"High performance computing plays an important role in different industrial areas. Especially in research and development, computational intensive applications are needed for simulations and visualisations. To create the right preconditions for such extensive applications, parallel computing and powerful computer architectures are required", Professor Wolfgang Schreiner of the RISC - Research Institute for Symbolic Computation - explained. He is the main national co-ordinator in Austria of the European-wide initiative PRACE.
Participants from the EU and the world - e.g. from Brazil and China - attended the lectures of the international speakers of RISC Software GmbH, Maxeler Technologies, Catalysts GmbH, CSC, Finland, Universities of Malaga, Spain, and Virginia, USA, and the Leibniz Supercomputing Centre from Munich, Germany. The participants got deep insights into solving multi-physical problems with the Elmer open source software on large-scale HPC resources. Moreover, the multi-various application possibilities of HPC in Europe were discussed.
"Even the fastest computer in the world will not come to a conclusion, if the applied software doesn't use the scale of the machine. If you can use 100.000 or even more processors, it's best to use them all simultaneously. We showed the participants of the PRACE Spring School what to take in account in their software to use large-scale computers efficiently", stated Alexander Leutgeb, speaker at the PRACE Spring School 2014 and employee of RISC Software GmbH in Hagenberg.
Solving Problems in Research and Industry with Supercomputers Key issues of science and society can only be met by bundling all available solution methods. Software solutions are an increasingly important part to solve these problems. For example, in the further exploration of the factors and interactions related to climate change; for innovations in materials with the help of nanotechnologies; in the development of energy-efficient cars and airplanes; and in the field of life sciences. For these tasks, computing power and storage are required, which only can be provided by distributed, massively parallel computer architectures ("High Performance Computing").