Collaborations

  • ICL/UTK about MPI process placement with TreeMatch
  • UWLAX about network locality with hwloc and netloc
  • Cisco about hardware and network locality with hwloc and netloc
  • The OpenMPI consortium about hardware and network locality with hwloc and netloc
  • Eth Zurich on topology mapping problems
  • PPL (U. Illinois at Urbana Champaign) on topology-aware load-balancing (through the Inria-Urbana-Argonne Joint Lab).
  • University of Tokyo on the adaptation of MPI and runtime systems to MIC processors.
  • ORNL about high performane networking with CCI.
  • Inria team Bacchus on mesh handling and (re) distribution with PaMPA and Scotch.
  • CEA/DAM on runtime systems for clusters of multicores.