120 credits in science/engineering including Introduction to Scientific Computing or Scientific Computing I. Participation in High Performance Programming or participation in Low-level Parallel Programming. Proficiency in English equivalent to the Swedish upper secondary course English 6.
There is an everlasting demand for increased computer capacity and performance. High-performance computers with only one processor are expensive, and there are physical limitations that put an upper limit on the performance. These problems are solved by connecting many cheap standard-type processors in one computer and letting them work simultaneously on one task. The amount of performance gained by the use of multi-core or multi-processor computers is strongly dependent on the software algorithms and implementation. In particular, the software must be "parallelised" to run on multiple cores simultaneously. In this course, you will learn how different types of parallel computers are built up and how they work. Parallel algorithms for fundamental computational problems are presented. Important questions related to this are if there is parallelism inherently in a particular algorithm, or if reformulation of the algorithm can increase the parallelism. A good part of the course is hands-on parallel programming using programming models such as MPI (Message Passing Interface).
