A framework for adaptive collective communications for heterogeneous hierarchical computing systems

Luiz Angelo Steffenel (1) , Grégory Mounié (2)

1 CRESTIC - Centre de Recherche en Sciences et Technologies de l'Information et de la Communication - EA 3804
2 ID-IMAG - Informatique et Distribution

Collective communication operations are widely used in MPI applications and play an important role in their performance. However, the network heterogeneity inherent to grid environments represent a great challenge to develop efficient high performance computing applications. In this work we propose a generic framework based on communication models and adaptive techniques for dealing with collective communication patterns on grid platforms. Toward this goal, we address the hierarchical organization of the grid, selecting the most efficient communication algorithms at each network level. Our framework is also adaptive to grid load dynamics since it considers transient network characteristics for dividing the nodes into clusters. Our experiments with the broadcast operation on a real-grid setup indicate that an adaptive framework allows significant performance improvements on MPI collective communications.

Keywords: Performance modeling; Adaptive techniques; Polyalgorithms; Collective communication; MPI

Domains: Distributed, Parallel, and Cluster Computing [cs.DC]

Luiz Angelo Steffenel, Grégory Mounié. A framework for adaptive collective communications for heterogeneous hierarchical computing systems. Journal of Computer and System Sciences, 2008, 74 (6), pp.1082-1093. ⟨hal-00261436⟩