Dan Reed
[email protected]
University of North Carolina at Chapel Hill
Jack Dongarra
[email protected]
University of Tennessee
and Oak Ridge National Laboratory
Ken Kennedy
[email protected]
Rice University
———-
Inscribed on the wall of the U.S. House of Representatives Science committee hearing room is a quotation from Tennyson that is an apt metaphor for both the promises offered and the reality delivered by high-end computing, “For I dipped into the Future, far as human eye could see; saw the vision of the world, and all the wonder that would be.” Computational science is the enabler of wonder, and enabled via high-end computing, it has emerged as the third element of the research portfolio, complementing theory and experiment.
Today, despite the unprecedented opportunities and a clear record of success, buttressed by formal reviews, Congressional recommendations and a plethora of studies and reports, the prospects for continued deployment and support of high-end facilities for open scientific research are in more serious doubt than they have been in decades. The NSF supercomputing centers are on transitional life support – two years after the cyberinfrastructure report was released, no successor to the Partnerships for Advanced Computational Infrastructure (PACI) program has been announced, and the centers have little funding for substantive hardware upgrades. Similarly, funding for DOE's open high-end computing vision has yet to be appropriated. Death can come via indecision just as surely as by inappropriate choices. The path is clear; we must act to preserve our national high-performance computing assets and competitive advantage.
In the 1980s, the National Science Foundation, based on community urging, created the NSF supercomputing centers. Via this program and its PACI successor, the NSF centers have provided community access to high-end computers that previously had been available only to researchers at national laboratories. To broaden participation and access, DOE also opened a portion of its high-end computing facilities for national use, notably at NERSC and ORNL.
Today, opportunities abound for application of high-performance computing in both science and industrial sectors. Integrated vehicle designs with lifetime warranties, based on coupled electrical, mechanical and power train models, are within reach. Higher resolution cosmological models would allow testing of competing theories of the evolution of the universe, with sufficient resolution to simulate galaxy formation. Personalized medicines, tailored to minimize toxicity and maximize efficacy based on individual genetics, are possible based on drug chemistry models. All require a new generation of high-performance computing systems that can deliver high sustained performance for coupled models.
Despite these opportunities, our current national investment strategy for high performance computing seems to have wandered from the path that has brought so much success. In an era of constrained budgets, there is a great temptation to view distributed infrastructure (Grids, cyberinfrastructure and other elements) as an alternative successor to high-end computing. This view is divisive and inimical to leading edge science. High-end computing was the midwife to cyberinfrastructure, and it continues to support and nurture distributed scientific communities. High-end computing and cyberinfrastructure are both part of a broad and empowering vision of computing-enabled science; it is not an “either or” situation. Both are central, both are critical, and neither can be sacrificed for the other.
Last year, when Dan Reed testified to the House Science Committee on the future of high-end computing, he made several points. What he told the committee remains apt, though our implementation window continues to shrink. We believe an interagency initiative in high-performance computing should be based on the following principles:
- An integrated, long term, multi-decade, strategic roadmap that articulates the responsibilities, scope and financial scale of each agency's responsibilities.
- Regular deployment and support of the world's highest performance computing facilities for open scientific use, as part of a broad ecosystem of supporting infrastructure, including high-speed networks, large-scale data archives, scientific instruments and enabling software.
- Coordination and support for national priorities in science, engineering, national security and economic competitiveness.
- Vendor engagement to ensure technology transfer and economic leverage
- Verifiable metrics of interagency collaboration, community engagement and technical progress that are tied to agency funding.
- Active recruitment of computational scientists and encouragement of cross-disciplinary collaboration and education.
As a community, we must speak clearly and with unanimity if we are to ensure continued support for high-end computing facilities and the software infrastructure that enables their use; we cannot let the scientific opportunities slip away.
There is a second, more foreboding inscription on the wall of the House Science Committee hearing room, drawn from Proverbs: “Where there is no vision, the people perish.” We hope for the wonder that would be, based on an implemented vision.
DISCLAIMER: These are the authors' personal opinions, and they should not be attributed to any group in which they might be participants.
Please send any comments/questions to HPCwire editor Tim Curns at [email protected].
Dan Reed is the Chancellor's Eminent Professor at the University of North Carolina at Chapel Hill, as well as the Director of the Renaissance Computing Institute (RENCI), which is exploring the interactions of computing technology with the sciences, arts and humanities. He is the author of over one hundred research papers and monographs on algorithms, architectures, and performance evaluation techniques for high-performance computing. Professor Reed also serves as Vice-Chancellor for Information Technology for the University of North Carolina at Chapel Hill. Professor Reed was previously Director of the National Center for Supercomputing Applications (NCSA) at the University of Illinois at Urbana-Champaign, where he also led National Computational Science Alliance, a consortium of roughly fifty academic institutions and national laboratories that is developing next-generation software infrastructure of scientific computing. He was also one of the principal investigators and chief architect for the NSF TeraGrid. He was chair of the community workshop for the High-end Computing Revitalization Task Force (HECRTF). He is a Fellow of the ACM and the IEEE.
Jack Dongarra is a University Distinguished Professor of Computer Science at the University of Tennessee and has the position of a Distinguished Research Staff member in the Computer Science and Mathematics Division at Oak Ridge National Laboratory. His research includes numerical algorithms in linear algebra, parallel computing, the use of advanced-computer architectures, programming methodology, and tools for parallel computers. He has contributed to the design and implementation of the following open source software packages and systems: EISPACK, LINPACK, the BLAS, LAPACK, ScaLAPACK, Netlib, PVM, MPI, NetSolve, Top500, ATLAS, and PAPI. He has published approximately 200 articles, papers, reports and technical memoranda and he is coauthor of several books. He was awarded the IEEE Sid Fernbach Award in 2004 for his contributions in the application of high performance computers using innovative approaches. He is a Fellow of the AAAS, ACM, and the IEEE and a member of the National Academy of Engineering.
Ken Kennedy is the John and Ann Doerr University Professor of Computer Science and Director of the Center for High Performance Software Research (HiPerSoft) at Rice University. He has supervised thirty-six Ph.D. dissertations and published two books and over two hundred technical articles on compilers and programming support software for high-performance computer systems. In recognition of his contributions to software for high performance computation, he received the 1995 W. Wallace McDowell Award, the highest research award of the IEEE Computer Society. In 1999, he was named the third recipient of the ACM SIGPLAN Programming Languages Achievement Award. He is a Fellow of the AAAS, ACM, and the IEEE and a member of the National Academy of Engineering. From 1997 to 1999, he served as co-chair of the President's Information Technology Advisory Committee (PITAC).