Höchstleistungsrechenzentrum Stuttgart (HLRS)

Based on a long tradition in supercomputing at the Universität Stuttgart, the HLRS was founded in 1995 as the first national center for High-Performance Computing in Germany. Since then it has assumed an internationally renowned position in supercomputing. The National Science Foundation Award for High Performance Distributed Computing in 1999 and the HPC Challenge Award 2003 have emphasized this again. Today, HLRS provides services to researchers at universities and research laboratories in Germany and at the same time does a great deal of research in the field of modeling and simulation.

One of the key goals of the HLRS is the provision of the best system for each application. Hence, the decision for any hardware provided is always driven by demand of the users aiming to provide the best solution for a given problem. In consequence HLRS sees its role in evaluating the strengths and weaknesses of existing systems. At least as important is the feedback that HLRS is able to provide to its industrial hardware vendors in intensive partnerships. Thus HLRS has helped to improve HPC-systems by bringing to bear the requirements of simulation in the hardware design process.

On top of this HLRS is as a provider for simulation solutions. This requires taking the lead in bringing together various disciplines working towards solutions in research and engineering. So HLRS has become a breeding place for innovative simulation. Not the least because it has successfully managed to integrate means of virtual reality visualization and supercomputing providing the user with a real simulation workbench that virtually puts the scientist directly into the simulation process.

The NEC-HLRS Teraflop-Workbench Cooperation (www.tflop-workbench.de)

To improve the quality of services and research HLRS works closely with hardware and software vendors. The most important cooperation is the joint program with NEC called the Teraflop-Workbench Partnership.

The aim of this initiative is twofold. On the one hand it is designed to help close the gap between peak performance and sustained performance in supercomputing. On the other hand supercomputers have ceased to be stand alone large systems that inhale huge amount of data to exhale even larger amounts. As a standard tool current systems are integrated into a workflow of pre-processing, simulation and post-processing. This workflow itself is integrated into the overall workflow of either science or engineering. This puts the supercomputer into a heterogeneous environment of hardware and software. To handle this complexity and heterogeneity is the second part of the cooperation.

High Performance Computing Center Stuttgart (HLRS)
Universität Stuttgart
Allmandring 30
D-70178 Stuttgart
Germany

Phone: +49 (0) 711 685 2504
http://www.hlrs.de