Research

New digital business models and the digital transformation of the traditional industry require a scalable IT infrastructure that dynamically adapts to business development. This makes cloud computing the most important IT paradigm for the operation of modern IT applications: IT resources can be procured to virtually any extent and charged according to usage.

However, applications need to be designed specifically for operation in the cloud. In order to ensure high scalability of cloud applications, concepts from the field of parallel and distributed systems are used and adapted. Currently, however, the cloud is mainly used for operating web-based applications, which can be comparatively easily parallelized. In addition, current trends such as artificial intelligence (deep learning), computer vision and virtual / augmented reality could also benefit from the cloud. The execution of these more complex application classes requires novel concepts to unlock the full potential of cloud computing.

We are working to develop methods and architectures for non-trivially parallel distributed systems that open the cloud as execution environment for a variety of complex application classes. We use current technologies such as OpenStack, Ceph, Hadoop, Mesos, Docker and Kubernetes.