples. 14 Computation is also an essential tool in experimental 15 science. The most advanced instruments, optical and radio 16 telescopes, particle accelerators, and computers themselves, 17 are studied, designed, optimized, and verified with the help 18 of computer simulation. Data collection is automated. The 19 reduction to comprehensible data sets involves enormous 20 computations in some cases. The exchange of large data sets 21 will require very heavy use of high-capacity data networks. 22 An important step in modern science, I believe, was the 23 creation by the Congress and the National Science Foundation ) 24 of the National Supercomputer Centers. That was the mark of 25 the entry by the mainstream of American research into this new ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 76 1 era of computational science. The entire scientific and 2 engineering community of the Nation has the opportunity to 3 exploit these new tools. Students and young scientists, 4 always the very heart of any important scientific change, are 5 now involved. They will carry the message to the rest of our 6 society and to the future. 7 I would like also to comment that the present program 8 includes attention to education. The NSF program, the 9 supercomputer programs include for example at Cornell the 10 Superquest program which is bringing knowledge and training of 11 supercomputing to high schools around the Country. 12 I will show some videos showing significant scientific 13 advances made possible by supercomputing, and I would like to 14 comment to Senator Pressler, in particular, that some of the 15 advances that I have in my written testimony are those that 16 come from small schools. So these centers provide this power 17 to large institutions and small, primarily research 18 institutions, primarily undergraduate institutions, and this 19 is a very important balance. 20 Another vital role of computational science is that of 21 permitting quantitative connections among different 22 disciplines. Every one of the large problems that confront 23 our society, and to whose solutions we expect science to 24 contribute, is in some sense a multidisciplinary problem. 25 Issues of the environment and medicine, to site only two, ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 77 1 involve many sciences working together; chemistry, physics, 2 engineering, fluid flow, biology, understand of materials. 3 Bringing the knowledge from these fields together to make 4 quantitative predictions about the effect of some 5 technological or regulatory proposal would be utterly 6 impossible without the use of high-performance computational 7 modeling, which is the natural language, the indispensable 8 lingua franca of quantitative multidisciplinary research. The 9 supercomputing community will soon find itself at a major 10 crossroads where the increases in performance needed for our 11 scientific mandate will demand parallel architectures. 12 To exploit these new machines, a major retooling of 13 software and algorithms will have to take place. This must be 14 started very soon if we are to make progress on the grand 15 challenge problems in the mid-1990's. The high-performance 16 computing and communications program will offer us an 17 essential opportunity to bridge the gap between today's 18 high-performance vector machines and tomorrow's highly 19 parallel systems. 20 I have emphasized how science and its application to 21 societal problems involve the national scientific community. 22 Bringing to bear this transformation of computational science 23 in the most complete and positive way requires that its 24 techniques and strategies be learned, used, and shared by the 25 widest possible group of researchers and educators. All of ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 78 1 these are necessary, and the appropriate level and balance 2 among them is essential. The High Performance Computing Act 3 of 1991 is a vital step in that direction. 4 And now I will move to the screen and I am going to show 5 three videos with different scientific themes, and each has a 6 different theme to bear on the application of science to 7 industry or medicine. 8 [First videotape shown.] 9 Dr. Kalos: The first one shows the investigation, which 10 involves some important algorithmic advance by Mike Teeter, 11 who is a professor of Cornell and an engineering fellow of 12 Corning Glass. 13 We are going to see -- please start it. We are going to 14 see a model, first a simple ball and stick model, of a quartz 15 crystal. Then, the silicon and oxygen atoms will be dressed 16 in fields that represent the electron fields at various 17 densities. There are three of them, and the lowest of level 18 of electron density is shown in blue. 19 Now we see the blue level only, and we see the ramified 20 electron field that permeates the entire crystal and gives it 21 its structure. The importance of this for Corning is that 22 understanding the physics of quartz means understand the 23 physics of glass, and this was translated into making better 24 optical fibers and an important competitive advantage for 25 Corning. ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800). FOR DEPO 79 1 [Second videotape shown. ] 2 Dr. Kalos: The next movie is going to show something you 3 have talked about. The composition of a set of data 4 representing a sedimentary oil field in the Gulf of Mexico 5 assembled by a team of 19 organizations, 11 petroleum and 8 6 academic institutions. 7 We see a 30 by 10 mile area of the floor of the Gulf of 8 Mexico. The green zones show places where oil or gas has been 9 seen. This is a set of salt domes. They are 6 miles high 10 under the ocean. And what we are going to see -- well, now we 11 see sand and shale zones. But more important, we are going to 12 see patterns of heat-flow throughout these salt domes. And 13 the patters of heat-flow are correlated with the presence of 14 petroleum. 15 Now, this is simply the assembling for the first time of 16 disparate data from all of the partners in this. Assembling 17 it, producing this video, has presented new insights for the 18 geologists. They think they understand better than ever 19 before how to improve the recovery of petroleum from existing 20 sources worldwide. They are also going on to do serious, very 21 heavy computational modeling to try to understand in a still 22 more fundamental way the processes that are going on. 23 We are flying through the data; we are understanding what 24 is really happening. 25 Senator Gore: This is an example of what they mean when ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 80 1 they say that computational science has now joined inductive 2 reasoning and deductive reasoning as a third new branch of 3 knowledge creation. 4 Dr. Kalos: Absolutely. Yes. Incidently, these partners 5 are also far apart geographically and will need to exchange 6 these data through high-capacity networks as they work. 7 The third video has to do with the uses of ultrasound, 8 low-intensity and high-intensity ultrasound in probing the 9 human eye. 10 [Third videotape shown.] 11 Dr. Kalos: First, we will see how low-intensity 12 ultrasound is used to survey the existence of tumors in a 13 human eye. A data set is being built up, reduced to a form 14 understandable by the computer. The location of the tumor is 15 indicated. Incidently, this is the retina, which is so 16 distorted by the tumor that it is torn away from its usual 17 position in the eye. A data set is being sliced off and 18 assembled into a three-dimensional data set that the computer 19 can understand. 20 And we see in animation how this is done. We see the 21 retina there, the distorted retina. And here a 22 cómputer-usable model has been assembled. And now we see the 23 tumor in three dimensions as it rotates. Now the model is 24 used in a mathematical way to understand how the illumination 25 by high-intensity ultrasound would affect that tumor. And we ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 81 1 see the simulation of the effects of heating by a focused, 2 high-intensity ultrasound beam. 3 You see, of course, the high temperature at the center of 4 the focus. In therapy, that focus would be steered around the 5 tumor and would literally cook it into oblivion. 6 Thank you. 7 [The prepared statement of Dr. Kalos follows:] 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. . SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 82 1 Senator Gore: Boy, that was really impressive. Well, we 2 will save our questions, but thank you so much for your 3 presentation. Very impressive. 4 Our third witness on this panel is Mr. Tracey Gray, vice 5 president for marketing with the Government Systems Division 6 of US Sprint. Mr. Gray, thank you so much for joining us 7 today, and please proceed. 