Werner Hanke - Böcker

This volume presents a selection of reports from scientific projects requiring high end computing resources on the Hitachi SR8000-F1 supercomputer operated by Leibniz Computing Center in Munich. All reports were presented at the joint HLRB and KONWHIR workshop at the Technical University of Munich in October 2002. The following areas of scientific research are covered: Applied Mathematics, Biosciences, Chemistry, Computational Fluid Dynamics, Cosmology, Geosciences, High-Energy Physics, Informatics, Nuclear Physics, Solid-State Physics. Moreover, projects from interdisciplinary research within the KONWIHR framework (Competence Network for Scientific High Performance Computing in Bavaria) are also included. Each report summarizes its scientific background and discusses the results with special consideration of the quantity and quality of Hitachi SR8000 resources needed to complete the research.

Leading-edge research groups in the field of scientific computing present their outstanding projects using the High Performance Computer in Bavaria (HLRB), Hitachi SR8000-F1, one of the top-level supercomputers for academic research in Germany. The projects address modelling and simulation in the disciplines Biosciences, Chemistry, Chemical Physics, Solid-State Physics, High-Energy Physics, Astrophysics, Geophysics, Computational Fluid Dynamics, and Computer Science. The authors describe their scientific background, their resource requirements with respect to top-level supercomputers, and their methods for efficient utilization of the costly high-performance computing power. Contributions of interdisciplinary research projects that have been supported by the Competence Network for Scientific High Performance Computing in Bavaria (KONWIHR) complete the broad range of supercomputer research and applications covered by this volume.

Leading-edge research groups in the field of scientific computing present their outstanding projects using the High Performance Computer in Bavaria (HLRB), Hitachi SR8000-F1, one of the top-level supercomputers for academic research in Germany. The projects address modelling and simulation in the disciplines Biosciences, Chemistry, Chemical Physics, Solid-State Physics, High-Energy Physics, Astrophysics, Geophysics, Computational Fluid Dynamics, and Computer Science. The authors describe their scientific background, their resource requirements with respect to top-level supercomputers, and their methods for efficient utilization of the costly high-performance computing power. Contributions of interdisciplinary research projects that have been supported by the Competence Network for Scientific High Performance Computing in Bavaria (KONWIHR) complete the broad range of supercomputer research and applications covered by this volume.

High-Performance Computers (HPC) have initiated a revolutionary develop­ ment in research and technology since many complex and challenging prob­ lems in this area can only be solved by HPC and a network in modeling, algo­ rithms and software. In 1998 the Deutsche Forschungsgemeinschaft (German Research Association) recommended to install an additional Federal High­ Performance Computer followed by the one in Stuttgart. In January 1999 the Wissenschaftsrat (German Science Council) decided that the Leibniz­ Rechenzentrum (Computing Center) of the Bavarian Academy of Sciences in Munich should run the second Federal High-Performance Computer in Ger­ many. The investment cost of this Hochstleistungsrechner in Bayern (HLRB) was borne by the Federal Government of Germany and the Free State of Bavaria whereas the operating cost was at the expense of the Bavarian Gov­ ernment only. The operation of the HLRB is organized in combination with the - Leibniz-Rechenzentrum (LRZ) of the Bavarian Academy of Sciences as the operating authority of the HLRB - Steering Committee of the HLRB - Competence Network for Technical/Scientific High-Performance Comp- ing in Bavaria (KONWIHR). In 2000 a Hitachi SR8000-Fl was installed. It was the first Teraflops Com­ puter in Germany and reached a peak performance of two Teraflops after an extension at the end of 2001. The goal of HLRB is to provide computer facil­ ities necessary to solve challenging scientific and technological problems that cannot be solved on big servers but require large (storage) high-performance (very fast) computers and efficient software.