Teaching Introductory Parallel Computing Course with Hands-On Experience

This paper presents an innovative course designed to teach parallel computing to undergraduate students with significant hands-on experience. This course represents an introduction to the main topics of parallel, distributed and high-performance computing (HPC). The course introduces main concepts and architectures used in parallel computing today, and improve students' skills to develop parallel programs using major parallel programming paradigms: MPI (Message Passing Interface), OpenMP (Open-Multiprocessing). The main objective of the course is to teach practical parallel programming tools and techniques for MIMD with shared memory, MIMD with distributed memory and SIMD. Each of these software tools can be used to give students experience with parallelization strategies, and ability to rate the quality and effectiveness of parallel programs. We evaluate the success of our approach through the use of testing and survey and provide directions for further improvements in teaching parallel programming.