Advanced Scientific Computing Research
Associate Director of the Office of Science for Advanced Scientific Computing Research
SC-30/Germantown Building
U.S. Department of Energy
1000 Independence Ave., S.W.
Washington, D.C. 20585-1290
http://www.sc.doe.gov/ascr/home.html
General Goal 5, World Class Scientific Research Capacity: Provide world-class scientific research capacity needed to: ensure the success of Department missions in national and energy security; advance the frontiers of knowledge in physical sciences and areas of biological, medical, environmental, and computational sciences, or provide world-class research facilities for the Nation’s science enterprise.
Computer-based simulation enables us to predict the behavior of complex systems that are beyond the reach of our most powerful experimental probes or our most sophisticated theories. Computational modeling has greatly advanced our understanding of fundamental processes of Nature, such as fluid flow and turbulence or molecular structure and reactivity. Through modeling and simulation, we will be able to explore the interior of stars and learn how protein machines work inside living cells. We can design novel catalysts and high-efficiency engines. Computational science is increasingly central to progress at the frontiers of almost every scientific discipline and to our most challenging feats of engineering.
In the past two decades leadership in scientific computation has become a cornerstone of the Department’s strategy to ensure the security of the nation and succeed in its science, energy, environmental quality, and national security missions. According to a number of authorities, ranging from the President’s Science Advisor and the President’s Council of Advisors on Science and Technology to the National Research Council and the Council on Competitiveness, this scientific leadership is critical to the economic health of the nation.
Mission: The mission of the Advanced Scientific Computing Research (ASCR) program is to deliver forefront computational and networking capabilities to scientists nationwide that enable them to extend the frontiers of science, answering critical questions that range from the function of living cells to the power of fusion energy.
Program Goal 05.23.00.00: Deliver forefront computational and networking capabilities to scientists nationwide that enable them to extend the frontiers of science, answering critical questions that range from the function of living cells to the power of fusion energy.
5.0 Objectives and Performance Targets:
At the core of this program, and underpinning all of our objectives, is a fundamental quest for knowledge. Our program history provides a compelling story of how computation and the resulting impacts in the pursuit of science have shaped the world around us. The future appears even more promising as we strive to find answers to the following questions:
What new mathematics are required to effectively model systems such as the Earth’s climate or the behavior of living cells that involve processes taking place on vastly different time and/or length scales?
Which computational architectures and platforms will deliver the most benefit for the science of today and the science of the future?
What advances in computer science and algorithms are needed to increase the efficiency with which supercomputers solve problems for the Office of Science?
What operating systems, data management, analysis, model development, and other tools are required to make effective use of future-generation supercomputers?
Is it possible to overcome the geographical distances that often hinder science by making all scientific resources readily available to scientists, regardless of whether they are at a university, national laboratory, or industrial setting?
Below are the main objectives for Advanced Scientific Computing Research. Complementing this stand-alone Program Plan, and providing additional details of our program objectives are the Office of Science Strategic Plan (February 2004), the Facilities for the Future of Science: A Twenty Year Outlook (November 2003), as well as the most recent Office of Science budget.
Objectives:
Scientific Discovery: Broadly enable scientific discovery and the understanding of complex systems with crosscutting research in computer science, research in applied mathematics, and corresponding development of advanced fundamental algorithms and software.
Scientific Simulation: For a critical, select set of science challenges, extend the frontiers of scientific simulation with a new generation of tailored computational models that fully exploit the power of advanced computers… and the power of collaboratory software, making scientific resources available to scientists anywhere, anytime.
Computing at the Petascale: Deliver the key hardware, including tailored supercomputing architectures for science, computing resources at the petascale and beyond, and network infrastructure that will enable a new chapter of U.S. scientific investigation and collaboration.
An accompanying timeline (Road Maps) provides a roadmap for these objectives, including our planned future facilities, performance targets, and the primary connections and program interdependencies. Two important caveats, described below, must be observed when viewing the timeline.
The Objectives, Performance Targets and schedules identified on the timeline are for planning purposes only and do not constitute financial or contractual commitments by the Federal government. More often than not, there are significant discrepancies between planning levels and subsequent, enacted budgets. It is reasonable to anticipate that resources may not be available to fully support every performance target, including but not limited to the schedule for performance. Subsequent annual updates of this plan will reflect and adjust for those fiscal constraints based on the latest available information.
Additionally, there are many more connections (lines) and interdependencies (footnotes in red) than are displayed on the actual timelines. The very nature of science is multi-disciplinary and interdependent. Consequently, those relationships that are depicted are only illustrative, although they are believed to largely representative of the primary relationships.
6.0 Program Evaluation:
The ASCR program conducts frequent and comprehensive evaluations of every component of the program. Progress against established plans is evaluated by periodic internal and external performance reviews. These reviews provide an opportunity to verify and validate performance. Quarterly, semiannual, and annual reviews consistent with specific program management plans are held to ensure technical progress, cost and schedule adherence, and responsiveness to program requirements.