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 83 1 STATEMENT OF TRACEY GRAY, VICE PRESIDENT OF MARKETING, 2 GOVERNMENT SYSTEMS DIVISION, US SPRINT COMMUNICATIONS COMPANY 3 LIMITED PARTNERSHIP 4 Mr. Gray: Thank you, Mr. Chairman, and Members of the 5 subcommittee. Now, these are two tough acts to follow, and I 6 do not even have my video to support me. I would like to give 7 you some comments from the perspective of the -- of a business 8 poised to take advantage and to offer to the Government and to 9 the many users of academia and industry in this country, the 10 kind of capabilities that you have designed this bill to 11 support. 12 First of all, we endorse and support this bill. I would 13 like to reference my comments, really, along the lines of how 14 we as a business in the industry see fulfilling and 15 participating in the objectives of this legislation and 16 initiative, and how we believe that we can bring some of those 17 objectives to bear. 18 I would like to depart just briefly, though, to remind 19 everybody of the power of the private sector to bring to bear 20 resources, capability, technology, and the business resources 21 and energy to meet goals of this type, providing the 22 incentives and the understanding and the perceptions are 23 properly put in place. And I think that is what this bill is 24 all about. 25 I would like to remind everybody here that just about 6 ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800). FOR DEPO 84 1 years ago, this country saw one of the most significant 2 economic realignments in the industry that we have ever had. 3 That was the divestiture of AT&T. US Sprint was an outgrowth 4 of that. 5 US Sprint today has a network that is just a little under 6 4 years old. It is a fully deployed, fiber-optic network that 7 transcends the Nation. It has 23,000 miles of fiber in it. 8 We support millions of residential, business, large 9 corporations and Government customers. We support 500,000 or 10 more Government customers today on that network, supporting 11 them with voice, data, video, imaging, high-speed facsimile, 12 electronic messages, packet services, and a vast array. of 13 private line services. 14 Indirectly we support hundreds of thousands of other 15 Government services through inter-operability of our networks 16 with others. We are also deploying at this time -- bear in 17 mind, this is a network slightly less than 4 years old -- we 18 are deploying at this time the capability to support sonic 19 technology. 20 What that really means is very wide band high-speed 21 network and data transmission capability within the network. 22 Bear in mind, again, this is a network slightly over 3 years 23 old. We have well over $3.5 billion invested in this network. 24 We support this bill from the standpoint that it provides 25 the seed money, the initial stimuli that we think is necessary ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 85 1 to develop the next level of applications and to bring about 2 the incentives in the private sector to make the next step, 3 which could be a quantum leap, in the deployment and the 4 investment of technology that will support the multi-gigabyte 5 transmission paths that are necessary to achieve the 6 objectives you are talking about. 7 We are very heartened to hear and see the recognition of 8 the importance of the communications link, the superhighway 9 network. We believe that the National Research Education 10 Network objectives can be achieved with this initiative, with 11 this measure. 12 I would like to bring two issues to the fore that, to 13 give us, if not some concern, some reason to watch what you 14 are doing and look at what you are doing and participate in 15 it, to ensure that these issues are dealt with. And I believe 16 Senator Pressler articulated one of those very well. And that 17 is to ensure that the funding and the development and the 18 participation is broad enough to encourage and to support the 19 users of something other than supercomputers and the users who 20 may only have to rely on multi-gigabyte networks. 21 A reason for believing this, and promoting this, is we 22 know from experience that the cost benefits and the likelihood 23 of seeing timely development, timely deployment, of the types 24 of technology you are talking about is very dependent upon a 25 broad base of users. The more fully cost can be allocated and ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. . SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 86 1 distributed along many users, the better off everybody will 2 be, and the faster and the more timely development of these 3 technologies will take in the private sector. 4 I think an example I would like to cite where we have had 5 the technology, we have had the capability for some time, and 6 yet we have not seen the full benefits and the full 7 optimization of the technology and the applications in the 8 area of integrated services digital networks, this thing 9 called ISDN. 10 Certainly the technology and the capability is there, but 11 the applications are not. And Dr. Bromley spoke eloquently 12 about the need for software and the development of basic 13 applications to drive and to take advantage of the technology 14 of that type. 15 So we encourage the planners and the architects of this 16 legislation to ensure that there is a broad participation in 17 the academic, educational, and industrial community, beyond 18 those that just rely on supercomputers. Secondly, we 19 encourage and we will do all we can through our participation 20 to ensure that the development of the network itself is a 21 development and a plan that will permit the utilization of 22 public networks to support these services and to support the 23 capabilities. 24 We do not believe that it requires a private network 25 development or application to support your objectives. We do ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 87 1 believe, and our experience tells us, that the maximum cost 2 benefits, the long-term interest of the users of a network can 3 be found with shared network applications. We have seen time 4 and again the problems that develop with private networks, 5 where you have a group of users stranded with a set 6 technology. 7 We also believe that public services and the 8 commercialization of these applications and products will 9 ensure that the Government, the need for the infusion of 10 Government money, will cease over time, will minimize over 11 time. 12 I thank you, Mr. Chairman, and members of the 13 subcommittee. That is a business perspective that differs 14 slightly from my panel members. I welcome this opportunity to 15 have spoken to you. Thank you very much. 16 [The prepared statement of Mr. Gray follows.] 17 18 19 20 21 22 23 24 25 ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 88 1 Senator Gore: Thank you very much. Extremely valuable 2 statement. And may I say, in echoing the words of Dr. 3 Bromley, that the real-life perspectives or reality checks, if 4 you will, which we have gained from our dialogue with your 5 company and others, have been invaluable in shaping this 6 legislation. We really appreciate your statement here today 7 as well. 8 Next, Dr. David Nagel, vice president for advanced 9 technology with Apple Computers. You are invited to proceed 10 at this time. Welcome. 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800). FOR DEPO 89 1 STATEMENT OF DAVID C. NAGEL, VICE PRESIDENT, ADVANCED 2 TECHNOLOGY, APPLE COMPUTER, INC. 3 Dr. Nagel: Thank you, Mr. Chairman. I am appearing 4 today not only on behalf of Apple Computer but also on the 5 behalf of the other members of the Computer Systems Policy 6 Project. We are very appreciate of the opportunity to appear 7 for the subcommittee on a favorite topic of high-performance 8 computing and networking. 9 In the fall of 1989, the 11 largest computer systems 10 companies in the U.S. formed the Computer Systems Policy 11 Project to address what we felt were some fundamental problems 12 facing our industry. It was a measure of the importance of 13 this activity that the CSPP is an association of the chief 14 executives of our companies; the CEO's are supported by the 15 chief technologist for each company, like myself, and by a 16 permanent professional staff in Washington. 17 We began our study more than a year ago with an internal 18 look at the health of our industry. We assessed technologies 19 that we believe are critical to our industry. We assessed how 20 the U.S. is doing relative to other countries and our foreign 21 competitors in those technologies, and we developed a 22 prognosis for U.S. industry performance into the future. 23 While by almost any measure, our industry is still the 24 strongest in the world, our lead appears to be diminishing 25 rapidly by almost all of the measures that we examined. In ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 90 1 1983, for example, U.S. companies held an 83 percent share in 2 the world market of computer systems, including software. 3 Between 1983 and 1989, our share of the world market declined 4 by more than 20 percent, from 83 percent to about 61 percent. 5 During the same period, Japan's share rose from 8 percent to 6 22 percent and our European colleagues' share grew from 10 7 percent to 15 percent. 