Implementation and Evaluation of MLEM-Algorithm on GPU

In a current project at our chair, we analyse modern HPC systems with heterogeneous architectures towards exascale computing. A central element in this project is to find a real-world problem-solving application, which is representative and stands for a class of typical HPC problems. As a good example we have chosen the maximum likelihood expectation maximization (MLEM) algorithm.

In this work, an efficient MELM Implementation on GPU is to be developed. To verify the efficiency, the GPU Implementation will be benchmarked against our parallel MLEM implementation on our HPC- Cluster.


Contact: Dai Yang (German or English)