High Performance Algorithms for Electronic Materials

University Of Minnesota-Twin Cities, Minneapolis MN

Investigators

Abstract

This grant is supported by the Divisions of Materials Research and Advanced Computational Infrastructure and Research. The research is in the area of computational materials research and involves a collaborative effort in both materials research and computational science. The overall goal of the research is to exploit high performance computers for solving large scale and complex problems that arise in modeling real materials. The research will center on electronic materials with extended and point defects, and non-ideal interfaces and in the form of amorphous solids, atomic clusters, liquids, and glasses. The continuing evolution of high performance computers is creating new opportunities for the application of sophisticated electronic structure techniques to the study of technologically important materials. A few orders of magnitude improvement in speed are feasible in the near future and will enable us to step beyond the current computational limits in materials science and achieve an unprecedented understanding of the physical and electronic properties of materials. Such achievements are not made by progress in computer hardware alone, but depend more so on innovations in algorithms. The development of new algorithms in multidisciplinary research requires an understanding and appreciation of the different disciplines involved by the collaborating members. Based on prior experience of this group, progress can be best made by a close collaboration between computer scientists and materials scientists. Thus, algorithms will be developed, implemented and applied to a number of materials problems. Applications will include: localized systems, atomic clusters and quantum dots; liquids and disordered semiconductors; dilute magnetic semiconductors. Algorithms to be developed include: out-of-core methods; special methods for computing eigenvectors and eigenvalues; avoiding eigenvectors and eigenvalues. %%% This grant is supported by the Divisions of Materials Research and Advanced Computational Infrastructure and Research. The research is in the area of computational materials research and involves a collaborative effort in both materials research and computational science. The overall goal of the research is to exploit high performance computers for solving large scale and complex problems that arise in modeling real materials. The research will center on electronic materials with extended and point defects, and non-ideal interfaces and in the form of amorphous solids, atomic clusters, liquids, and glasses. The continuing evolution of high performance computers is creating new opportunities for the application of sophisticated electronic structure techniques to the study of technologically important materials. A few orders of magnitude improvement in speed are feasible in the near future and will enable us to step beyond the current computational limits in materials science and achieve an unprecedented understanding of the physical and electronic properties of materials. Such achievements are not made by progress in computer hardware alone, but depend more so on innovations in algorithms. The development of new algorithms in multidisciplinary research requires an understanding and appreciation of the different disciplines involved by the collaborating members. Based on prior experience of this group, progress can be best made by a close collaboration between computer scientists and materials scientists. Thus, algorithms will be developed, implemented and applied to a number of materials problems. ***