All on-going projects undergo regular (every three to five years) peer review and merit evaluation based on procedures set down in 10 CFR 605 for the extramural grant program, and under a similar process for the laboratory programs and scientific user facilities. Results of these evaluations are used to modify program management as appropriate. Additionally, all new projects are also selected through peer review and merit evaluation.
The Advanced Scientific Computing Advisory Committee (ASCAC) was established by DOE to provide independent advice on complex scientific and technical issues related to the ASCR program. The ASCAC is charged with providing advice on:
Promising future directions for advanced scientific computing research
Strategies to couple advanced scientific computing research to other disciplines The relationship of the ASCR program to other Federal investments in information technology research. A Committee of Visitors (COV) has also been appointed under the guidance of ASCAC to review the management practices of the ASCR program. In particular the COV will examined the decision process for awarding grants and for determining priorities of funding among the various activities within the ASCR program.
Change control and off-ramps:
Science changes rapidly and breakthroughs in knowledge by our science programs, other agencies, industry and the international science community create a constant state of flux. Although there are long-term research themes and lengthy horizons for new cutting-edge tools, basic research must be constantly revisited in a context of new discoveries and the most promising current opportunities.
Additionally, basic research is, by its nature, unpredictable. Results that appear to mark a failed experiment are often much more significant to progress in the field than a “successful” result. This is the reason that expert review will be used to assess progress toward our objectives. It is critical that all evaluations take this unique aspect of research into account so that success will be judged as advancing the field rather than meeting the specifics of an objective or target.
Underpinning the Office of Science change control process and our off-ramps are a strong dependence on our program advisory committees, for us the ASCAC. Our program and our advisory committee are driven by the following three major criteria for evaluating change and possible off-ramps: Quality, Relevance, and Performance. These criteria are also the criteria that the Office of Management and Budget (OMB) applies to basic research.
As part of the Office of Science Strategic Planning process, our advisory committee is consulted on the actual Objectives for the program. A broader array of stakeholders from government, industry, and academia are also consulted. Such input helps form the basis for a new focus or direction at this more aggregate level, and the current objectives for this program were the result of a recently completed cycle and preparation of a new Office of Science Strategic Plan. The objectives from the Strategic Plan form the basis for this Program Plan.
Key Targets were also developed in consultation with ASCAC as part of OMB’s Program Assessment Rating Tool (PART) process. Progress reviews for these key targets will be conducted by ASCAC every three years. These reviews will allow us to assess progress so that the program can continue, redirect or discontinue the efforts that support those targets.
Ultimately, all decisions on focus, emphasis, resources, and possible shifts are vetted at the appropriate levels within our program - from the researchers to the program managers, and often to the level of the Associate Director. Depending on the scope of the issue and the venue, the Director of the Office of Science may be involved. For major off-ramps, the Director of the Office of Science is always involved and assumes final responsibility.
7.0 Key External Factors:
The activities funded by the ASCR program are coordinated with other Federal efforts through the Interagency Principals Group, chaired by the President’s Science Advisor, and the Information Technology Working Group (ITWG). The ITWG evolved through an interagency coordination process that began under the 1991 High Performance Computing Act as the High Performance Computing, Communications, and Information Technology (HPCCIT) Committee. The Federal IT R&D agencies have established a 10-year record of highly successful collaborative accomplishments in multiagency projects and in partnerships with industry and academic researchers. The multiagency approach leverages the expertise and perspectives of scientists and technology users from many agencies who are working on a broad range of IT research questions across the spectrum of human uses of information technology. DOE has been an active participant in these coordination groups and committees since their inception and the ASCR program will continue to coordinate its activities through these mechanisms including an active role in implementing the Federal IT R&D FY 2002-2006 Strategic Plan under the auspices of the National Science and Technology Council and the President’s Science Advisor.
The ASCR program has many connections with other organizations, and is dependent on their planning needs, identified challenges, information, scientific data sharing, and more. Some of the key external factors for the ASCR program, include: the SC-wide SciDAC effort and other scientific computational activities with the other SC programs, and interactions with industry on advanced networks, software, and computational hardware.
External factors that affect the programs and performance include: (1) mission needs as described by the DOE and SC mission statements and strategic plans; (2) evolving scientific opportunities, which sometimes emerge in a way that revolutionizes disciplines; (3) results of external program reviews and international benchmarking activities of entire fields or subfields, such as those performed by the National Academy of Sciences; (4) unanticipated failures, for example, in the evaluation of new computer architectures for science, that cannot be mitigated in a timely manner; (5) strategic and programmatic decisions made by other (non-DOE) Federal agencies and by international entities; and (6) the evolution of the commercial market for high performance computing and networking hardware and software.