DOE Greenbook: Needs and Directions in High Performance Computing for the Office of Science
Edited by: Steve Jardin
Date of Report: June 2005
This report was prepared by the NERSC User Group for the U.S. Department of Energy Office of Science
Executive Summary
Computational science plays an essential role within each of the research programs in the U.S. Department of Energy (DOE) Office of Science (SC). There are a diversity of traditional scientific disciplines conducting research within the SC programs; including chemistry, materials science, quantum physics, geophysics, biology, plasma physics, nuclear physics, and high energy physics, among others. Researchers from within each of these disciplines are finding that computation at the largest scale provides a capability that is now considered essential for the advancement of each of the respective disciplines. Today's most powerful computers and the associated software are being used to produce new and more precise scientific results at the cutting edge of these disciplines, and this trend is destined to continue for years to come.
Many examples of the impact of large-scale computations on the sciences are presented in this document. Our basic understanding of chemical processes has increased. New insight as to the molecular basis for vision has been obtained. A comprehensive study of the impact of pollutants on climate change has been carried out. Our understanding of the complex process of photosynthesis has increased. Significant progress has been made in our ability to simulate the folding of proteins.
The improved models within fusion energy sciences have allowed researchers to design a new class of fully 3D magnetic confinement configurations, known as quasi-symmetric stellarators, with a "hidden symmetry" that should allow greatly improved confinement and stability properties. The inertial-confinement fusion computational program has developed high-resolution beam simulations that are being used to develop and guide experiments, and fast-ignition simulations that are elucidating the physics of that potential breakthrough option.
The high energy physics program has made great progress both in accelerator modeling and in obtaining detailed computational predictions of the masses of strongly interacting particles, an important test of quantum chromodynamics (QCD) and the Standard Model. Large-scale computation has provided new insight into the properties of atomic nuclei and how they behave at extremely high energy density. Massive supernovae have been modeled with unprecedented detail, shedding light on the evolution and fate of the Universe.
These and many more scientific advances have been enabled by the extraordinary computational science resources available at the National Energy Research Scientific Computing Center (NERSC). However, a common theme in each of the disciplines is that the current computational resources available through NERSC are saturated, and this lack of additional computing resources is becoming a major bottleneck in the scientific research and discovery process. A large increase in computer power is needed in the near future to take the understanding of the science to the next level and to help secure the U.S. DOE SC leadership role in these fundamental research areas.