Academic Career and Research Areas
Prof. Gerndt (*1960) has been responsible for parallel computer architecture at the Department of Informatics since 2000. His research focuses on programming tools for the efficient use of parallel supercomputers and for optimizing the efficiency of cloud applications. In particular, he deals with methods of automatic performance analysis and optimization in the context of research projects related to the Periscope Tuning Framework.
Prof. Gerndt studied computer science at the University of Bonn and did his doctorate there in 1989 on the German supercomputer project SUPRENUM. He did postdoctoral work at the University of Vienna for two years. He worked at TUM’s Forschungszentrum Jülich research center until he was appointed professor in 2000. He completed his lecturing qualification at TUM in 1998. His work was based on the efficient use of shared virtual memory systems.
Key Publications (all publications)
Kumaraswamy M, Chowdhury A, Gerndt M, Bendifallah Z, Bouizi O, Locans U, Říha L, Vysocký O, Beseda M, Zapletal J: "Domain knowledge specification for energy tuning". Concurrency and Computation: Practice and Experience. 2018: e4650.Abstract
Jindal A, Podolskiy V, Gerndt: "Autoscaling Performance Measurement Tool". ICPE '18 Companion of the 2018 ACM/SPEC International Conference on Performance Engineering. Berlin. 09-13 April 2018: 91-92.Abstract
Gerndt M, Benkner A, César E, Navarrete C, Bajrovic E, Dokulil J, Guillén C, Mijakovic R, Sikora A: "A multi-aspect online tuning framework for HPC applications". Software Quality Journal. 2018; 26(3): 1063-1096.Abstract
Gerndt HM, Zima HP: “SUPERB: Experiences and Future Research”. In: Languages, Compilers, and Run-Time Environments for Distributed Memory Machines. Editors: Saltz J, Mehrotra P. 1992: 1–15.Abstract
Gerndt HM, Zima HP: “MIMD-Parallelization for SUPRENUM”. In: International Conference on Supercomputing (ICS '87). LNCS 297. 1987: 278–293.Abstract