Computational Materials Science

The things we rely on in our daily lives—cars, planes, power plants, calculators, computers, medical scanners, to name just a few—work only as well as the materials they are made of. Materials underpin major industries, steer global markets, and are critical to national and economic security. Enhanced material performance is essential to the mission of the U.S. Department of Energy—for increased efficiency in energy production and use, and for the successful deployment of environmentally benign products.

To date, the development of technologically advanced materials has largely been accomplished through empirical approaches. This approach, though successful in the past, is time consuming and expensive, and has intrinsic limitations. The drive to discover novel materials for existing and new applications necessitates a new approach based on a fundamental knowledge of the detailed interactions that occur at the atomistic scale and an understanding of how these interactions affect the properties of the material. To successfully pursue this science-based approach to materials design requires computational resources that far exceeds today's computational capabilities. The new computing capabilities that are just now becoming available are critical to realizing this goal.

This opportunity is timely because the computational materials science community is ready to use the terascale (trillions of calculations per second) computers to become available in 2000-2005. Algorithms already exist that reach to terascale performance levels. For example, the 1998 Gordon Bell performance prize, which is awarded annually for the best achievement in high-performance computing, was received by a group of computational materials scientists for their simulations of magnetic materials. This work is a step along the path to designing magnetic materials for various specialized purposes.

Success in this new era of computational materials science will require multidisciplinary teams of theoretical and computational materials and molecular scientists, computer scientists, and applied mathematicians working side by side to create the new theoretical methods, computational approaches, and mathematical algorithms for high fidelity simulations of materials. By fully integrating this new simulation capability with materials characterization, synthesis, and processing, progress toward understanding the relationship between the structure and properties of materials will be dramatically accelerated. This is the foundation for realizing the goal of designing materials on the computer.