EXECUTIVE SUMMARY Nuclear physics is manifest in areas as seemingly disparate as the history of the early universe, the generation of energy in the sun, and the creation of nearly all the elements in stellar furnaces and explosions. A major focus of nuclear physics is the strong force and the atomic nuclei whose binding is a direct result of it, and whose stability underlies that of the atoms and thus molecules forming the familiar matter of all life forms and everyday objects. The strong nuclear force is a complex one, much more so than the gravitational and electromagnetic forces familiar from daily life. Understanding this nuclear force places particularly difficult demands not only on experiment but also on theory and calculation, particularly on computational power needed for these calculations. Nuclear physics pursues several major areas of research, including the following: 1. Structure of nuclei 2. Nuclear forces and quantum chromodynamics (QCD), the quantum field theory of the strong interaction 3. Fundamental interactions and symmetries that govern the interactions and the quantum states observed 4. Nuclear astrophysics and the synthesis of nuclei in stars and elsewhere in the cosmos 5. Hot QCD for the highest temperatures known and phases of strongly interacting matter 6. Science and design of the accelerators, detectors, and other tools essential for pursuing these scientific questions. The analysis and interpretation of results, and the development of an encompassing theoretical and intellectual framework depend on significant and focused investments, particularly in large-scale computing facilities and their ancillary capabilities for data visualization, storage, retrieval and transmittal. These are necessary in parallel to the perhaps more familiar investments in state-of-the-art accelerators and detectors to support experimental investigations. Most forefront problems now driving active research depend on access to large-scale computing facilities for tasks ranging from designing the facilities, making predictions based on current theories, and determining the implications for scientists' overall understanding of the experimental results, all with more demanding requirements for precision. The scale of computing facilities needed for state-of-the-art scientific exploration far outweighs what is possible with desktop-to departmental-scale (few to hundreds of computing cores) resources-instead, requiring access to the (few) large national centers where tens of thousands of computing cores, capable of delivering sustained performance of hundreds of teraflops, are available. Many problems, coming from all areas of nuclear physics, still require that significant approximations be made before such powerful centers can process the problem, or cannot even be attempted at present centers as they require computing timescales on the order of years. Extrapolating to the computing power needed to remove such approximations indicates a major shift in the ability to address such forefront questions that will oc...