8 More troubling, the computer systems industry went from a 9 significantly positive contribution to the U.S. trade balance 10 all throughout the 1980's to a position in 1990 where our 11 imports almost exactly balance our exports. While the U.S. 12 ratio of computer exports to imports moved steadily downward 13 over the last decade, Japan even more dramatically increased 14 its export-import ratio from about 2 in 1980 to more than 6 at 15 the end of the 1980's. 16 While these findings are clearly troubling to us, the 17 members of CSPP recognize that the primary burden for staying 18 competitive in the global marketplace rests squarely with our 19 own industry. So we began with an internal assessment. We 20 examined our own investment levels and competitive positions 21 in the key technologies which we think are critical to 22 success. We identified, for example, 16 critical pre- 23 competitive generic technologies, and concluded that the U.S. 24 still leads the world in half of these. And most of these are 25 software intensive. ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 91 1 We also concluded that the U.S., once leading the 2 remainder, now lags the world in several critical 3 technologies, and is losing a lead in the remainder. And most 4 of these, and in contrast to the technologies for which we 5 hold a lead, the lagging technologies are mostly capital- 6 intensive ones. We also believe that, without further 7 positive action, the U.S. position will erode further in all 8 of these 16 technology areas over the next few years. 9 The computer systems industry spends 21 percent of the 10 private sector R&D, or about 10 percent of the total national 11 investment, in research and development. The investment of 12 the computer industry in 1989, more than $18 billion, is more 13 than that of any other industrial sector, and represents a 26 14 percent increase over the amount we spent in 1988, a period 15 when many other industrial sectors were reducing their R&D 16 spending. 17 In contrast to the level of investment in private 18 industry, the U.S. Government only invested about 2 percent of 19 its R&D portfolio in generic technologies related to our 20 industry. 21 Taken as a whole, we conclude that the Federal R&D budget 22 in computing is not today focused or managed in ways that are 23 needed to preserve and enhance our economic competitiveness, 24 given the rapid pace of innovation and the R&D practices of 25 other countries. In short, we believe the Federal R&D has not ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 92 1 been as helpful to the computer industry as it might be. 2 Based on our analysis and this conclusion, we have 3 outlined an initial set of technology policy recommendations. 4 These provide a strategy, we believe, for better focusing the 5 Federal R&D investment in these pre-competitive generic 6 technologies, and will help us meet very stiff international 7 competition. 8 We believe that the Government and industry must work 9 together, and jointly must take the following first steps to 10 improve the effectiveness of the spending in the U.S. First, 11 we think that there should be an improvement in the mechanisms 12 within OMB for reviewing the Federal R&D spending program. In 13 many cases, these have become so complex, it is very difficult 14 to actually figure out what is being spent. 15 Number two, we need to increase the industry input, we 16 believe, in setting Federal R&D priorities and to better 17 manage the Federal R&D budget. Number three, we think 18 industry should work with the Federal labs and with Government 19 agencies to improve, to set Federal laboratory priorities and 20 improve the return on the national investment made in these 21 labs. 22 And fourth, we look forward to working with the 23 Government in implementing high-performance computing 24 industries, including a national network capability of 25 bringing the benefits of computing to every institution, ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 93 1 household, business, and school in the Nation. 2 We have created three CEO-level working groups to address 3 our industry's participation in the Federal R&D priority 4 setting. And we are looking here at structural and legal 5 impediments of which there appear to be a variety. We are 6 increasing the degree of interaction between industry and the 7 programs in the Federal labs. And finally, we are looking at 8 ways in which we can better participate in implementing the 9 high-performance computing and communication initiatives. 10 We fully support and recommend full funding for the 11 national high-performance computing and communication 12 programs, including a National Research and Education Network. 13 We recognize and applaud the pioneering role that this 14 subcommittee and its chairman have played in recognizing the 15 importance of the development of a national information 16 infrastructure and an effective, high-performance computing 17 program. 18 We believe this efforts are critical in providing the 19 research infrastructure in maintaining our Nation's leadership 20 in basic computer and information research. The CSPP believes 21 that the high-performance computing and communication 22 initiatives will be instrumental achievement of the national 23 education in work force training goals. 24 Now, much has been written and said about the benefits of 25 high-speed networking at the institutional level, of higher ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800). FOR DEPO 94 1 education levels. While we agree with and support these uses, 2 high-speed networks will allow the rate of scientific and 3 engineering progress a major grand challenge problems to 4 accelerate significantly. 5 But we also believe the benefits of high-speed networking 6 and high-performance computing should ultimately find their 7 way beyond institutions and become available to the rest of 8 us. And I am echoing this, both what my colleagues have 9 already said and what have been said by the senators. 10 I would like to briefly touch on some other benefits of 11 what we believe are a truly universal high-speed network, 12 benefits that will eventually impact a much larger number of 13 our citizens. 14 Actually, we are beginning to see the precursors of the 15 benefits of networking and distributive computing, even with 16 the overly complex low-speed network systems currently in 17 place in the U.S. Apple Computer, for example, in a project 18 called Apple Global Education, or AGE, has made our own 19 internal slow-speed electronic mail system available so that 20 school children all over the world can communicate and 21 exchange ideas. 22 For example, recently on Earth Day, 12 schools from 23 around the world collaborated and prepared and produced local 24 newspapers that featured environmental issues using this 25 network. On other networks, we have begun to see a variety of ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 95 1 education applications develop. We have seen collaborations 2 between teachers in the preparation of educational materials. 3 We have seen collaborations between students conducting 4 scientific investigations. We have seen networks allow 5 students in elementary and high school to benefit from access 6 and experts in universities. And we have seen a variety of 7 on-line courses and instructional materials being prepared and 8 disseminated electronically every day. 9 Both students and teachers have access to a widening 10 range of information, databases and computing resources, all 11 remote to their physical locations. We believe these 12 applications are springing up everywhere, even though there 13 are many impediments, because our educational system is 14 discovering the value of electronic information delivery, even 15 with the very slow speed networks available today. 16 With the arrival of data exchange capabilities like those 17 that will be provided by the NREN, capabilities that will 18 allow graphics and images to be transmitted as easily as text, 19 we should see an explosion in new uses of high-speed 20 networking and education, uses which we think will 21 fundamentally transform the whole process of education. 22 Teachers and students who are in remote, rural areas -- 23 in some cases, remote urban areas -- far from major libraries 24 and universities, will have access to information and 25 expertise every bit the same as their counterparts in the most ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 96 1 favored settings. Physical separation will no longer matter 2 when on-line video conferencing and other high-speed network 3 supported applications are available to every school in the 4 U.S. 5 We believe also that U.s. business will benefit from 6 high-speed networking. Apple, for example, has greatly 7 benefited from our own internal electronic mail system, as 8 have many other companies; in our case, an information 9 exchange system called Apple Link. Using Apple Link, 10 individual contributors exchange ideas and documents with one 11 another, with their managers, and with the executives of the 12 organization. From time to time, they exchange insults with 13 one another. 