TUESDAY, NOVEMBER 13

  • 3:30-5:00pm
    Room A201/205
    Access Grid Enabled

    Supercomputing in a Shoebox: The Convergence of High-Performance Computing (HPC) and High-Performance Embedded Computing (HPEC)
    Moderator: Jeremy Kepner, Massachusetts Institute of Technology, Lincoln Laboratory
    Panelists: Jack Dongarra, Professor, University of Tennessee; Charles Holland, Director of Information Technology, DUSD (S&T); Craig Lund, CTO, Mercury Computers, Inc.; Charles Seitz, CEO & CTO, Myricom, Inc.; Tony Skjellum, President, MPI Software Technology, Inc.; Tom Sterling, California Institute of Technology

    The majority of the world's computing is performed in embedded systems (wireless apps, network routers, set-top boxes, etc.). Larger and larger parallel embedded systems are being deployed and the world of High-Performance Embedded Computing (HPEC) is increasingly dealing with traditional HPC issues (bisection bandwidth, shared vs. distributed memory, scalable software, etc.). The increasing physical size of today's largest HPC systems is problematic for all but a few users. Increasingly the HPC community is dealing with traditional HPEC issues (Mflops/ft3, Mflops/watt, etc.). This panel will introduce these issues and will discuss how they relate to both the HPC and HPEC communities. Questions that will be addressed include:

    • DENSITY: HPEC systems provide a 10x increase in Mflops/ft3 over their HPC counterparts. Is the need for compute density limited to special-purpose applications, or will high-density computing be required for next-generation supercomputers to be feasible?

    • CONVERGENCE: HPC systems are increasingly becoming more focused on density while HPEC systems are becoming more focused on programmability. Will these two regimes converge?

    • GROWTH: Both HPC and HPEC are driven by commercial and DoD markets (servers, medical, wireless, etc.). Will one of these areas become so dominant as to create a "victor"?

WEDNESDAY, NOVEMBER 14

  • 10:30am-noon
    Room A201/205
    Access Grid Enabled

    Computational Biosciences: The Computational Continuum from Human Genomes to Human Health
    Moderator: Raquell M. Holmes, Research Assistant Professor, Center for Computational Science, Boston University Panelists: Thomas Bartol, Computational Neurobiology Laboratory, The Salk Institute for Biological Studies; Charles Delisi, Professor Biomedical Engineering, Boston University; Micah Dembo, Professor Biomedical Engineering, Boston University; Joel Stiles, Senior Scientific Specialist, Computational Neuroscience, Pittsburgh Supercomputing Center

    The success of the human genome project has shot the field of bioinformatics and computational biology into the national limelight. As more and more scientists turn to the frontier of computational biology, it is important to identify the current trends and future computational needs of research across the biological spectrum. This panel brings together computational scientists who have been and are pushing the envelope in the areas of genomics, cellular processes, and human health. Panelists will discuss computational resources and approaches that are transforming the way life sciences are done and understood.

    The panel is directed towards scientists who are interested in better understanding computational challenges in the biosciences. Questions to be addressed include the following:

    • What are the computational approaches used in current research of human genomes, cellular processes, and human health?

    • What computing resources are utilized in these areas?

    • What technological and computing advances have aided in bringing the research to where it is today?

    • What are the computing advances that are needed to improve research in these areas?

  • 3:30-5:00pm
    Room A201/205
    Access Grid Enabled

    Computational Science: Do Undergraduate Faculty Know that Computational Science is the Future?
    Moderator: Scott Lathrop, Program Manager for Education, Outreach and Training, National Center for Supercomputing Applications
    Panelists: Rubin Landau, Professor Physics, Director of Computational Physics Program, Oregon State University; E. Bruce Pitman, Professor Mathematics and Vice-Provost for Educational Technology, State University of New York, Buffalo; Kris Stewart, Professor Computer Science, San Diego State University and Director of NPACI/CSU Education Center on Computational Science and Engineering; Gabriele Wienhausen, Provost for the Sixth College, University of California, San Diego

    Computational Science is making possible significant innovations and break-throughs within academia, industry, and government. These improvements have followed the efforts of schools, colleges, and universities to prepare mathematicians, scientists, and engineers with the skills and insights needed to effectively utilize modern computational and informational technologies. Yet many educators are either unaware of how to integrate computational science into their courses, or feel that the mere use of computers in their teaching and research automatically provides them with the necessary "computer savy."

