Planning for the Leadership-Class Computing Facility

点击翻译按钮获取中文摘要

NSF has played a central role in the transformation of scientific research and as a leader in enabling the use of High-Performance Computing (HPC) to advance discovery for almost four decades. The leadership-class computing program is intended to provision advanced computational capabilities to enable transformative research for all of Science and Engineering (S&E) that would not otherwise be possible. NSF's current leadership-class computing investment is the support of the Frontera project at the Texas Advanced Computing Center (TACC) at The University of Texas at Austin. Frontera is the Phase 1 system of a two-phase approach to eventual deployment of a much more capable Leadership-Class Computing Facility (LCCF) in approximately 2024. This award will support the conceptual design of a Phase 2 system and facility.The project design approach will extend best practices in the design/operate/evaluate cycle that TACC has used to successfully deploy some of the largest computing systems in the world for open scientific research. This process will have four keystones: science user requirements, future technology assessment, alternative design evaluation, as well as risk management for cost, scope and schedule. Testbeds feature prominently in the design process and will provide users with early access to future computational hardware and will offer system designers quantitative insights into the effectiveness of these technologies. Testbed partners will include major cloud and technology vendors as well as early exascale computing technology adopters. In addition, the design process will be guided by science inputs and technology drivers identified by science and technology engagement teams. These teams will be composed of distinguished scientists representing a broad range of expertise and experience to guide the Phase 2 system design. The team members will also serve as community leaders to bring together colleagues in workshops, meetings, and other engagement venues to inform the process further. Moreover, the project will engage with the scientific software community, including library and community code providers, to ensure the project optimizes the selection of Phase 2 computing technologies to enable transformative S&E discoveries for the future.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.