International Journal of
Information and Education Technology

Abstract

In the rapidly evolving landscape of education, the integration of digital technologies has led to the emergence of Massive Open Online Courses (MOOCs), which are becoming increasingly available across various educational domains. Despite a prevailing preference for traditional educational methods, MOOCs present viable alternatives that complement rather than replace conventional classroom instruction. The EUMaster4HPC is a newly established pan-European master’s program that focuses on High Performance Computing (HPC). This consortium aims to develop a distinctive curriculum centred around modular education in HPC. Additionally, EUMaster4HPC is committed to promoting its own MOOC offerings. In this paper, we provide a detailed analysis of design strategies for producing MOOC focused on OpenACC, OpenMP Offloading, and Heterogeneous-Compute Interface for Portability (HIP) programming models, all of which are crucial for accelerating computations on GPUs. These models emphasize high-level parallel programming, while HIP also addresses low-level programming, necessitating familiarity with CUDA for effective learning. Furthermore, the discussed programming models support heterogeneous programming, allowing for the simultaneous utilization of both CPU and GPU resources. This work investigates the instructional design methodology, course content, structure, and strategies employed in this MOOC, thereby contributing to the broader discourse on digital education in HPC.