The International Conference for High Performance Computing, Networking, Storage and Analysis
Preparing Computer Science Students for An Increasingly Parallel World: Teaching Parallel Computing Early and Often.
Authors: Martin Burtscher (Texas State University), Wuxu Peng (Texas State University), Apan Qasem (Texas State University), Hongchi Shi (Texas State University), Dan Tamir (Texas State University)
Abstract: The widespread deployment of multicore systems over the last decade has brought about drastic changes in the computing landscape. However, most undergraduate computer science (CS) curricula have yet to embrace the pervasiveness of parallel computing. In their first years, students are typically exclusively trained to think and program sequentially. Too firm a root in sequential thinking can be a non-trivial barrier for parallel thinking and computing. Thus, there is an urgent need of including parallel computing concepts earlier in the CS program. This poster describes an innovative early-and-often approach of systematically integrating parallel computing into existing curricula. Through a series of self-contained, context-aware course modules, parallel concepts are introduced in early CS courses; reinforced throughout the curriculum, and collated in an upper-level capstone course. The project bears special significance as it leverages Texas State's diverse student body to broaden the participation of Hispanic students; a traditionally underrepresented group in STEM.