CyberTraining: Pilot: Fostering Computational Excellence (FOCEX): Addressing the Disconnect between Advanced CyberInfrastructure and Educational Preparedness

该项目将以讲习班和教材包的形式为科学、技术、数学和工程(STEM)的学生和研究人员开发培训材料，以帮助将计算研究转化为现代高端计算平台。该项目解决了研究人员在使用计算和数据密集型应用程序时遇到的长期挑战，这些应用程序依赖于现代高性能计算集群中部署的新计算加速器硬件。培训的重点是计算流体力学(CFD)的应用，即预测飞机等物体周围的流动。此外，培训材料将与其他学科共享许多共同元素，并将很容易适用于除CFD以外的其他领域。研讨会将介绍一般概念和示范代码，并在实际应用程序中演示其功能。该项目包括一项档案和传播计划，并将继续向广大受众进行广告宣传。该项目将培训一批研究人员和学生，通过使用最先进的数值技术和在先进的网络基础设施(CI)中有效地利用现代异质计算体系结构来解决流体力学中的挑战性问题。该项目有两个主要组成部分：直接培训和教学促进。直接培训将作为为期4周的暑期班试点项目，面向对CFD和编程有基础了解的研究生和本科生。在这个项目中，参与者将获得研究CFD管道的实践强化培训，包括了解数值技术、高级CI中可用的异构计算体系结构的知识、有效利用这些体系结构所需的编程技能，以及机器学习和优化。将开发培训模块，以准备在整个NSF Access计划中分发给教师的教学工具包。目的是促进这些科目在全国各地的不同课程中的整合。直接培训试点预计将直接接待学生，而教学促进具有扩大规模并使更多学生受益的潜力。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

This project will develop training materials in the form of workshops and teaching kits for students and researchers in science, technology, mathematics, and engineering (STEM) to aid in translating computational research into modern high-end computing platforms. This project addresses the longstanding challenges researchers encounter when working with computing and data-intensive applications that rely on new computing accelerator hardware deployed in modern high-performance computing clusters. The training will focus on computational fluid dynamics (CFD) applications, i.e., predicting the flow around objects like an aircraft. Additionally, the training material will share many common elements with other disciplines and will be easily applicable to other fields besides CFD. The workshop will present general concepts and exemplary code and demonstrate its functionality in a real-world application. An archival and dissemination plan is included in the project, and advertising to a broad audience will be pursued.This project will train a cohort of researchers and students to tackle challenging problems in fluid mechanics by employing state-of-the-art numerical techniques and efficiently utilizing modern heterogeneous computing architectures in advanced cyberinfrastructure (CI). This project has two major components; direct training and teaching facilitation. Direct training will be piloted as a 4-week-long summer school designed program for graduate and undergraduate students with a foundational understanding of CFD and programming. During this program, the participants will gain hands-on intensive training in research CFD pipelines, which include understanding numerical techniques, knowledge of the heterogeneous computing architectures available in advanced CI, and the programming skills needed to effectively utilize them, as well as machine learning and optimization. The training modules will be developed to prepare teaching kits to be distributed to instructors throughout the NSF ACCESS program. The aim is to facilitate the integration of these subjects in different coursework across the nation. The direct training pilot is expected to host students directly while teaching facilitation has the potential to scale up and benefit many more students.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.