High Performance Computing and especially GPU Computing are a main focus area in research and teaching at the Institute of Applied Mathematics (IAM). For example, the IAM has been one of the first groups in Germany to actively work in the field of GPU Computing since 2004. Driven by innovative research, GPU Computing has grown into a major discipline, with applications in almost all scientific domains. In the academic year 2012/2013, NVIDIA supports our teaching activities in this field with the entry of the IAM into the renowned CUDA Teaching Center program. NVIDIA provides matching funds for teaching assistants and a set of GPU boards, which will be included into the department's servers so that students can use them.
In the winter term 2012/13, the new lecture "Applied Scientific Computing" is offered for the first time. The lecture provides both the foundations as well as practical aspects of parallel computing like parallel thinking in algorithm engineering, the implementation using CUDA C/C++ and OpenMP, and illustrated examples from different application domains. The lecture is open to students from all math degree programs, and via DoWiR, the Dortmund Center for Scientific Computing, also for interested participants from other departments.
Contact: Dominik Göddeke