Building a Universal Nuclear Energy Density Functional George F. Bertsch, University of Washington

Density functional theory has been spectacularly successful in condensed matter and quantum chemistry; the new SciDAC of the above title hopes to repeat that success in nuclear Density functional theories use drastically reduced sets of variables to describe a quantum system of interacting particles. The variables are fields that extend over 3-D space, with at least as many fields as there are particles in the nucleus. The energetics is determined by local or nearly local functions of these fields. Stated in this way, the theory of the 2000 or so known nuclei based on the 100 or so orbital fields per nucleus is certainly within the scope of present day high performance computing.

The present theory is now being benchmarked, but it is clear that it is inadequete for the purposes in mind, to make predicative calculations of nuclear properties needed for technological applications. Generalizations can be made extending the nonlocality or adding correlations explicitly, but they will certainly require several orders of magnitude increase in the computational demands of the theory.