14 Some of the fundamental administrative activities within 15 Apple, activities that were used to generate large piles of 16 paper, now are done almost entirely electronically. The 17 advanced technology group which I head uses a video 18 conferencing network to tie together our four separate 19 physical labs in the U.S. Without this system, it is clear we 20 would all have to do a great deal more traveling and generate 21 waste along a variety of dimensions. 22 Finally, Apple uses high-performance computing to great 23 advantage both in the development of technology for products 24 and in the development of products themselves. Over the past 25 year, for example, we have used our Cray to develop advanced ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 97 1 algorithms for data compression, high-performance graphics, 2 and speech recognition, among other applications. 3 Once we prove to our satisfaction that these algorithms 4 work on our Cray, we can design special circuits, again using 5 the Cray, which makes Cray levels performance for specialized 6 applications available on our advanced personal computers at a 7 tiny fraction of the cost of a supercomputer. 8 So high-performance computing helps us both in the 9 product sense and in the technology sense. 10 In conclusion, we recognize that improving U.S. 11 technology policy is a long-term process, cannot be addressed 12 by any one organization, any single set of recommendations or 13 any given piece of legislation. Improvement of U.S. 14 technology is, nonetheless, an essential process that will 15 require the cooperative R&D investments and partnership of 16 both the private sector and the Government. We believe that 17 improving U.S. technology requires a long-term commitment and 18 a series of changes by both industry and Government over time. 19 Whether as independent CEO's or as an industry, the members of 20 CSPP are committed and will remain involved in this process. 21 Thank you very much. 22 [The prepared statement of Dr. Nagel follows.) ] 23 24 25 ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 98 1 Senator Gore: Thank you, Dr. Nagel, and my thanks to 2 CSPP for the input. Your company, your member companies have 3 provided throughout this project -- and I now you are 4 representing the project, but please convey my personal 5 gratitude to John Sculley for his personal support and 6 encouragement of this legislation as it has gone along. 7 Dr. Nagel: I will be happy to do just that. Thank you. 8 Senator Gore: I have many questions, as I know Senator 9 Pressler does. But we have one more witness, and it is one we 10 have been looking forward to hearing: Dr. John Wold, 11 executive director with the Lilly Research Lab at Eli Lilly. 12 And you are accompanied, Dr. Wold, by Dr. Riaz Abdulla, head 13 of supercomputer applications and molecular design at Lilly. 14 Dr. Wold, please proceed. 15 16 17 18 19 20 21 22 23 24 25 ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800). FOR DEPO 99 1 STATEMENT OF JOHN S. WOLD, EXECUTIVE DIRECTOR, LILLY 2 RESEARCH LABORATORIES, ELI LILLY AND COMPANY; ACCOMPANIED BY 3 RIAZ ABDULLA, MANAGER, HIGH-PERFORMANCE COMPUTING PROGRAM, ELI 4 LILLY AND COMPANY. 5 Dr. Wold: Thank you, Mr. Chairman. Eli Lilly and 6 Company is a global corporation based in Indianapolis, 7 Indiana, that applies advancement in science to basic human 8 needs, health care, and nutrition. We compete in the 9 pharmaceutical, medical devices, diagnostic product, and 10 animal health products industries. 11 My responsibilities at Lilly include the company's high- 12 performance computing program. And with me, as you just 13 alluded to, is my colleague, Dr. Riaz Abdulla, who manages 14 this program on a day-to-day basis, and is himself a 15 practicing supercomputer user. 16 I would be pleased to have this opportunity to present my 17 company's views about the importance of a national commitment 18 to high-performance computing and to high-capacity network. I 19 am sure that this subcommittee has heard -- it will hear much 20 more in the future -- about the underlying technology required 21 to support the evolution of supercomputers and high-capacity 22 networks. 23 It is important that you share our computering 24 technologists' excitement about their visions of 25 supercomputing systems. But I think it is also important that ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 100 1 you share the visions of a research-based institution like 2 ours which have motivated us to invest in supercomputers. 3 Long-term success in the research-based pharmaceutical 4 industry depends on one factor: innovation. We must discover 5 and develop cost-effective new products that improve the 6 quality of life and offer economic benefits to patients, 7 payors and society as a whole. 8 Pharmaceutical R&D has traditionally been a high-risk, 9 complex, time-consuming, and costly enterprise. Over the past 10 half-decade, the research-based pharmaceutical industry has 11 experienced major changes. The rapid escalation of R&D costs 12 has helped precipitate major structural changes in the sector 13 of the global economy in which the U.S. is an established 14 leader. 15 An unprecedented wave of mergers, acquisitions, joint 16 ventures, has led to fewer, larger competitors. Competition 17 in the research-based pharmaceutical industry will only become 18 more challenging during the 1990's and beyond. Consequently, 19 my company, Lilly, has evaluated many opportunities to 20 reinforce its capacity to innovate, to enhance its capacity to 21 compete. Supercomputing is a case in point. 22 Our supercomputing experience was initiated through our 23 partnership with the National Center for Supercomputing 24 Applications at the University of Illinois, the NCSA. The 25 NCSA has prepared a video segment that describes our ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 101 1 involvement in their program. If we can run the video, I will 2 conclude my remarks after the video. 3 [A video was shown.] 4 Dr. Wold: Mr. Chairman, as you pointed out earlier, 5 supercomputing has created a new common language for research. 6 In recent years, scientists have developed mathematical 7 methods describing the realistic shape and motion of atoms in 8 large molecules, such as receptors or enzymes that exist in 9 the human body. These models are now emerging as important 10 tools for scientists probing new investigations into how 11 potential drug candidates would likely affect these molecular 12 targets. 13 These mathematical descriptions are based on equations 14 involving billions of numbers. Conventional computers take 15 days or weeks to perform these calculations, but 16 supercomputers can do this work in minutes or hours and permit 17 previously impossible calculations. 18 Graphic representations of the data serve as a new 19 communications medium, a new language for scientists. Teams 20 of scientists can share the same visualized image of how a 21 specific chemical agent would likely affect the receptor in 22 question. They can quickly evaluate the probable effects of 23 modifications in a chemical. They can focus the painfully 24 slow efforts required to synthesize and test new agents on 25 those compounds that appear to have the greatest potential. ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 102 1 Our experience to date suggests three interrelated 2 advantages high-performance computing to our industry. These 3 systems will speed up the identification of promising drug 4 candidates. Supercomputing will enable our scientists to 5 design new drug candidates that they otherwise would not have 6 even considered. 7 These systems will foster greater collaboration among 8 scientists from various disciplines who are involved in 9 pharmaceutical research and development. Supercomputer- 10 generated graphic simulations help scientists with diverse 11 academic training to share the same vision of crucial data. 12 And thirdly, these systems will encourage truly visionary 13 exploration. Now, supercomputers are motivating our 14 scientists to ask "what if" more boldly than ever before in 15 helping them to quickly consider many possible answers to 16 their questions. 17 I want to stress that supercomputing is only a tool. But 18 it is a very powerful scientific tool, a tool that will become 19 all the more powerful with networking capabilities. A high- 20 capacity network will greatly facilitate the dynamic 21 collaboration among scientists at different locations and 22 often different institutions. The network will help us 23 optimize scarce scientific talent during a period when we are 24 almost certain to experience major shortfalls in availability 25 of highly trained scientists. ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 103 1 Finally, a high-capacity network will help scientists 2 raise questions that they could never seriously ask before. 