CC* Campus Compute: UTEP Cyberinfrastructure for Scientific and Machine Learning Applications

当今高性能计算（HPC）领域发生的一场重大革命是越来越多地使用机器学习来补充模拟。德克萨斯大学埃尔帕索分校的这个项目为计算科学和工程领域的研究人员提供了访问最先进的gpu加速硬件和软件资源的途径。资源的选择是由科学驱动程序的需求驱动的，这些驱动程序解决了可再生能源、先进制造、先进材料、电力系统、网络安全以及使用模拟和/或机器学习的量子计算等领域的重要社会和国家问题。新的硬件与现有的校园HPC集群集成在一起。硬件通过参与开放科学网格与更广泛的学术研究团体共享。除了为教师提供宝贵的资源外，增强的HPC集群还可供学生用于与他们的论文和论文工作相关的项目以及课堂项目。校园HPC集群上的软件堆栈反映了具有类似体系结构的NSF超级计算机系统的可用性。学生研究人员为集群的运营和维护提供帮助，从而使他们具备未来在高性能计算和大规模机器学习领域工作所需的技能。该项目的预期成果包括：1)增加跨学科对高性能计算的使用，2)提高教师和学生使用最先进的高性能计算技术的技能，以及3)提高产生研究成果和出版物的速度。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

A major revolution taking place in High Performance Computing (HPC) today is the increasing use of machine learning to complement simulation. This project at University of Texas at El Paso is providing researchers in computational science and engineering fields with access to state-of-the-art GPU-accelerated hardware and software resources. The choice of resources was motivated by the requirements of science drivers that address important societal and national problems in the areas of renewable energy, advanced manufacturing, advanced materials, electric power systems, cybersecurity, and quantum computing using simulation and/or machine learning. The new hardware is integrated with the existing campus HPC cluster. The hardware is shared with the broader academic research community through participation in the Open Science Grid. In addition to providing a valuable resource for faculty, the augmented HPC cluster is available for use by students for projects related to their thesis and dissertation work as well as for class projects. The software stack on the campus HPC cluster reflects that available on NSF supercomputer systems with similar architectures. Student researchers provide assistance with operation and maintenance of the cluster, thus equipping them with skills needed for future careers in HPC and large-scale machine learning fields. Expected outcomes of the project include: 1) increased use of HPC across disciplines, 2) improvement in faculty and student skills in using state-of-the art HPC technologies, and 3) increased rate of producing research results and publications.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.