![]() This issue... View from the Inside Humble Weed Becomes Genome Model E-mail Reminder
This issue... View from the Inside Humble Weed Becomes Genome Model E-mail Reminder
This issue... View from the Inside Humble Weed Becomes Genome Model E-mail Reminder
This issue... View from the Inside |
View from the InsidePioneering a New Branch of ScienceInterview with Pat Dehmer, Acting Director of the Office of Computational and Technology ResearchEnergy Science News: Dr. Dehmer, you've recently become Acting Director of our newest Office. Please tell us about the research activities within this Office? Dehmer: We're the focal point for long-term computational and technology research in applied mathematical sciences, high performance computing, communications, and the current information infrastructure. The work crosscuts and supports many other programs within the Office of Science. Because of the vastness and diversity of the work, we've organized into three main divisions:
Energy Science News: What do you see as the role of the MICS division? Dehmer: MICS research provides scientists with the capabilities to analyze, model, simulate, and--most importantly--predict complex phenomena of importance to the Office of Science and to the Department of Energy (DOE). Some of the complex problems that require advanced computing research include climate modeling; combustion processes; subsurface transport of pollutants; rational design of new materials to produce new alloys, superconductors, polymers, and catalysts; the effects of aging on the nuclear stockpile; and the analysis and storage of vast amounts of real data from experiments at our facilities for high energy physics, nuclear physics, and materials sciences. For many problems, traditional theoretical and experimental approaches may not be suitable; theory may be inadequate to handle the complexity, and experiments may not be feasible because they are too dangerous, too expensive, or simply impossible owing to the length and time scales involved. Computer simulation has become a third branch of science that complements theory and experiment and allows a new approach to previously intractable problems. Energy Science News: I'm sure there are challenges. Dehmer: Yes. The challenges are associated with using these capabilities effectively. For example, the significant changes in hardware, which can occur as often as every 18 months, may require completely new approaches to computing software. This has already changed the way computation is performed in materials science, biology, and fusion energy. Teams involving scientists from a discipline plus mathematicians, computer scientists, and experts in computer graphics, data management, and advanced computer networks are now necessary to make use of the most modern computer and communications capabilities. This situation is quite different from that of even 10 years ago when individual scientists could code and perform calculations with little or no support from others. Energy Science News: Was this the impetus for the Collaboratories? Dehmer: Yes, the collaboratories are part of this evolution. Collaboratories such as the Materials MicroCharacterization Collaboratory (MMC) and the William R. Wiley Environmental Molecular Sciences Laboratory (EMSL) are already "laboratories without walls," making it possible for large, geographically distributed teams to collaborate using major experimental facilities, computational resources, and data resources. They are sharing both state-of-the-art equipment and staff expertise with other scientists, educators, industry, and students across the country. In the coming years, collaboratories are expected to play an increasingly important role in the Nation's scientific enterprise. The programs that conduct the R&D for the supercomputers and Internet technologies that make the operation of collaboratories possible are MICS programs. In all of these areas, the requirements far exceed the current state-of-the-art as well as the tools that the commercial marketplace can deliver. These facilities, such as the National Energy Research Scientific Computing Center (NERSC) and the Energy Sciences Network (ESnet), enable scientists to use the tools developed. Another type of facility is itself a research project--the Advanced Computing Research Facilities (ACRFs)--which represents the evolution of the High Performance Computing Research (HPCR) Centers. These centers combine research in computer software and hardware with targeted applications. One of the major issues is how different choices in computer architecture affect the ability of a system to scale to very large numbers of processors and very high performance. Leadership in this area was made evident this year at the SC98 Supercomputing Conference: the predominant number of awards were given for the performance of computers operated by these facilities. Energy Science News: Anything else to tell us, Dr. Dehmer? Dehmer: In pioneering this new branch of science, MICS recognizes the importance of educating future scientists with the breadth of skills needed in advanced computing research and in interacting with the disiplinary sciences. To that end, MICS supports a Computational Science Graduate Fellowship program. Energy Science News: Thanks, Dr. Dehmer. We'll look forward to your progress reports. Related Articles Supercomputing: Collaboratories: Modeling: Internet Support: |
![]() ![]() ![]() ![]() ![]() |
![]() ![]() | ||
![]() | ||
www.pnl.gov/energyscience/ |