3 In conclusion, I want to stress two points. We believe that 4 supercomputers in a national, high-capacity network are 5 important to our company, to our industry, and to the medical 6 professionals and patients we serve. And we believe that 7 high-performance computing will play a crucial role in the 8 many technology-based industries and in the growth of national 9 economies that depend on these industries. 10 We strongly recommend the enactment of the High 11 Performance Computing Act of 1991 and thank you for this 12 opportunity to share our thoughts with the committee. 13 [The prepared statement of Dr. Wold follows.] 14 15 16 17 18 19 20 21 22 23 24 25 ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 104 1 Senator Gore: Thank you very much. 2 Senator Pressler will have to depart for another 3 committee hearing, and I want to recognize him first. 4 Senator Pressler: Thank you very much, Mr. Chairman. I 5 am going to ask most of my questions for the record. I just 6 want to ask one question of the panel, if I may. And that is, 7 is the current fiber-optic infrastructure sufficient to handle 8 supercomputer network envisaged in S. 272 or envisaged in the 9 administration's proposal? 10 I do not think it is. But the question is how do we get 11 the fiber-optic infrastructure? How do we accomplish that? 12 Some people say we let the telephone companies, the regional 13 telephone companies into cable TV, and they will do it, so 14 they say. Others say we require the cable TV companies to 15 start laying fiber optics rather than copper. 16 And I guess a second question to that, and might be 17 addressed to Mr. Gray, is what type of user or consumer demand 18 needs to occur before private companies like yours will begin 19 to connect homes and small businesses with fiber optics to the 20 supercomputer network? 21 So, in other words, the underlying question, and some of 22 you want to think about it a little bit more, but I think that 23 is a basic question. I am working on that in the 24 communications subcommittee and some other legislation that is 25 related to this. How do we get the fiber-optic infrastructure ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 105 1 to support this? 2 Mr. Gray: I will try a first shot at some of that. It 3 certainly_will not be all inclusive. But certainly we as a 4 carrier, as well as our competitors, the more evident it 5 becomes to us and the greater the probability there is for 6 applications of the type that we are talking about here, and a 7 large user base to be established. 8 And frankly, that is what I see this initiative really 9 precipitating. It becomes a coalescing force to bring those 10 things together. We as individual members of the industry 11 cannot bring all that together; we cannot get the computer 12 industry, the users, academia, we cannot pull them together. 13 What you are proposing here does begin to coalesce those 14 forces and bring some focus, and at least provide some, some 15 perspective on our part that this could happen, as a private 16 industry. Therefore it gives us the incentive to divert and 17 to reorganize priorities to shift our investment toward these 18 kind of capabilities. 19 And within the industry, we have worked with the exchange 20 carriers; we are very dependent on them to extending the 21 capabilities of our network to their users. And more and more 22 competitive forces are at work there, because we do have 23 options other than the exchange carriers to get to our 24 customers. 25 So there are fairly powerful marketing influences that ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. e SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 106 1 can drive this, providing there is a, the infrastructure 2 provided by the Government to fund and seed some of this, and 3 provide the stimuli to make those things happen. 4 Dr. Nagel: It may well be that the, that the political 5 regulatory and problems of that category far outweigh the 6 difficulties of getting the technical infrastructure in place. 7 I think even from the limited look we have taken at this so 8 far, the technical problems are very minimal relative to the 9 problems that we have just talked about, getting people 10 together to work on something like this. 11 Dr. Abdulla: Senator Pressler, as a user, I would like 12 to very directly state that the answer to your first question 13 is no, the existing infrastructure is not sufficient. And 14 what the proposals that we have heard today really tell us 15 about is a paradigm shift. 16 You are talking about the difference between a 17 teleprinter and a telephone, is how different we are talking 18 about. It is going to completely change the way we do things. 19 Senator Pressler: I think we have got a big job to get 20 that fiber-optic infrastructure built somehow. It is like 21 wiring the Nation, and we have to find a way to do it. The 22 big telephone companies say they can do it, but they will only 23 do it if we let them get in the cable TV business, stuff like 24 that. 25 We have to find a way, and then we have to find a way ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 107 1 that everybody has access to use, it is kind of like a gas 2 pipeline; people will have a right to have access to that 3 fiber-optics cable somehow. And if one company controls it or 4 something, or does not let other people use it, the whole -- 5 it is a very difficult problem, as I see it. 6 So this whole supercomputing thing is great. But I see 7 this fiber-optic infrastructure thing as a very great problem 8 we have to solve. And I do not really have all the answers; I 9 am working on some legislation that I hope will solve part of 10 it. But any of you, if you think of any great ideas as you 11 ride the train home, tell me. 12 Senator Gore: If I could supplement this for the record, 13 let me just offer my 2 cents' worth on this. I have supported 14 the entry of telephone companies into the cable television 15 business, but that is an extremely controversial proposal 16 which may not finally pass. 17 I do believe that the measures included in this 18 legislation will result in the unleashing of forces which will 19 inevitably lead to the wiring of the Nation. In fact, the 20 fiber-optic capacity which is already in place is adequate for 21 the long distance links, provided we make available the new 22 switches, the new software, and the new algorithms, which will 23 upgrade the capacity of the existing fibers without requiring 24 the placement of new fibers in the ground or on the poles. 25 The inadequacy of the present fiber network, leaving ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 108 1 aside the software switches and algorithms, the inadequacy of 2 the fiber itself is mainly in what is called the last mile, 3 from the last switching station to the home. It is for that 4 reason that I have supported the Telco entry into cable. 5 But let us assume that that does not happen in the near 6 term. I believe very deeply that once the backbone network is 7 in place, we will witness the emergence of a new generation of 8 information services, a new generation of ways to configure 9 information to make it understandable to people, that we will 10 unleash enormous demand for access to that backbone network. 11 There will be a new set of financial incentives to 12 encourage people to provide that last mile. There will also 13 be the ongoing efforts of merit; just to name one, I know that 14 that is not a -- it is not universally participated in by 15 everybody, but it is a not-for-profit corporation, perhaps one 16 of several that will be active in rapidly expanding the reach 17 of the backbone network. 18 Just as the interstate highway system led to initiatives 19 by States and cities and even private turnpike authorities to 20 connect to the interstate highway system with new, four-lane 21 limited-access roads that were not part of the federal system, 22 as it was initially designed. So this backbone network will 23 quickly, in some cases even simultaneously, lead to the 24 completion of access links, which will themselves encourage 25 access links. ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800). FOR DEPO 109 1 Just as arteries and capillaries are related on down, I 2 think there will be a growing network, a growing network, with 3 lines going to more and more people. 4 What we have now is a chicken-and-egg problem. The 5 market place is not perceiving the demand for these new 6 information services because the network is not there to 7 deliver them. The market is not perceiving the demand for the 8 network to deliver them because the new services are not yet 9 there. Once that chicken-and-egg conundrum is overcome, then 10 the, we will be in a new system of supply, we will be in a new 11 reality. And the demand for these new services, I think, will 12 drive the forces that will encourage the market to complete 13 the national network. 