    This panel is directed to faculty, scientists, technologists, deans, and administrators. We aim to stimulate their interests in the developing resources and opportunities for expanding computational science on their own campuses.

    The proposed panel session will address Computational Science and its integration into undergraduate education. Topics to be discussed will include the following:

    • What is computational science?

    • How do Computational Science programs differ from Informatics and Information Technology programs?

    • What balance between mathematics, computer science, and scientific discipline is necessary to ensure a successful program?

    • What are the challenges in keeping this a strong and vibrant triad?

    • Do we really need to educate students in computational science?

    • Will this benefit careers in business and industry? How will this improve research in industry and academia?

    • Which fields will benefit the most?

    • The perspective of needed skills and knowledge within industry.

    • The impact of programs as experienced by their alumni.

    • Review of exemplary programs.

    • What are the qualities in these programs that make them successful and attractive to students?

    • What are the aspects of programs that have not succeeded and should be avoided?

    • The computational science continuum from K-12, to undergraduate to graduate education.

    • What resources and opportunities (including funding) are available or needed to aid educators in integrating computational science into their courses?

    • How do we ensure that these educational innovations address all fields including the social sciences, and are fair to the developers?

    • Cultural and societal issues as we move into a global IT society.

    • Intellectual property, privacy, multi-cultural, and multi-lingual issues.

THURSDAY, NOVEMBER 15

  • 10:30am - Noon
    Room A201/205
    Access Grid Enabled

    The Access Grid: Where the Vision Meets Reality
    Moderator: Emilee Patrick, Motorola Labs
    Panelists: Crysta Metcalf, Anthropologist, Motorola Research Labs; Don Morton, Associate Professor Computer Science, The University of Montana, Missoula; Rick L. Stevens, Director Mathematics and Computer Science Division, Argonne National Laboratory, and Professor of Computer Science, University of Chicago; Jennifer Teig von Hoffman, Senior Analyst, Boston University

    The Access Grid (AG) is the future of the Internet: a network of multimedia nodes, connected by multicast over high-bandwidth networks. Each AG node is an ensemble of resources enabling human interaction across the Grid, from high-end audio and visual technology to interfaces to grid middleware and visualization environments. In addition, large-format displays, multiple cameras and microphones, and dedicated meeting rooms provide support for real-time communication among entire groups, not just individuals.

    Experiencing the AG can feel as intimate as a private telephone call between two locations, or as impersonal as attending a talk in a large auditorium. This is not your mother's idea of videoconferencing. This panel will take place on the Access Grid. Not only will the session be made available to anyone with the requisite equipment and high-speed connection, but attendees will experience firsthand the nature of interaction over the AG. Four panelists will each present their own vision for the future of communication and collaboration via the Internet, and then engage in a discussion of how this compares with the usage, utility, and usability of the Access Grid as it exists today.

    Specific questions will include, but will not be limited to:

    • Should the AG support large scheduled meetings or smaller, informal gatherings?

    • Is it possible to do both?

    • What is the nature of the AG community today—who are the people that use the AG and how do they use it?

    • What is the most compelling application for the AG in the future?

    • How should it evolve?

    • What are the privacy implications of AG technology everywhere?

    • From the perspective of a novice user, what is the single most important problem that, if it were solved, would bring the biggest improvement to the AG?

    For more information about the Access Grid, see the AG Home Page at http://www-fp.mcs.anl.gov/fl/accessgrid/default.htm

FRIDAY, NOVEMBER 16
  • 8:30-10:00am
    Room A207/209

    HPC Software: Have We Succeeded in Spite of It or Because of It?
    Moderator: John M. Levesque, Senior Technologist, Cray Inc.
    Panelists: Walt Brainerd, Owner, The Fortran Company; Chris Doehlert, President & CEO, Etnus; Michael Gittings, Guest Scientist, Los Alamos National Laboratory, and Assistant Vice President and Chief Scientist, Science Applications International Corporation; Bill Gropp, Senior Computer Scientist and Associate Division Director, Mathematics and Computer Science Division, Argonne National Laboratory; David Kuck, Intel Fellow and Director KAI Software Lab, Intel Corporation; James R. Taft, Technical Director, Advanced Computing Technologies, NASA Ames Research Center

    The moderator believes that we have succeeded in spite of HPC software. The state of HPC software is poor at best and the future of HPC software is dismal. The current thrust towards Linux and open source software does not bode well for software. My questions are targeted at a few specific problem areas.

