MRI: Acquisition of HPC at AU (Expanding Capabilities for Research and Research Training at American University through Shared High-Performance Computing)

该奖项允许美利坚大学获得高性能计算（HPC）系统，以扩展研究和研究培训的能力。获取高性能计算资源是实施非盟2008年战略计划的关键一步，该计划要求全校范围内协调努力，扩大研究和研究生教育。这一面向全校研究人员的共享资源将建立在非盟在社会科学方面的优势之上，同时支持扩大科学项目的计划。它将通过将可用的计算速度提高近100倍来进一步提高研究人员的生产力。HPC系统将支持越来越多的研究人员，他们将应用计算科学的工具来理解经济学、教育、统计学、物理科学和计算机科学中的复杂问题。在经济学方面，研究人员将使用提议的HPC系统来研究整合所有可用行为信息来设计监管市场的方法，探索信息技术在增长地理中的作用，并模拟扩大联邦儿童福利补贴对儿童健康的影响。在教育方面，研究人员将开发手持设备，有可能彻底改变有特殊需要的学生的教育，他们将开发研究工具，将跨学科的方法整合到教育史和教育政策中。在统计学和科学领域，研究人员将使用HPC系统将“超级学习者”算法应用于流行病学中的关键问题，为所有科学家提供利用志愿者计算机网络解决常见问题的手段，为创建和探测超冷物质的量子态设计一种新方法，并为从医学到“绿色”工业化学的各种应用绘制酶的共变残基图。研究人员将使用算法、建模、仿真、3D渲染和地理可视化来解决这些科学和经济上重要的问题。计算科学中最先进的方法需要快速处理、大块RAM、专门的图形处理和高磁盘输入/输出速度，这些速度远远超过了目前在AU工作站上可用的速度，而这些将由提议的HPC系统提供。这台高性能计算系统的获得将大大推进非洲大学的研究任务和研究培训。当研究人员聚在一起管理这个系统，培训使用这个系统，并参加有关用这个系统完成的研究的研讨会时，他们将共同努力解决实质性的计算问题。该系统将改善研究生的研究密集型学习环境。已经是许多研究团队成员的本科生将得到更好的训练，以便在研究生项目和工作场所做出贡献。非盟校园内交流和协作的增加将扩大现有的努力，将代表性不足的群体纳入研究，因为共享强大的计算资源将导致形成同伴和导师支持的临界质量。

This award permits American University to acquire a high-performance computing (HPC) system to expand capabilities for research and research training. Acquisition of HPC resources is a crucial step in the implementation of AU's 2008 strategic plan, which calls for campus-wide, coordinated efforts to expand research and graduate education. This shared resource for researchers across the campus will build on AU's strength in the social sciences, while supporting plans to expand programs in the sciences. It will further the productivity of researchers by increasing the available computing speed by a factor of almost 100-fold. The HPC system will support a growing community of researchers who will apply the tools of computational science to understand complex problems in economics, education, statistics, the physical sciences, and computer science. In economics, the researchers will use the proposed HPC system to investigate ways that incorporate all available behavioral information to design regulated markets, to explore the role of information technology in the geography of growth, and to simulate the effect of expanding federal subsidies for child welfare on the health of children. In education, researchers will develop handheld devices that have the potential to revolutionize the education of students with special needs, and they will develop research tools to integrate interdisciplinary methods into the history of education and educational policy. In statistics and the sciences, researchers will use the proposed HPC system to apply the "super learner" algorithm to key problems in epidemiology, to give all scientists the means to leverage volunteer computer networks to solve common problems, to design a new method for creating and probing quantum states of ultracold matter, and to map covarying residues in enzymes for applications ranging from medicine to "green" industrial chemistry. The researchers will address these diverse scientifically and economically important questions using algorithms, modeling, simulation, 3D rendering, and geographic visualization. State-of-of-the art methods in computational science require fast processing, large blocks of RAM, specialized graphics processing, and high disk input/output speeds that far exceed what is currently available on workstations at AU, and which will be provided by the proposed HPC system. The acquisition of this HPC system will significantly advance the research mission and research training at AU. As researchers come together to manage the system, to train on use of the system, and to attend seminars on research completed with the system, they will work together to solve substantive computational problems. The system will enhance the research-intensive learning environment of graduate students. Undergraduate students, already members of many research teams, will be better trained to contribute in graduate programs and in the workplace. Increased communication and collaboration across AU's campus will amplify existing efforts to include underrepresented groups in research because the shared resource for powerful computing will lead to the formation of a critical mass for peer and mentor support.