The World Teaching of Parallel and Distributed Programming

Abstract

The parallel and distributed programming is widely used in cloud computing. This is due to factors such as the popularization of multicore and heterogeneous CPU, which allow significant performance gains over sequential processing. ACM and IEEE recommend to teach the parallel and distributed programming and also provide details on how this topic should be studied in computing courses. However, the teaching of parallel programming is not trivial and differs in various undergraduate courses throughout countries. The differences include teaching approaches, theoretical and practical classes, instructional materials, embedded topics, number of hours, required and elective topics, prerequisites, among others. This article presents how the main institutions of higher education in the world approach the teaching of parallel programming, with the purpose of evaluating its adherence to the syllabus proposed by ACM and IEEE. The universities considered in this study were chosen according to a specific ranking and geographically separated by continent. After that, we compared with the ACM and IEEE-Computer Society reference curricula, highlighting the main differences. Our results show that there are still significantly differences regarding teaching HPC, mainly in relation to syllabus and topics covered, and that is hard to find available documents that show clearly how such subjects are conducted.