The section JARA-HPC is structured in such a way that interdisciplinary research questions can be addressed and answered in economically relevant periods based on the use of supercomputers.
For some years now there has been a growing realisation that the application software in computational sciences is lagging behind HPC hardware developments. While several Petaflop-scale supercomputers are now available worldwide, it is becoming increasingly difficult to exploit these machines efficiently with single applications. Substantial efforts are needed in order to enable the involved computational science communities to solve problems with high scientific impact through efficient use of high-end supercomputing resources.
JARA-HPC and the Jülich Supercomputing Centre are jointly committed to meeting this challenge: With the so-called Simulation Laboratories (SimLabs) we combine domain-specific research with the required methods for highly parallel computing on high-performance computers and supercomputers. Three corresponding JARA-HPC SimLabs are currently established:
The Cross-Sectional Groups (CSGs) comprise methodical activities that are relevant for the users of high-performance computers with utterly diverse scientific backgrounds. For instance, members of the CSG "Immersive Visualization" support neuroscientists of the section JARA-BRAIN in their investigations of the human brain.
So far, two CSGs have been established:
Development of algorithms for the immersive and interactive, as well as spatially distributed and cooperative visualization, and their application within the Vista software Framework.
Development of algorithms for performance analysis and performance increase of parallel programs, and their application in tools such as Scalasca or SIONlib.
Seed Funds are a means to promote projects focussing on promising fields of research that are still in their initial phase and not yet ready for third-party funding.
Seed Funds contribute to consolidating already existing core fields and ensuring interdisciplinary cooperation with the engineering sciences. Several state-of-the-art research projects have already been funded, possibly developing into the future top fields of research.
Since 2012, scientists at RWTH Aachen University and Forschungszentrum Jülich have been using the JARA-HPC partition for the realization of interdisciplinary research projects.
Currently, more than 50 projects are being processed using the computing resources of the JARA-HPC partition.
A selection of older projects in the first two computing time periods (2012) and a brief project description can be found here