link to homepage

Institute for Advanced Simulation (IAS)

Navigation and service

Cooperation with companies concerning further developments of cluster operation systems

Cluster operation systems provide the basic layer to create a parallel
HPC-system out of the massive collection of stand-alone server systems
interconnected by fast networks in cluster computers. It implements on
a parallel-system level the required functionality provided by
operating-systems in standard single- and multi-core systems. Without
such middleware a coherent use of cluster computers is
unimaginable. Together with ParTec Cluster Competence Center GmbH
( JSC forms the core-team of the ParaStation
consortium. Its central objective is the advancement of the
ParaStation cluster middleware providing a highly scalable,
fault-tolerant high-performance implementation of the MPI
standard. This includes facing the challenges on the way to Exascale
like the support novel many-core processors, innovative highly
scalable interconnects or the enhanced requirements on resiliency

Contact: Norbert Eicker