You are in the main area:Events
Headerimage for: HPC Challenge

HPC Challenge

Computational Grid

This computational Grid will be distributed over all continents, spreading the calculation over a dozen computers with thousands of processors. An application is thereby able to take advantage of this 12 TeraFlops setup.


The application shown is a modified version of fastDNAml.
The metacomputing library PACX-MPI will be used as communication middleware to distribute the application across HPCs, including:

  • Terascale Computing System (TCS)
  • several Top-500 cluster
  • several Cray T3Es
  • several Origin 2000 & 3800
  • several IBM SPs
  • Hitachi SR8000
  • Fujitsu PrimePower
  • NEC SX 5
  • ia32 & ia64 cluster


Visualization will be performed in a collaborative VR environment using Covise and Cover.

Private internal Project Pages

Please enter here for the internal pages.