You are in the main area:Events
Headerimage for: Meet Us

Meet Us

Meet Us

10. 11. 01
HLRS Presentation at SC 2001 in Denver/Colorado

November 10-16, 2001 The Grid is not enough

The High Performance Computing Center Stuttgart (HLRS) is a national HPC Center in Germany. In addition, together with debis Systemhaus GmbH, debis-sfr and Porsche, it has formed a company to provide access to supercomputers for research and industry. RUS/HLRS tries to establish a distributed working environment for its users to see and use all resources (T3E-512, SX-5/32, SR8000-16) in a seamless way. At SC2001 HLRS will demonstrate its activities in the field of Grid Computing. We will show the main building blocks HLRS is working with and how they are working together. Examples from industry include the car and aerospace sector. Scientific research is demonstrated in medicine, biology, chemistry and physics. The results are visualized by our own collaborative visualization tool COVISE. To show the potential of the resulting framework HLRS will integrate computing resources at Sandia National Laboratories (SNL), Pittsburgh Supercomputing Center (PSC), Tsukuba Advanced Computing Center (TACC)/Japan, Manchester Computing Centre (MCC), National Center for High Performance Computing (NCHC)/Taiwan, Japan Atomic Energy Research Institute (JAERI)/Japan, University of Houston, and HLRS/Germany.

SC Global Participation:

HLRS together with the John von Neumann-Institute for Computing in Jülich will host a workshop on Grid Computing as part of SC Global. SC Global consists of live speaker presentations transmitted from Denver via group conferencing as well as live speaker presentations from Jülich and Stuttgart shown at the conference in Denver.

 

Pictures from the HLRS Booth
Pictures from the European Village

Scenarios planned for demonstration:

SNL will kindly provide most of the equipment for the HLRS/SNL collaboration booth. A number of projects are demonstrated that focus on designing and implementing middleware components, on putting them together to create a framework for grid computing and on running applications to exploit the potential of such frameworks.

 

DAMIEN (Distributed Applications and Middleware for Industrial use of European Networks) is one of three European funded Grid-Computing projects. In close cooperation with datagrid and eurogrid it develops and integrates software tools for metacomputing used for industrial codes from aerospace industries.

 


PACX-MPI (PArallel Computer eXtension) allows to extend the functionality of native MPI implementations to clusters of MPPs and PVPs. PACX-MPI covers the complete MPI1.2 standard.

 


COVISE (COllaborative VIsualization and Simulation Environment) is a visualization software that was developed in several European projects. It aims to allow distributed visualization, collaborative working and distributed simulation in a HPCN environment.

 

Blood flow simulation is a project together with the University hospital at Ulm. CT data are created at Ulm and transformed into 3D-representations of arteries. After automatic mesh creation the arterial blood flow is simulated at Stuttgart and the data is prepared for visualization. This will help surgeons to improve the quality of their work.

 

URANUS (Upwind Algorithm for Nonequilibrium Flows of the University of Stuttgart) has been developed at the IRS to simulate the reentry phase of a reusable space vehicle in a wide altitude velocity range. To compute large 3-D problems the URANUS code was parallelized at HLRS. The parallel multiblock program is able to run on several hundreds of processors and is used as an example application for the metacomputing testbed.

 


MARMOT is a MPI debugging and verification tool that aims to increase the portability of MPI programs and to track down bugs like race conditions and deadlocks.

 

The P3T toolkit is a object oriented framework for general particle tracking applications. Examples are Molecular Dynamics, Direct Simulation Monte Carlo or the simulation of sedimenting particles in fluids.

 

HLRS will also show the collaboration with the John von Neumann Institute for Computing (NIC) that extends to the collaboration in the field of seamless computing (UNICORE), gigabit networking and common use of software, developed by the two centers.

 

 

Demonstration Coordination at HLRS:

  • SC2003 coordination supervisor and booth contact: Matthias Müller and Michael Resch
  • Metacomputing coordination: Edgar Gabriel
  • Visualization supervisor: Ulrich Lang
  • Visualization coordination, COVISE/COVER activities: Uwe Wössner.