Efforts in the first 15 years of the SciDAC program have conclusively demonstrated that significant advances in science can be realized through close and sustained interactions among domain scientists and applied mathematicians. The SciDAC-4 FASTMath team has a proven record of engaging application scientists to enable new scientific discovery through strong and sustained interactions and a demonstrated ability to tackle difficult algorithmic and implementation issues as computer architectures undergo fundamental shifts.
We have impacted domain science by providing application scientists new structured and unstructured adaptive mesh simulations, high-order discretizations, adaptive time integrators, robust linear, nonlinear and eigensolvers, numerical optimization methods, and uncertainty quantification tools. Our contributions in SciDAC-3 ranged from providing the foundations for next generation application codes to developing key functionalities that enabled faster time to solution, higher resolutions, or more robust capabilities.
Application advancements through prior work by the FASTMath and QUEST teams enabled next-generation codes, faster solutions, and more robust simulations. Examples of successful advancements are available.
The FASTMath SciDAC Institute is building on successes such as these and provides the mathematical algorithms, software tools, and human expertise to enable effective use of the high-end computing facilities by Department of Energy and other application scientists. The FASTMath team partners with DOE applications scientists from across the Office of Science, the NNSA, and the applied energy offices. For the SciDAC-4 partnership teams that we are involved with, we give a brief overview of our collaborations and progress in advancing scientific goals. More information can be obtained by contacting the indicated FASTMath team members.
Office of Nuclear Energy
Office of Basic Energy Sciences
The Department of Energy’s Office of Basic Energy Sciences (BES) supports fundamental research to understand, predict, and ultimately control matter and energy at the electronic, atomic, and molecular levels, to provide the foundations for new energy technologies and to support DOE missions in energy, environment, and national security. Through SciDAC partnerships, BES is developing new algorithms and computational approaches which could dramatically accelerate the discovery of new materials and processes as well as provide fundamental understanding and improvement of current materials and processes. Implementing these new algorithms on current and next generation massively parallel computers requires a team approach that includes materials and chemical scientists, applied mathematicians, and computer scientists.
Develop simulation tools to model catalysis from electronic structure to device cell.
Develop simulation tools to study strongly correlated topological materials.