14 Now, I believe that is a realistic vision. But just in 15 case, I also support the Telco entry into CATV, and I will 16 look forward to working on any other proposals that people 17 have to address that. 18 Dr. Langenberg: Mr. Chairman, I might be able to provide 19 some support for that view on a relatively small scale. There 20 is a generic type of community that tends to run from 20,000, 21 perhaps to 40,000 population, that contains an unusually high 22 proportion of serious users of computing facilities. It is 23 called a university. 24 And one after another over the past decade or so, I have 25 watched universities all across the country, fiber-optic ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 110 1 backbone -- provide fiber-optic backbones for themselves. It 2 is a very complex kind of drive for that effort, but it is 3 partly-demand, it is partly based on leadership of university 4 officials who can see the future coming and who want to be 5 prepared to hook into it, once it is there, as a part of 6 maintaining that institution's competitive edge. And it is 7 partly push, partly pull, but it does work. 8 Senator Gore: To use another example, the state of 9 Tennessee Public Service Commission has already embarked on a 10 very ambitious plan to provide high-performance networking to 11 virtually the entire State in anticipation of the completion 12 of the backbone network. 13 I know that there are some other states that are doing 14 the same thing. So I think that it will happen, once the 15 network is there. 16 But let me say that we have a statement for the record 17 from the U.S. General Accounting Office on supercomputing in 18 industry -- thank you very much, Senator Pressler -- and we 19 will include this for the record. 20 [The material referred to follows.] 21 22 23 24 25 ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 111 1 Senator Gore: We also have statements for the record 2 from the American Library Association, the Computer Research 3 Association, the Association of Research Libraries, and there 4 are other associations that have also provided statements for 5 the record. Without objection, they will be included. 6 [The material referred to follows:] 7 [SUBCOMMITTEE INSERT] 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 112 1 Senator Gore: Most, if not all -- the GAO was not asked 2 to provide a statement of support or opinion one way or 3 another on this; it is a factual recounting, accounting of 4 some of the questions we addressed to them. These other 5 statements, or most, if not all, are extremely supportive of 6 the legislation. 7 Just to pick up where I left off in my last comments. 8 Dr. Nagel, you talked about making this available to the rest 9 of us, beyond the institutions. I was thinking of the supply 10 and demand forces that will be unleashed when I heard your 11 statement there, and I wanted to refer back to that. 12 Dr. Kalos, I enjoyed the videos that you showed there. 13 How many industrial partners use the Cornell facility? 14 Dr. Kalos: We have about 15 industrial partners. 15 Senator Gore: Are they concentrated in a few industries? 16 Dr. Kalos: No, they span many industries. I should also 17 mention that our major industrial partner is IBM, which is a 18 well-known manufacturer of computers of all kinds. 19 Senator Gore: I have heard of them. 20 [Laughter.] 21 Dr. Kalos: And when IBM decided to reenter the high- 22 performance computing arena, they did it in partnership with 23 Cornell University. We have been pioneers with IBM in 24 conceiving, testing, shaking down certain aspects of their 25 supercomputing, and especially their parallel computing ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 113 1 effort. And that is a partnership that will continue as IBM 2 enters the highly parallel computing arena. 3 Senator Gore: Now, I understand that Cornell runs a 4 program called Superquest to give high school students access 5 to supercomputers. How does that work? 6 Dr. Kalos: Well, we run a national competition; we 7 announce to schools around the Nation that this program is 8 available. Teams at high schools submit ideas for original 9 research that requires supercomputing for its accomplishment. 10 The proposals are evaluating by a group of independent 11 reviewers, and the winners come to Cornell. Their prize is 2 12 weeks in Ithaca -- 13 [Laughter.] 14 Dr. Kalos: -- where they are provided with pizza, 15 softball, and access to high-performance computing, among 16 other essential parts of life. 17 [Laughter.] 18 Dr. Kalos: And the winners are chosen on the basis of 19 the originality of the research. Some have done videos, and 20 so on. I myself had the privilege of introducing the two 21 winners in Gainesville, Florida, and each gave a research talk 22 about research that I found interesting and original. 23 One had to do with surgical treatment for strabismus, 24 crossed eyes. And I thought it was quite original. And 25 another, in fact, was research in my own area, which is ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800). FOR DEPO 114 1 stochastic simulation. And I found that extremely 2 interesting. So I was a little overwhelmed by the quality of 3 these students. And this program continues, and in fact is 4 being broadened this year with the participation of other of 5 the national centers. 6 Senator Gore: Am I wrong that relatively few U.S. 7 companies seem to be using supercomputers compared to what the 8 potential would appear to be? 9 Dr. Kalos: Compared to the potential, yes, I think that 10 is absolutely correct. 11 Senator Gore: Why is that? 12 Dr. Kalos: Well, I think that is, there are a number of 13 issues; perhaps our colleagues from Lilly could speak better 14 to this. I think first of all, the role of computational 15 modeling in science is a relatively new development. The 16 recognition by scientists is complementary to what they have 17 learned -- 18 Senator Gore: Inductive and deductive reasoning. 19 Dr. Kalos: Exactly so. The idea that this is another 20 way of doing science that sheds valuable information, it is a 21 way of connecting to the knowledge they already have. This is 22 rélatively new. In addition, of course, new techniques have 23 to be learned. 24 First of all, the basic techniques of mathematical 25 representation of the problems at hand. the translation of ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 115 1 that into correct and efficient computer algorithms, the 2 realization and testing on computers of all kinds, and the 3 realization and testing especially on supercomputers. 4 So these are a number of new challenges that scientists 5 face everywhere, and I think that, as the applications grow 6 throughout the country, as our young scientists are trained in 7 computational scientists almost as a matter of course, that 8 industry will very naturally take this up. 9 Senator Gore: Dr. Wold. 10 Dr. Wold: I can say that, certainly in our case, our 11 entry into supercomputing would have been clearly impossible 12 without the national center at the University of Illinois. 13 That was our introduction. We felt it was quite a leap of 14 faith to get involved even to that level; I had not even 15 considered at that time purchasing our own supercomputer. 16 Our usage of supercomputing time at the national center, 17 as well as our own usage after we finally got our 18 supercomputer, has just in every case exceeded our 19 expectations; in fact, it exceeded our ability to plan for it. 20 So once the tool was there, the utilization just 21 increased dramatically. The key is to get that first 22 opportunity into any researcher's hands to see what can be 23 done. 24 Senator Gore: Do you work with Larry Smarr at Champaign- 25 Urbana? ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 116 1 Dr. Wold: We certainly do, yes. 2 Senator Gore: Now, geographically, you are about, what, 3 50 to 100 miles from there? 4 Dr. Wold: It is about 120 miles, yes. 5 Senator Gore: 120 miles. Do you have to go to his 6 center still? Or do you have a link? 7 Dr. Wold: We do have a link. But perhaps I could have 8 Dr. Abdulla address that, since he drove that 140 miles many 9 times. 10 Dr. Abdulla: Senator Gore, we have, and continue to have 11 an important program at the National Center for Supercomputing 12 Applications. When we started off, obviously our people had 13 to go to the center where they learned all of the different 14 kinds of technologies that Larry Smarr had over there. There 15 was a tremendous commitment to new hardware, to work stations, 16 to networks, to software, and to new algorithms. 17 We had educational programs that are ongoing, even to 18 this day. We call them FOCUS; obviously we focus on chemistry 19 and the -- it stands for Focus on Chemistry Using 20 Supercomputers. So we expose, in very tight-knit workshops, 21 the .latest in supercomputing technology. Our staff goes 22 through all these training sessions, comes back to an 23 environment which is very similar to what we have at the NCSA, 24 and continues in collaboration with consultants that we have 25 in our shop, and expand and grow. ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 117 1 And you know, if you remember Jack Walton's talk in 2 Washington, D.C., when he talked about the stages of change, 3 one of the things he said was that until somebody establishes 4 a gate post, there is not going to be any diffusion. 