SC2000 STATE OF THE FIELD TALKS


STATE OF THE FIELD TALKS
The State-of-the-Field talks give conference attendees the opportunity to hear distinguished leaders in four different fields of computer and computational sciences conduct brief tours of the state-of-the-art in these fields and explore possible ways in which the field will change in the coming years.

All four talks will be in Ballroom C


STATE OF THE FIELD CHAIR
PAT TELLER, UNIVERSITY OF TEXAS, EL PASO


WEDNESDAY NOVEMBER 8
Wednesday, 8:30 AM
Ballroom C Webcasts

Chair: James R. McGraw, Lawrence Livermore National Laboratory

COTS Cluster Systems for High-Performance Computing

Dr. Thomas Sterling
NASA Jet Propulsion Laboratory, High Performance Computing Group, and California Institute of Technology, Center for Advanced Computing Research

Note: In addition to being webcast, this presentation includes a COTS clusters slide show that will be presented simultaneously with the webcast. That slide show has been reproduced here in HTML format.

In recent years an alternative strategy to achieving high performance, which overcomes the combined problems of cost and architecture variability as well as stability of single suppliers, has emerged. A product of nearly a decade of applied research on workstation clusters, PC clusters, and non-dedicated LAN-connected user desktop and server facilities for cycle harvesting has yielded a rapidly maturing methodology for aggregating and employing low-to-moderate range computer systems in distributed complexes for both capacity and capability workload processing requirements.

COTS clusters are now having significant impact on the realm of processing once reserved to supercomputing. But not all applications are suitable for such loosely-coupled ensembles, and system software environments are still in evolution.

In this talk, Dr. Sterling will explore the history, methodologies, capabilities, and limitations of COTS clusters. He will examine in detail hardware component capabilities and configuration. Sterling will describe software systems and tools for cluster system programming and management, and will present performance and scaling data on successful applications as well as those demonstrating poorer suitability. Finally, Dr. Sterling will discuss examples of near-term research and technology trends that are likely to determine the future directions and capabilities of the next generation of COTS clusters for high-performance computing.

Biographical Sketch: Dr. Sterling holds a joint appointment with NASA's Jet Propulsion Laboratory (JPL) and the California Institute of Technology (CalTech), serving as Principle Scientist in JPL's High Performance Computing group and Faculty Associate in CalTech's Center for Advanced Computing Research. He received his Ph.D. as a Hertz Fellow from MIT in 1984.

For the last 20 years, Sterling has engaged in applied research in parallel processing hardware and software systems for high-performance computing. He was a developer of the Concert shared-memory multiprocessor, the YARC static dataflow computer, and the Associative Template Dataflow computer concept, and has conducted extensive studies of distributed shared-memory cache-coherent systems. In 1994, Sterling led the NASA Goddard Space Flight Center team that developed the first Beowulf-class PC clusters. Since 1994, he has been a leader in the national Petaflops initiative. He is the Principal Investigator for the interdisciplinary Hybrid Technology Multithreaded (HTMT) architecture research project sponsored by NASA, NSA, NSF, and DARPA, which involves a collaboration of more than a dozen cooperating research institutions. Dr. Sterling holds six patents, and was one of the winners of the 1997 Gordon Bell Prize for Price/Performance.

Wednesday, 9:15 AM
Ballroom C Webcasts

A Small Dose of INFOSEC

Dr. Eugene H. Spafford
Purdue University, Department of Computer Sciences

Melissa. ILOVEYOU. Web page defacement. Denial-of-service against e-commerce sites. Theft of laptops from the State Department. Information Warfare. Y2K. Napster. Hacking.

It is no longer possible to avoid stories of information loss, fraud, and compromise. Open any paper or magazine, or listen to the news (on-line as well as in the standard media) and stories relating to information security are sure to be present.

So, what is the current state of information security? Are things getting better, or are they getting worse? And what are the challenges that we are likely to see in the near future?

In this talk, Dr. Spafford will present some highlights of what is happening in information security, and what is yet to happen. He also will include some discussion of the nature of infosec-related challenges that we are likely to face - and few of them are based solely in technology.

Biographical Sketch: Dr. Spafford is a professor of Computer Sciences at Purdue University, the university's Information Systems Security Officer, and Director of the Center for Education Research Information Assurance and Security (CERIAS). CERIAS is a campus-wide multi-disciplinary center, with a broadly-focused mission to explore issues related to protecting information and information resources. Spafford has written extensively about information security, software engineering, and professional ethics. He has published over 100 research articles and reports, and has written or contributed to over a dozen books; he serves on the editorial boards of most major infosec-related journals.

