Motivated by Japan's recent accomplishment of building the world's fastest supercomputer, this report discusses a methodology for creating "a new class of computational capability in the United States that is optimal for science." With the needs of several different scientific disciplines in mind, the report outlines the goals of the IBM Blue Gene and the proposed Blue Planet projects. Specific architectural characteristics are defined that will be needed to accommodate the scientific community. Milestones are laid out through 2009 and a number of alternative technologies that should be explored are also mentioned. This report provides an excellent glimpse into the future of large-scale scientific computing.