The course advances the students knowledge of how problems can be solved by parallel
processing and accelerators (e.g., graphic processors), how problems are decomposed
into parallel parts, and how programs can be optimized for different computer and
accelerator architectures.
The following topics are covered:
- Introduction to homogeneous and heterogeneous parallel computers.
- Introduction to graphics processors and accelerators.
- How problems are decomposed to enable parallel execution.
- OpenMP.
- How to write programs for computer clusters using, e.g., MPI.
- How to write programs for graphics processors using, e.g. CUDA.
- Parallel patterns, such as prefix sum, mapreduce, matrix calculations, merge sort, and graph search.
- Examples of how parallel processing is used within different domains, such as machine learning, image
processing, and image analysis.
- How to plan and deploy (virtual) clusters and parallel application in a cloud.
- Common benchmarks and how they are used to evaluate performance.
- Tools to test and debug parallel programs.
The course will be given during the period: 20 Jan - 22 Mar, 2020
Please sign up no later than 10/1 2021
For more information see the course plan:
https://kursplan.lnu.se/kursplaner/syllabus-4DV657-1.pdf If you have any questions please turn to
