CC* Compute: High Performance Campus Computing for Institutional Research at the American Museum of Natural History

点击翻译按钮获取中文摘要

Through the National Science Foundation CC* program, the American Museum of Natural History (AMNH) expands the High-Performance Computing (HPC) capabilities that directly support the Museum's research. AMNH conducts scientific research and educational activities across astrophysics, anthropology, biology, and geosciences. The increasingly data-intensive nature of this research requires greater access to computational resources and to ever more sophisticated tools, including local and remote HPC clusters.In this project, AMNH is expanding its on-premise computing cluster capacity and consolidating all existing clusters into a unified open-source software framework. These clusters are connected to the Museum's Science DMZ, a high-performance network specifically designed for research data flows, which provides high-speed network access between the Internet2 and the AMNH on-premise clusters. Additionally, AMNH researchers can execute complex workloads at scale using cloud resources at Amazon via the same local HPC management framework. Federation with InCommon provides both AMNH researchers and outside collaborators with secure access to these resources via a common authentication and authorization framework. Finally, the AMNH clusters are integrated with the Open Science Grid allowing AMNH to offer idle computing cycles to the wider research community while providing AMNH researchers with the same access to remote computing resources. These improvements greatly expand the overall HPC capacity available to AMNH scientists, increasing the speed and effectiveness of their research and decreasing time to discovery. Additionally, the work of AMNH scientists informs the Museum's educational and curatorial programs, directly benefiting AMNH students and the public.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.