BERLIN Germany, October 26, 2017—The Gauss Centre for Supercomputing (GCS) is pleased to announce that it is continuing to serve as a co-sponsor of undergraduate students participating in the Student Cluster Competition (SCC) at the Supercomputing Conference 2017 (SC17) in Denver, Colorado/USA (Nov. 12-17, 2017). Team TUMany segFAUlts (pronounced too many segfaults), is a joint six-person team representing the Technische Universität München (TUM) and the Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU). The team--evenly divided between males and females--is the only German participant among the 16 teams taking part in the competition.
As Germany's national supercomputing institution, GCS is committed to encourage a wide variety of activities that help promote up and coming experts in the field of high-performance computing (HPC). "We are convinced that these kind of contests can be an enormous motivational spur for young HPC talent," explains Dr. Claus Axel Müller, Managing Director of GCS. "We don't want financial constraints related to international travel to be an insurmountable obstacle for the students, so we are happy to step in and help out. Independent of the eventual outcome of this contest, participating in an international competition - a friendly yet tough one - will be an invaluable experience for the undergraduate students. We wish them lots of success, and tons of fun."
According to the SCC rules, the six-student-teams have to compete in a non-stop, 48-hour challenge held on the SC17 showground that will, by definition, contain some unexpected elements of surprise, such as a mystery application (to be revealed during the competition) and an unannounced power outage. Apart from application performance and accuracy, the teams will be judged on the energy efficiency of their pre-assembled compute cluster and its power consumption. Equally important for the overall evaluation of the teams will be whether the students manage to impress conference attendees and especially interview judges with their HPC knowledge of their systems and the applications.