5 So we are at that state now; we have establishes gate 6 posts. And people are saying, gosh, if it works for 7 scientists, let me try it with my problem. And then it 8 diffuses, until finally you bring about a complete revolution 9 in the way you do things. So it is a process. 10 Senator Gore: In parallel. 11 Dr. Abdulla: Yes. It is a process rather than an event. 12 And it is happening because of all of these infrastructure- 13 related items that we talked about. 14 Senator Gore: Am I wrong that Japanese companies seem to 15 be relatively more willing to explore the potential of 16 supercomputing? 17 Dr. Nagel: Well, I was going to comment. I think -- and 18 I do not necessarily have the data to support this assertion 19 -- but I think that what, one of the things that you will find 20 is that the use of high-performance computing in industry is 21 jointly, is really a function of how competitive the industry 22 is, or how competitive the people in the industry are. 23 And one of the things that we know about the Japanese is 24 that they are very, very competitive, and effectively so. So 25 I think even in the U.S., you will find those industries which ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 118 1 are the most competitive, and are, you know, frantically 2 searching for ways of getting a sustainable competitive 3 advantage, will be using advanced techniques like 4 supercomputers and high-performance computing networks and SO 5 forth, because they will give that advantage and they will, 6 you know, make that initial threshold jump to get over the 7 difficulties. 8 Senator Gore: I think the testimony of a witness from 9 Cray last year indicated that one-third of U.S. supercomputers 10 are in industry; two-thirds of Japanese supercomputers are in 11 industry. That does not clash with the impressions that you 12 all have, does it? All right. 13 Mr. Gray: Mr. Chairman, could I elaborate on that point 14 just for a moment? 15 Senator Gore: Please. 16 Mr. Gray: We happen to be well aware, through all of our 17 international negotiations and dealings, that your Japanese 18 have got very ambitious programs for upgrading their networks 19 for deploying fiber in a very ubiquitous fashion. 20 They are currently purchasing advanced technological 21 switching capability that would support multi-gigabyte 22 networks and this kind of thing, and have every evidence they 23 will be deploying this technology and is capable of, within 24 the next 2 to 3 years. 25 Senator Gore: Well, there is no question about that. We ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800). FOR DEPO 119 1 have by most estimates about an 18-month lead over the 2 Japanese in network technology. But if we choose not to 3 exploit it, we will lose it in about 18 months because they 4 are not standing still at all. 5 Dr. Nagel: I would like to make, if I might, just one 6 more comment on this business of networking, going beyond just 7 institutional support. I think one of the sometimes 8 unappreciated consequences of setting a goal that we are going 9 to move beyond sophisticated users, as Dr. Langenberg 10 mentioned you find in the universities, to people that are not 11 necessary sophisticated in the use of networking and 12 computers, is that it will force us to make them easier to 13 use, and therefore the barriers to use in industry and in 14 education and everywhere else, will be, you know, reduced 15 greatly. 16 And that is really one of the, you know, it is still a 17 fairly arcane business to use high-performance computing and 18 to use our Internet and the various range of networks that we 19 have available in this country. 20 Senator Gore: And not only with the user interface, as 21 they call it, become a lot friendlier, as they say, the cost 22 is coming down dramatically. One estimate given to the 23 subcommittee -- I do not know whether it is accurate or not 24 -- but the estimate was that a supercomputer which costs 25 between $10 million and $20 million today will almost ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 120 1 certainly with 5 years be in the $400,000 to $500,000 range. 2 If that is the case, then, and if simultaneously the ease 3 of use-improves dramatically, we will see a sudden sharpening 4 of this conflict between data processing capability, demand 5 for data processing, all on the one hand, and our ability as a 6 nation to communicate over our existing communications lines, 7 the information visualizations and packages of data that we 8 need to convey in order to communicate with each other. 9 Dr. Kalos? 10 Dr. Kalos: Mr. Chairman, I would like to amplify 11 somewhat on the issue of the contribution, in particular of 12 high-performance networking to industrial productivity. I 13 would like to call your attention to a joint program of the 14 Xerox Corporation and Cornell, Cornell Theory Center, which 15 they call the Xerox Design Research Institute. 16 That is a very broad-based program which is concerned 17 with what it takes to bring better products more quickly to 18 market. And it concerns simulation and modeling on the 19 supercomputers, but much more than that. For example, one of 20 the issues is the product history and how one learns from 21 previous products how to design better products. 22 Another issue is that Xerox, like many other companies, 23 is spread all over the country. And for people in Park, in 24 Palo Alto, to collaborate with people in Webster, near 25 Rochester, to collaborate with people in their laboratory in ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 121 1 Tarrytown, New York, and design better products that are more 2 manufacturable, that are more maintainable requires a 3 collaboration at great length. 4 And they are very concerned about their ability to 5 communicate, not only the results of supercomputer 6 calculations, but the results of many other ideas and records. 7 I consider that also important, and I believe that the present 8 bill will contribute to productivity very much. 9 Senator Gore: In what Bill Wolf again has called a "CO- 10 laboratory." 11 Dr. Kalos: Exactly. 12 Senator Gore: And I might just note for the record, 13 while we are talking about industry, that until quite 14 recently, Toyota had more supercomputers than Ford, GM and 15 Chrysler combined. That has now barely changed, but it is 16 significant. 17 Dr. Wold, could you name any products for which, that 18 Lilly has been able to develop that might not have been 19 possible in this time frame without supercomputing? 20 Dr. Wold: Well, I certainly wish I could. But our 21 business is a very long-term business in terms of research. 22 It takes 10 years after the discovery. The supercomputer 23 impacts the discovery phase of research, SO if and when we 24 have a compound, a new drug that can be linked to the 25 supercomputer, it is a number of years in the future. ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 122 1 I feel, however, though, that when that day does come, we 2 probably will not remember or notice that it was discovered by 3 the supercomputer, that computational sciences will have been 4 woven into the fabric of research such that it will no longer 5 be remarkable such that we would notice that. 6 Senator Gore: I have a number of questions which I may 7 have to ask for the record because we are running out of time. 8 But Mr. Gray, one of the goals of S. 272 is to provide a 9 catalyst for development of the extensions of the network by 10 the private sector, as I was saying earlier. 11 We want the technology developed under this bill put to 12 use by commercial network providers so that every office and 13 home will have access to the information resources available 14 on the NREN. It is my view that because of the unique nature 15 of this challenge, namely, adding to fiber already in place 16 new switches and software and algorithms which will vastly 17 upgrade the capacity of the fiber. 18 And since other fiber not dedicated to this national 19 network is also there, the discovery of the new switches, et 20 cetera, will present the possibility of making them quickly 21 available for private fiber, so that the network can be very, 22 very quickly expanded. It will also serve as a sort of 23 national demonstration project, showing what is possible with 24 a national gigabyte network, and providing that a commercial 25 market exists for such services. ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 123 1 One thing that might clearly hinder development of 2 commercial, high-speed networks would be if the federal 3 government ran the NREN in a way that competed directly with 4 the private sector, and set up an unnecessary conflict. 5 S. 272 states clearly that the NREN will be phased into, 6 and I am quoting here from the bill, "will be phased into 7 commercial operation as commercial networks can meet the 8 networking needs of American researchers and educators. 