Johannes Gebert

Profile

Johannes Gebert leads a research group in future computing at the High‑Performance Computing Center Stuttgart (HLRS). His 10 years of experience as a design, structural, and software engineer bridges the gap between domain specialists and high‑performance computing (HPC). He finished his PhD on a new biomechanical method on HPC systems with summa cum laude. A research visit at the Innovative Computing Laboratory (ICL) of the University of Tennessee, Knoxville (UTK), also strengthened his background in HPC. Several lectures and a Future Computing Lab foster young talents in undergraduate and graduate studies. Johannes is an Early Career Editorial Board Member for Advances in Engineering Software.

CV GitHub Google Scholar LinkedIn Mail

Future Computing

The working group in future computing was set up right after completing my PhD. We aim for intrinsically motivated undergraduates, PhD students, and post-docs to move the limits of high-performance computing. We are open to collaborating in all ways that serve that goal.

High-performance computing centers support and conduct research to enable large-scale simulations of domain-specific researchers. However, existing hardware architectures quickly approach the limits of physics and technical feasibility. While significant speed-ups are still possible with available technologies, research on new hardware and software paradigms may enable crucial performance gains in the future of computing. Improvements for specific niches of mathematics, e.g., optimization problems, are expected to justify the use of specialized accelerators.

Consequently, research in future computing focuses on innovative computing machinery and its usage. It aims to close the gap between computing paradigms and existing and future HPC simulations. In addition, AI has emerged to shape HPC applications and the HPC environment itself. Heterogeneous hardware approaches promise to accommodate specialized routines, AI, and HPC. Extensive research must ensure technical feasibility, software that uses the probable advantages of new accelerators, and reasonable total cost.