Spafford is a Fellow of the ACM, Fellow of the AAAS, senior member of the IEEE and is a charter recipient of the Computer Society's Golden Core award. Among other activities, he is chair of the ACM's US Public Policy Committee, a member of the Board of Directors of the Computing Research Association, and is a member of the US Air Force Scientific Advisory Board. Dr. Spafford regularly serves as a consultant on information security and computer crime to law firms, major corporations, U.S. government agencies, and state and national law enforcement agencies around the world.


THURSDAY NOVEMBER 9
Thursday, 8:30 AM
Ballroom C Webcasts

Chair: Pat Teller, University of Texas, El Paso

Numbers, Lots of Numbers, and Insight Too: Scientific Computing 2000

Margaret H. Wright
Bell Laboratories, Lucent Technologies

Hamming's famous 1962 dictum, "The purpose of computing is insight, not numbers," has been quoted countless times, and no one doubts its truth. But most insights derived from computing depend on numbers - numbers being produced today at speeds and in quantities that were inconceivable in 1962. And certain crucial numbers, such as sensitivities and uncertainties, allow us to associate precision and measure with otherwise fuzzy insights.

In this talk, Dr. Wright will sketch a few directions in scientific computing that are helping us to obtain, as quickly and reliably as possible, numbers that are as accurate as necessary, to solve as many important problems as possible, all with the ultimate goal of insight and understanding.

Biographical Sketch: Dr. Wright is head of the Scientific Computing Research Department in the Computing Sciences Research Center, Bell Laboratories, Murray Hill, New Jersey. She received a Ph.D. and M.S. in computer science and a B.S. in mathematics from Stanford University. From 1976-1988, she worked in the Department of Operations Research, Stanford University, and in 1988 she joined Bell Labs. Wright's research interests are numerical optimization, linear algebra, mathematical software, scientific computing, and the application of optimization to real-world problems. During 1995 and 1996, she served as president of the 9,000-member Society for Industrial and Applied Mathematics (SIAM) and in 1997 was elected to the US National Academy of Engineering. Dr. Wright currently chairs the Department of Energy's Advanced Scientific Computing Advisory Committee.

Thursday, 9:15 AM
Ballroom C Webcasts

Parallel/Distributed Programming: Research Success--Application Failure?

J.C. Browne
The University of Texas at Austin, Department of Computer Science

Development of parallel/distributed programs that are correct, efficient, scalable, and portable across parallel/distributed execution environments remains the major bottleneck to effective application of parallel/distributed computational environments. Research in parallel programming has generated innovative concepts and potentially effective models for parallel/distributed programming. Some of these concepts and models have been given experimental implementations. A few have demonstrated significant potential for development of correct and efficient parallel/distributed programs.

Little of this research has had any impact upon the practice of parallel/distributed programming. The emerging technologies for compositional development of software have not yet made a major impact on parallel/distributed programming. Current practice of parallel/distributed programming is still dominated by ad hoc extensions to conventional sequential programming systems. We are still trying to fly into space using a horse and buggy.

In this talk, Dr. Browne will give one perspective on why past research has failed to have a major impact on practice, will propose requirements for broadly applicable effective parallel/distributed software system development, and will assess past and current research with respect to these requirements. He will suggest an approach for bringing parallel/distributed programming into the mainstream of software development and suggest a basis for a quantum leap in the practice of parallel/distributed software system development.

Biographical Sketch: Dr. Browne is Professor of Computer Science and Physics and holds the Regents Chair #2 in Computer Sciences at The University of Texas at Austin (UT-Austin). He earned his Ph.D. in Chemical Physics at UT-Austin in 1960 and taught in the Physics Department at the university from 1960 through 1964. From 1965 through 1968, Browne was Professor of Computer Science and Director of the Computation Laboratory at Queens University in Belfast. In 1968, he rejoined UT-Austin as Professor of Physics and Computer Science. During 1968-69, 1971-75 and 1984-87, he served as Department Chair for Computer Science. Browne is a fellow of the Association for Computing Machinery (ACM), the British Computer Society, the American Physical Society, and the American Association for the Advancement of Science; he has been Chairman of the ACM Special Interest Group on Operating Systems.

Browne has published approximately 100 papers in computational physics and 200 papers in computer science. He has been active in research on parallel/distributed programming and computation for twenty years, developing parallel/distributed programming systems and using parallel/distributed programming systems. Dr. Browne's research on parallel/distributed programming spans from graphical/visual programming environments (CODE) to libraries for parallel implementation of adaptive solutions of partial differential equations (DAGH, SDDA).