Modeling Materials at the Petascale

Developing a massively parallel multi-scale method for the study of strongly correlated materials – such as magnets and superconductors

Mark S. Jarrell (project webpage)
Univerity of Cincinnati

Many important classes of materials are strongly correlated, that is, the interactions between electrons are large. Most magnets and superconductors fall into this category. Although these are materials of great technological promise, they are poorly understood due to long-ranged spin and charge correlations, competing ground states, and their complex phase diagrams. This project will develop a massively parallel multi-scale method for the study of strongly correlated materials that separates the problem into short length scales treated explicitly with quantum Monte Carlo methods, intermediate length scales treated diagrammatically using vertices obtained from the Monte Carlo, and long length scales treated in the mean field. Fundamental developments in linear algebra will be used to increase the explicitly treated length scales. The codes developed will scale to the petascale, efficiently employing thousands of processors, but have a simple user interface accessible to the average theoretical or experimental user. This formalism will be used to develop a better understanding of materials such as lanthanides, actinides, and complex transition metal oxides where correlations over many length scales are central to the phase diagram, or to aid in material design to improve and search for new correlated materials. The codes will be developed by an interdisciplinary team of computational physicists with expertise ranging from many body theory to the first principles electronic structure of actinides and lanthanides, in collaboration with applied mathematicians who specialize in the multiscale linear algebra problems and iterative solution of coupled nonlinear equations which are central to quantum simulations of strongly correlated materials.

Advances in materials and chemistry are often critical to progress in all three mission areas. This project will advance our understanding, simulation and design of magnetic materials and superconductors for energy and national security applications as well as basic research applications.


Science Application: Materials Science and Chemistry

Project Title: Next Generation Multi-Scale Quantum Simulation Software for Strongly Correlated Materials

Principal Investigator: Mark S. Jarrell
Affiliation: University of Cincinnati

Participating Institutions and Co-Investigators:
University of Cincinnati, Karen Tomko
Oak Ridge National Laboratory, Thomas A. Maier (co-PI), Eduardo F. D’Azevedo
University of California at Davis, Zhaojun Bai (co-PI), S. Y. Savrasov, Richard T. Scalettar

Funding Partners: U.S. Department of Energy - Office of Science, Advanced Scientific Computing Research, and the National Nuclear Security Agency.

Budget and Duration: Approximately $0.6 million per year for five years 1

Other SciDAC Materials Science & Chemistry efforts

1Subject to acceptable progress review and the availability of appropriated funds


Home  |  ASCR  |  Contact Us  |  DOE disclaimer