ATLAS computing on the HPC Piz Daint machine

Not scheduled
15m
OIST

OIST

1919-1 Tancha, Onna-son, Kunigami-gun Okinawa, Japan 904-0495
poster presentation Track8: Performance increase and optimization exploiting hardware features

Speaker

Michi Hostettler (Universitaet Bern (CH))

Description

The Piz Daint Cray XC30 HPC system at CSCS, the Swiss National Supercomputing centre, is in 2014 the highest ranked European system on TOP500, also featuring GPU accelerators. Event generation and detector simulation for the  ATLAS experiment has been enabled for this machine. We report on the technical solutions, performance, HPC policy  challenges and possible future opportunities for HEP on extreme HPC systems. In particular a custom made  integration to the ATLAS job submission system has been developed via the Advanced Resource Connector (ARC) middleware. Further, some GPU acceleration of the GEANT4 detector simulations were implemented to justify the allocation request for this machine.

Primary author

Sigve Haug (Universitaet Bern (CH))

Co-authors

Andrej Filipcic (Jozef Stefan Institute (SI)) Michi Hostettler (Universitaet Bern (CH)) Dr Rodney Walker (Ludwig-Maximilians-Univ. Muenchen (DE))

Presentation materials