Developing HPC Learning Pathways: Challenges and Recommendations

Abstract

Recent growth in computational research, touching all disciplines and all aspects of research and discovery, has resulted in increased demand for High Performance Computing (HPC) skills among members of the research community. Providing adequate training for practitioners coming from a wide range of technical and research background is not straightforward. Concurrently, learners now have a range of informal options for learning, not all of which provide accurate or recommended content. Curated learning paths are one of the strategies that can help address this challenge. Starting with data collected from a set of HPC practitioners we take a closer look at the diverse set of learning goals and concerns within the "Computational Researcher" persona. The range of responses highlights the challenges of building HPC learning paths for computational researchers wishing to develop and improve their HPC skills. To address the challenge we provide a set of recommendations that will enable learners to align their goals with a set of required skills and build a path appropriate to their learning needs.