WASHINGTON, D.C., Aug. 5, 2020 — The U.S. Department of Energy (DOE) announced it will provide $57.5 million to establish two multidisciplinary teams to develop new tools and techniques to harness supercomputers for scientific discovery.
The two teams, led respectively by Argonne and Lawrence Berkeley National Laboratories, are composed of leading experts in computer science, software development, applied mathematics, and related disciplines. The teams will provide expertise and develop tools to enable scientists to take full advantage of DOE’s high performance computing capabilities.
The need for such multi-disciplinary teams is growing as the world moves into the age of Exascale computing, with advanced systems embodying entirely novel architectures coming online.
“The Department of Energy is home to some of the world’s fastest supercomputers, and as we move into the Exascale computing era, these resources are continuing to rapidly advance in performance, architecture, and design,” said Under Secretary for Science Paul Dabbar. “These Argonne and Lawrence Berkeley-led teams are comprised of experts in computing and applied mathematics and will ensure that the American scientific community can fully harness our country’s leading capabilities in high performance computing.”
Both teams involve researchers from multiple DOE laboratories, universities, and industry and are part of a program known as SciDAC, or Scientific Discovery through Advanced Computing. The SciDAC program brings together experts in computer science and applied mathematics with researchers in specific scientific disciplines to develop new high-performance computing tactics for tackling scientific questions.
As a joint effort among the six major program offices within DOE’s Office of Science and DOE’s Office of Nuclear Energy, SciDAC addresses problems in disciplines including high energy and nuclear physics, condensed matter physics, materials science, chemistry, fusion energy sciences, Earth systems research, and nuclear energy.
Teams, known as “SciDAC Institutes,” are expected to take advantage of DOE supercomputing facilities at Argonne (IL), Oak Ridge (TN), and Lawrence Berkeley (CA) National Laboratories. Their work is aimed to spur advances in the use of artificial intelligence and machine learning.
The Argonne-led team, “RAPIDS2: A SciDAC Institute for Computer Science, Data, and Artificial Intelligence,” will focus on community outreach to support scientists with application development.
The Lawrence Berkeley-led team, “Frameworks, Algorithms, and Scalable Technologies for Mathematics (FASTMath) Institute,” will focus on the development of new mathematical techniques.
Total planned funding is $57.5 million over five years, with $11.5 million in Fiscal Year 2020 dollars and outyear funding contingent on congressional appropriations. A list of lead and partner institutions for the two teams can be found on the homepage of the DOE Office of Science, Office of Advanced Scientific Computing Research, under the heading, “What’s New.”
Source: Department of Energy