Abstract: Computing power and paradigms are perpetually evolving with time and technology. A corresponding growth in the volume and availability of data has driven the implementation of high performance computing (HPC) across a broader spectrum of disciplines, which in turn is driving demand for new HPC professionals. However, financial, logistical, and educational barriers exist which inhibit students from beginning HPC career pathways.
As a means of circumventing these challenges of learning foundational HPC concepts, viable low-cost methods can be implemented. The aim of this work is to create a portable low-cost cluster where students can be introduced to computational science and its applications within an informal environment. This is done by creating a graphical user interface based application that allows a user to run a modern numerical weather model on a cluster of Raspberry Pi computers.
This work helps introduce fundamental HPC concepts to a broader audience that may not have any prior exposure to computational science or possess technical skills. By creating an informal and simplistic supercomputing environment, the learning process for such students can be accelerated.