Dynamical Parallel Applications on Distributed and High Performance Computing Systems

This paper provides an extended overview of recent dvelopments for exploiting distributed, Grid and High Per-formance Computing (HPC) resources with applications like Geoscientific Information Systems (GIS) based on the GISIG actmap-project. Focus is on frameworks for optimising the dynamical parallel use of computing resources for future cooperation and development concepts, integrating software and hardware architecture aspects. Using parallel processing and the method of event triggering from within Active Source can be used to exploit the vast computing power of distributed MultiCore parallel systems for a multitude of purposes like geoinformation processing, geophysical data analysis, information systems, and e-Science. An extended case study for an application InfoPoint demonstrates the algorithm here. As various obstacles showed up with implementing and minimising the complexity of the application-resource workflow, the creation of future Web and HPC services on top of HPC and Distributed Systems will be a solution for dedicated issues. The new extended framework of the Grid-GIS house is presented here, showing the case study for using these concepts for exploration purposes. For implementation testing distributed resources and the new multisite supercomputer resources of HLRN-II (North-German Supercomputing Alliance) have been used