Cyber Training: Pilot -- Breaking the Compute Barrier, Upskilling Agri-Food Researchers to Utilize HPC Resources

University Of Minnesota-Twin Cities, Minneapolis MN

Investigators

Abstract

There is a dearth of scientists with expertise in the agri-food and environment domains that have compute-to-scale capabilities enabled by High Performance Computing (HPC) environments. Low adoption of HPC capabilities among agri-food researchers can be largely attributed to the real (or perceived) complexity of using HPC. Moreover, traditional training courses rooted in the CSE sciences often lack the contextualized problem focus and hands-on access to tailor-made learning data and problem sets that are familiar to and thus useful for upskilling this particular sector of the workforce. This project proposes to develop and deploy a multi-module learning curriculum tailored to CI-applications in the agri-food sciences that is provided as a synchronous, virtual offering with substantial hands-on application-based learning opportunities. The challenges that the proposed work will be made generalizable, such that other left-behind communities seeking to capitalize on core advances in data science and HPC can leverage the approaches and infrastructure developed under this proposal. The proposed multi-module course is focused on building the foundational, data-driven skills necessary to create a sustainable community of skilled CI Users through tailored, discipline-appropriate course materials targeted at bridging the gap between domain specific science and computer science for agri-food scientists. This proposal aims to develop and deploy a multi-module learning curriculum tailored to Cyberinfrastructure (CI)-applications, notably High-Performance Computing (HPC), in the agri-food sciences that is provided as a synchronous, virtual offering with substantial hands-on application-based learning opportunities. The 30-person course will be delivered via a containerized learning environment to ensure all learners have ready access to an identical set of tools. The first three course modules provide the basic building blocks for HPC-based analytics, followed by a series of hands-on application modules that enable agri-food researchers with the levels of competency needed to facilitate HPC analyses of critical agri-food problems. The course will be accessible to academic (undergraduate, graduate, and faculty/staff) audiences around the US and abroad (especially targeting underrepresented populations of students), as well as individuals working in US government agencies and agri-business firms. To enable both academic and non-academic accessibility, this pilot project will host the CI-focused agri-food analytics curriculum on Microsoft Azure cloud computing infrastructure, but the course will introduce learners to the portfolio of available private, academic, and cloud-based HPC resources. The project team will work with internal and external agri-food networks and leverage the capabilities of the ACCESS Knowledge Base Ask.CI and/or Community Affinity Groups. The team will engage in a series of external and internal content and delivery audits throughout the grant period to ensure the identification of optimal HPC learning pathways for agri-food researchers. After delivering alpha-, beta- and full-course instances of their HPC for Agri-Food Researchers course, the course will continue to be offered 2-3 times annually through their GEMS Learning portfolio beyond the life of the grant. 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.