Russian-German seminar Supercomputing in Scientific and Industrial Problems took place from 9th to 11th March at Keldysh Research Center (Moscow). The seminar had been organized by the High Performance Computing Center of the University of Stuttgart, which is the leading center of this kind in Europe, and Keldysh Institute of Applied Mathematics of the Russian Academy of Sciences (RAS), one of the recognized leaders in the field of high performance computing in Russia.
The High Performance computing Centers of the University of Stuttgart and Keldysh Institute of Applied Mathematics RAS have a similar strategy of evaluating scientific approaches to developing and creating effective scalabale parallel algorithms and technologies to provide solutions for calculations in breakthrough technologies.
Scientists from SPbPU were invited to take part in the seminar by its co-chairpersons – B.N. Chetverushkin, Member of the Presidium of the Russian Academy of Sciences, and Professor Michael Resch, Director of the High Performance Computing Center of the University of Stuttgart. The research results were presented obtained by the teams headed by Professor E.M. Smirnov (the Department of Hydro- and Aerodynamics at the Institute of Applied Mathematics and Mechanics, SPbPU), A.Yu. Snegiryov (the same Department) and N.N. Shabrov (the Department of Computer-Aided Mechanical Engineering of the Institute of Metallurgy, Mechanical Engineering and Transport, SPbPU).
On the whole, the seminar enabled its participants, namely the representatives of the academic organizations of RAS, as well as specialists in the sphere of high performance computing from the universities of Munich, Aachen, Dresden, Freiburg and Stuttgart, to share their experience and opinions, to inform each other about the achieved results, to discuss the ways to develop the architecture of multiprocessor supercomputers and algorithmic composition of solutions to fundamental tasks, and to prove the necessity to develop a specific architecture for an exaflop supercomputer that will help to solve problems at different scales effectively at the same time.