The PhD thesis: “Toward Resilience in High Performance Computing: A Prototype to Analyze and Predict System Behavior,” focuses on enhancing the resilience of high-performance computing (HPC) systems. As HPC systems grow in complexity and scale, they become more susceptible to failures, which can disrupt computations and lead to significant data loss. This research aims to address these challenges by developing a prototype framework that can analyze and predict system behavior, thereby improving the reliability and efficiency of HPC systems. The thesis explores various methodologies for monitoring and analyzing system performance data to identify patterns and anomalies that may indicate potential failures. By leveraging machine learning and statistical techniques, the framework can predict system behavior and suggest proactive measures to mitigate risks. This approach not only enhances the resilience of HPC systems but also optimizes resource utilization and reduces downtime, ultimately contributing to more robust and efficient computational environments.

The Center for Information Services and High-Performance Computing (ZIH) at TUD Dresden University of Technology is a central scientific institution managing the university’s communication infrastructure and operating its central IT services and high-performance computing servers. It is involved with the National High-Performance Computing (NHR) and the National Research Data Infrastructure (NFDI) in Germany, providing access to high-performance computing resources and expertise to support scientific research across German universities. This involvement aims to equip researchers with the necessary computing capacity and support to efficiently utilize complex algorithms and AI methods.

The Cluster for Advancing Electronics Dresden (cfaed) is a Cluster of Excellence, established in 2012 under the German Excellence Initiative. It focuses on interdisciplinary research to develop new technologies for electronic information processing, using innovative materials and concepts such as silicon nanowires, carbon nanotubes, and DNA-Origami. The cluster unites around 300 scientists from various institutions to explore and create cutting-edge electronic systems.