University brings HPC into the classroom
Southern Utah University (SUU) is teaming up with Interactive Supercomputing (ISC) to make HPC easy and accessible for students and faculty staff of all levels. It is part of a programme developed by SUU’s College of Computing, Integrated Engineering, and Technology (CCIET) to foster an integrated, interdisciplinary curriculum combining maths, engineering and computer science departments.
SUU is providing students and faculty staff access to the school’s Dell 128-node parallel cluster using ISC’s Star-P software. Star-P is a software platform that delivers parallel computing capabilities to desktop users, without them having to become parallel programming experts. Star-P enables SUU users to easily program models and algorithms using their familiar desktop languages by automatically transforming the application to run on the parallel clusters. Star-P eliminates the need to re-program the applications in complex languages such as C and Fortran or use MPI (message passing interface) to run in parallel.
‘Our mission is to help students achieve their academic goals and to compete on a global level for careers in government, industry, secondary education, and acceptance to graduate school. Supercomputing is increasingly playing a critical role in those career paths,’ said Mikhail Bouniaev, dean of CCIET and professor of mathematics. ‘While SUU is a small university, we believe our unique interdisciplinary undergraduate research program that integrates supercomputing will become a model for schools across the country.’
The college provides programmes in computer science, engineering, information systems, mathematics, technology and interdisciplinary studies. The CCIET programme has three goals:
- To create a solid integrated undergraduate curriculum;
- To provide HPC resources for applied research in mathematics, engineering and computer science, for use in applications such as differential equations, fluid dynamics, computer-aided design, etc; and
- To provide a resource for faculty staff who want to get started in HPC, but don’t know where or how to begin.
The programme kicks off this August at the Utah Advanced Computing Summer Institute in Cedar City, UT. Here, attendees will use Star-P to learn how HPC works in research, teaching, design and production. It will feature HPC experts from industry, government and academia. Star-P helped solve two problems in launching SUU’s interdisciplinary research programme, according to Bouniaev. First, the small college did not have a lot of parallel computing expertise on staff from which they could draw support. Secondly, faculty and students were constrained by the size and scope of problems that they could tackle on their desktops using tools like Matlab and Python. ‘Star-P will make the transition from desktop computing to HPC smooth and easy,’ he said. ‘I believe supercomputing is a key area that can unite almost all disciplines in our college. With it, we can create curricula from which students can develop a sound understanding of fundamentals, as well as investigate specialised theories, practices, and experiments that enhance their learning.’