new third dimension: roughly 500 to 1,000 personal computers. This translates to at least days or weeks of dedicated time on an ASCI-level parallel system. The Type Ia supernova problem seems to be worse. In addition to demands such as those just mentioned, at least two other issues arise: subgrid models for burning, and grid distortion in the strongly differentially rotating system of a binary merger. These crucial roles played by computations in HED astrophysics have been recognized by the Department of Energy (DOE) and strongly supported within the DOE National Nuclear Security Administration’s ASCI Alliances program and the DOE Office of Science’s Science Discovery through Advanced Computing (SciDAC) program.
In summary, there is a wide variety of areas in which theory, numerical simulations, and experiments on HED facilities can address aspects of astrophysical phenomena. Areas of promising overlap include the physics of supernova explosions and supernova remnant evolution, high-Mach-number astrophysical jets, planetary interiors, photoevaporation-front hydrodynamics of molecular clouds, photoionized plasmas around accreting black holes, and relativistic plasmas in gamma-ray burster fireballs. This selection of topics may well be just the “tip of the iceberg” as more experience is obtained in carrying out both large-scale numerical simulations and scaled HED experiments in support of astrophysics.