Monday, May 10, 2010

JUROPA MPI Buffer on demand

To enable huge runs with lots of MPI ranks you have to disable the per default allocated all-to-all send buffer on the NEC- Nehalem Cluster Juropa at FZ J├╝lich.


Here is an excerpt from the official docu:







Most MPI programs do not need every connection



  • Nearest neighbor communication
  • Scatter/Gather and Allreduce based on binary trees
  • Typically just a few dozen connections when having hundreds of
    processes
  • ParaStation MPI supports this with ?on demand connections?
  • export PSP_ONDEMAND=1
  • was used for the Linpack runs (np > 24000)
  • mpiexec --ondemand
  • Backdraw
  • Late all-to-all communication might fail due to short memory
  • Default on JuRoPA is not to use ?on demand connections?

Links


Juropa Introduction @ FZJ

No comments:

Post a Comment