    First, I believe standards have the most impact on the efficiency of the resulting software. Fortran 90 standards have ruined the efficiency of the Fortran language. On the other hand, the MPI standards group was able to define a set of calls that could utilize the efficiency of the system without many pitfalls.

    A second area of concern is the academic HPC software research. Has this research contributed to principal application developers? There are two different ways such research can benefit the end-user: (1) be implemented within the principal software components developed by the hardware/software vendors or (2) be used directly by the end-user. An important measurement for the panel is the efficiency of a software component on the state-of-the-art hardware. HPF formulated a language that was impossible to implement efficiently and subsequently was never used by the major application developers; MPI, on the other hand, formulated a language (library) that could be implemented efficiently and is the principal tool in the HPC industry today.

    Questions addressed to Bill Gropp and Walt Brainerd on standards:

    • How are features brought to a standards group?

    • When considering a feature, are the efficiencies of implementation considered? For example, if a feature is very difficult or impossible to implement efficiently on a majority of the hardware architectures, is it given serious consideration?

    • What recommendations do you have to improve HPC software standards?

    Questions addressed to David Kuck and Chris Doehlert on software implementation:

    • What are the priorities for the following characteristics of software development?
      (a) adherence to standards

      (b) efficiencies of implementation

      (c) ease of use

    • Please comment on the difficulties of targeting Fortran 90 in your software. How much of your work is based on academic research?

    • Has government funding of numerous universities for HPC software research aided your development?

    Questions addressed to Jim Taft and Michael Gittings on use of HPC software:

    • Considering all software components, prioritize the software that you use to get your job done?

    • What, if any, non-application software that you or your group developed out of necessity, do you feel should have been supplied by a hardware and/or software vendor?

    • Going back over the years, please list what, if any, software you have used which was developed by HPC academic research groups? If the list is small or empty, please give your views on the third question under the software implementers' list.
  • 10:30am-noon
    Room A207/209

    General: Supercomputing's Best and Worst Ideas
    Moderator: H. J. Siegel, Professor, Colorado State University
    Panelists: James C. Browne, Professor Computer Science, University of Texas, Austin; Cherri M. Pancake, Professor, Oregon State University; Guy Robinson, Research Liaison/MPP Specialist, Arctic Region Supercomputing Center, University of Alaska; Charles Seitz, CEO & CTO, Myricom, Inc.; Burton Smith, Chief Scientist, Cray Inc.; Marc Snir, Professor, University of Illinois, Urbana-Champaign

    The title of this panel is, to some extent, self-explanatory-but if that were all we said, the abstract would be too short. Thus, the questions below expand on the issues that the panelists may wish to consider. We are pleased to have a very distinguished group of panelists covering a wide range of aspects of supercomputing. We invite you to join in the discussion with your answers to the following questions, and with your own questions for the panelists and audience.

    Some questions to be put forth at the panel are as follows:

    • What makes an idea a "supercomputing idea"? (You cannot use the word "supercomputer" in your answer)

    • What are the criteria for deciding what is best or worst?

    • What ideas have gone, over time, from best to worst, worst to best, or both?

    • What ideas have stayed around too long?

    • From ideas that were bad, what good kernels have been extracted?

    • From ideas that were good, what bad impacts have occurred?

    • Are some ideas reminiscent of a "Tale of Two Cities"—it is the best of ideas and the worst of ideas?

    • Is one person's best idea choice another person's worst idea choice?

    • What best and worst ideas will hardware, software, and applications people agree to?

    • What ideas change from best to worst (or visa versa) if your perspective changes among hardware, software, and applications?

    • What best ideas have been stifled due to industry (i.e., economic) factors?

    • How much should wideness of applicability be used to judge the "bestness" of an idea?

    • What has been the impact of marketing (i.e., mass appeal) on the ability to develop an idea to find out if it is best or worst, or does this not affect supercomputers?

    • What has been the impact of government initiatives on what can become a best idea?

    • When have standards been enablers/stiflers of best ideas?

    • When have standards been enablers/stiflers of worst ideas?

    • Can we use a knowledge and understanding of supercomputing's best and worst ideas of the past to develop new best ideas and avoid new worst ideas in the future?