Concurrent Computing in an Upper-Level Computer Science Curriculum

Michigan Technological University, Houghton MI

Investigators

Abstract

omputer Science (31) The primary goal of this project is to design a set of comprehensive and flexible course materials that cover multithreaded, multiprocess, shared memory and distributed memory parallel, and distributed programming under a few common principles (e.g., concurrency and nondeterminism). These materials can be used across several courses or with a single dedicated course. A secondary goal is to develop software tools that abstract low-level details and accomplish all important concurrent operations under a single and unified interface. Furthermore, these tools include a visualization component to help students visualize the behavior of concurrent programs and the inner working of synchronization protocols, communication mechanisms, clock synchronization, and many other algorithms. With unified course materials, pedagogical tools, and a visualization system, it is expected that students can become familiar with the state-of-the-art of concurrency, acquire basic knowledge and skills, know how to handle concurrency under different environments, and be well-prepared to approach concurrent applications and their software development in their future careers. The work builds on a previously funded NSF project on teaching multi-threaded programming. An evaluation advisory committee is assisting in the formative and summative evaluations of this project to determine the impact of the materials on student understanding of the content and their productivity using the tools. The materials are being tested and evaluated at three additional sites to assess their effectiveness for diverse student populations. Dissemination is being accomplished by the development of faculty workshops and short courses at national and regional conferences. The materials are being developed for publication by a national publisher.