9 Is that consistent with your vision of where the NREN 10 plan ought to be headed? 11 Mr. Gray: Well, that is not, not totally clear to us. 12 The objectives as articulated in the legislation are clear; 13 what is a little unclear at the moment is how we, how that 14 path will evolve and how that plan will play itself out. 15 And the point I raise is that we would like to feel 16 certain we have a role to play in our participation to ensure 17 that the initial plan on the road map that is ultimately put 18 in place and followed will assure some evolution or some, at 19 least not exclude public network opportunities to support and 20 provide those services. 21 Senator Gore: Well, I raise this question now, so as to 22 reassert for the record of this hearing, as I have in other 23 hearings, the clear intent of the sponsors of the bill and the 24 advocates of the whole project, to make that work. 25 And even, we, like you, do not yet know how to dot every ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 124 1 I or cross very T. I want you to know that that is clearly 2 our intention and that is the way it is going to happen. 3 NSF seems now to be able to let the private sector 4 provide networking services when that makes sense. They have 5 contracted with your company, with companies like MCI and ANS 6 and regional networks to run NSFNET. Do you think that NSF is 7 presently taking the right approach? 8 Mr. Gray: From what we see and what our experience has 9 been, yes. 10 Senator Gore: Okay. Well, at least we have a model to 11 work with. It may need refinement; it may need modification. 12 But we have the intent, we have the model, we have the working 13 relationship with the companies involved. So that is a good 14 place to start. 15 Suppose we did not pass S. 272. Suppose the money was 16 not appropriated. Suppose there was no Federal leadership in 17 gigabyte networking. How long would it take for the private 18 sector to start providing gigabyte networking services on its 19 own? 20 Mr. Gray: Well, I could only speculate on that. I would 21 certainly think it is in the 5-year time frame or beyond. And 22 I think the more serious issue is whether or not the process 23 would optimize the capability. I think what you are 24 suggesting there would probably create a scenario where there 25 would be test bids developed or individual applications ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 125 1 develop one by one, and directed and designed and participated 2 in by a variety of participants. 3 And while you were setting down, you would probably have 4 a piecework of applications and standards, and you would have 5 a number of potential network services out there that could 6 possibly never inter-operate with one another. I think that 7 is the bigger danger over time. 8 Senator Gore: Can you estimate what the U.S. would use 9 if we do not build a gigabyte network? Or if we do not have 10 one, that can be used in a coherent fashion? 11 Mr. Gray: It would be kind of hard to estimate something 12 like that. I would not even want to take a stab at that after 13 te kind of figures that Dr. Bromley threw around. 14 [Laughter.] 15 Senator Gore: Yeah. We could just say, maybe we could 16 just agree that it would be a lot. I really wish that we had 17 more time to explore each of these questions with follow-ups. 18 I was looking for this number earlier about the increase in 19 the amount of traffic growing in NSFNET. It has been growing 20 at between 20 percent and 30 percent per month. That is a 21 pretty rapid rate of increase. 22 Voice: Have you calculated that out to an annual 23 increase? 24 Senator Gore: About 600 percent per year. That is a 25 pretty good growth rate, I would say. And with the new ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 126 1 information services, of course, it would expand dramatically 2 beyond that. 3 Because of the pressures of time, I am going to ask each 4 one of you if you would be willing to respond to further 5 questions in writing. I appreciate that very much. We are 6 going to proceed expeditiously this year. 7 [The information referred to follows;] 8 [SUBCOMMITTEE INSERT. ] 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202) 289-2260 (800) FOR DEPO 127 1 Senator Gore: I want to express my gratitude to Chairman 2 Hollings of the full committee and Senator Danforth for their 3 support and encouragement on this whole matter. And I 4 appreciate all our witnesses here today. And I think I speak 5 for most of my colleagues on the committee in saying that -- 6 most, if not all, because it passed unanimously last time -- 7 in saying that we are going to move expeditiously and get this 8 done. We appreciate your help today. Thank you. 9 [Whereupon, at 5:05 p.m., the hearing was adjourned.] 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 ALDERSON REPORTING COMPANY, INC. 1111 FOURTEENTH STREET, N.W. SUITE 400 WASHINGTON, D.C. 20005 (202)289-2260 (800) FOR DEPO Enclosure D High Performance Computing and Communications FY 1992 Budget (dollars in millions) 1991 1992 1993 1894 1995 1996 DARPA BA 183.0 232.2 263 553 399 447 OL 157.5 210.1 258 320 374 423 DOE BA 65.0 93.0 110 138 157 OL 169 54.6 78.7 97 121 143 153 EPA BA 1.4 5.2 5 5 5 OL 5 1.0 2.7 4 4 4 4 NASA BA 64.0 72.4 107 134 151 OL 145 49.1 62.9 89 118 140 146 HHS/NLM BA 13.5 17.1 17 17 17 OL 17 12.0 15.2 17 17 17 17 NIST BA 2.1 2.9 3 3 3 OL 3 1.8 2.7 3 3 3 3 NSF BA 169.0 213.0 262 305 354 OL 413 150.4 190.0 234 280 326 380 NOAA BA 1.4 2.5 3 3 3 OL 3 1.2 2.1 2 2 3 3 TOTAL BA 489.4 638.3 789 956 OL 1,087 427.7 1,202 564.6 704 865 1,009 1,129 200 OMB/ESD FAX# 3954817 OR 3953165 13:15 16/50/20 SENT BY:Xerox Telecopier 7021 ; 3-13-91 ; 5:50PM ; 2023953462- 2023951575:# 2 9/20727 GEORGE a BROWN, is., Celifornia, CHAIRMAN ROBERT 8. WALKER, Pennsylvania F. JAMES BENSENBAEHNER je., Wissonsin SHERWOOD L BOEHLERT. New York JAMES M. SCHEUER, New York TOM LEWIS, Rorida MARILYN LLOYD, Tennessee DON AITTER, Pennsylvania DAN GLICKMAN. Kansas SID MORRISON, Washington HAROLD L VOLKMER, Missouri NON PACKARD. California HOWARD WOLFS, Mianigan RALPH M. HALL, Texas U.S. HOUSE OF REPRESENTATIVESRECEIVE "RAUL a. HENRY, Michigan HARRIS W. FAWELL I DAVE MOCURDY. Oklahoma D. FRENCH BLAUGHTER Jan Virginis NORMAN Y. MINITA, California COMMITTEE ON SCIENCE, SPACE, LAMAR SMITH. Taxas TIM VALENTINE, North Cerolina CONSTANCE A. MORELLA, Meryiend ROBERT G. TOANSELLI. New Jersey RICK BOUCHER, Virginia AND TECHNOLOGY, DANA ROHRABACHER, California STEVEN M. SCHIFF, New Mexican TERRY L BRUCE Illinois RICHARD H. STALLINGS, leabe 91 MAR 12 P4: TOM CAMPBELL Celifornia OHE J. RHODES. III. Artzone JAMES A. TRAFIGANT, Jr., Onio SUITE 2320 RAYBURN HOUSE OFFICE BUILDING JOE BARTON. Texas HENRY J. NOWAK, New Yerk DICK ZIMMER, New Jersey CARL c. PERKINS. Kentucky TOM MOMILLEN. Maryland WASHINGTON, DC 20515 WAYNE T. GILCHAEST, Maryland DAVID R. NAGLE lows JIMMY HAYES. Louisians (202) 225-8371 OFFICE OF THE RADFORD BYERLY, Jr. JERRY R COSTELLO, Illinois JOHN TANNER, Tennessee DIRECTOR Chief of Staff MICHAEL RODEMEVER GLEN BROWDER Alabama PETE GEREN. Texas Chief Counsel RAY THORNTON, Arbansas DAVID a CLEMENT JIM BACCHUS, Plorida Republican Chief of Staff TIM ROEMER, Indians aud CRAMER. Alebama DICK SWETT, New Harnpainire March 11, 1991 MICHAEL J. KOPETBEL, Oregon JOAN KELLY HORN, Missouri BARBARA-ROSE COLLINS. Mishigan The Honorable D. Allan Bromley Assistant to the President for Science and Technology New Executive Office Building Washington, D.C. 20506 Dear Dr. Bromley: Thank you for your testimony last week outlining the President's High Performance Computing and Communications Program and your views on H.R. 656. I appreciate your contribution to the discussion of the future of high performance computing in this country. Several questions remain, however, on the Administration's position on this issue. I have attached several additional questions for your consideration. Your expeditious reply will greatly assist us in incorporating your views into relevant legislation considered by the Committee. Thank you for your attention to this matter. Bals Cordially, Walk Robert S. Walker Republican Chairman RSW/am Enclosure SENT BY:Xerox Telecopier 7021 ; 3-13-91 ; 5:51PM ; 2023953462- 2023951575:# 3 COMMITTEE ON SCIENCE, SPACE, AND TECHNOLOGY U.S. HOUSE OF REPRESENTATIVES WASHINGTON, D.C. 20515 QUESTIONS FOR THE WRITTEN RECORD HEARING ON HIGH PERFORMANCE COMPUTING 1. Please provide the Subcommittee with both the incremental and total run-outs for each of the agencies participating in the President's High Performance Computing and Communications Program (DOE, NASA, NSF, NIST, NOAA, EPA, NIH, and DARPA) over each of the next five years. 2. Does the President's initiative envision the government buying and owning supercomputers, high end switches, fiber optical cable, and other hardware? Is there any reason why the network cannot be established commercially from its inception? Is it necessary for the federal government to do more than fund research and development and contract for services from the network? 3. What specific and detailed changes would you like to