Parallelisierung von physikalischen Rechnungen auf GPUs

What it is about

You deal with methods for parallelizing physical calculations on GPUs. In the seminar you develop current topics, implement algorithms of computer-aided physics on highly parallel processors, and understand how parallelization parameters influence execution speed.

What you will be able to do

• Implement algorithms of computer-aided physics on GPUs
• Assess the influence of parallelization parameters on performance
• Prepare and present scientific literature on GPU parallel work
• Convincingly answer questions in the presentation

What the module consists of

• Seminar presentationCentral examination and performance form: preparation, creation and presentation of a scientific topic
• Programming projectPractical implementation of parallelization methods on GPUs

Teaching method

• Literature search and studyBasis for getting started with current scientific topics
• Presentation preparation and creationEnables structured processing and communication of the content
• Discussion with the lecturerClarification of open questions and deepening of the topic
• ProgrammingPractical application and testing of the developed methods
• Presentation and answering questionsDemonstrates learning success and presentation ability