StarPU-MPI: Task Programming over Clusters of Machines Enhanced with Accelerators

Cédric Augonnet (1) , Olivier Aumage (2, 3) , Nathalie Furmento (2, 3) , Raymond Namyst (2, 3) , Samuel Thibault (2, 3)

1 NVIDIA - NVIDIA
2 LaBRI - Laboratoire Bordelais de Recherche en Informatique
3 RUNTIME - Efficient runtime systems for parallel architectures

GPUs clusters are becoming widespread HPC platforms. Ex- ploiting them is however challenging, as this requires two separate paradigms (MPI and CUDA or OpenCL) and careful load balancing due to node heterogeneity. Current paradigms usually either limit themselves to of- fload part of the computation and leave CPUs idle, or require static CPU/GPU work partitioning. We thus have previously proposed StarPU, a runtime system able to dynamically scheduling tasks within a single heterogeneous node. We show how we extended the task paradigm of StarPU with MPI to easily map the task graph on MPI clusters and automatically benefit from optimized execution.

Mots clés: Accelerators; GPUs; MPI; Task-based model; Accelerators

Domaines: Calcul parallèle, distribué et partagé [cs.DC]

Cédric Augonnet, Olivier Aumage, Nathalie Furmento, Raymond Namyst, Samuel Thibault. StarPU-MPI: Task Programming over Clusters of Machines Enhanced with Accelerators. EuroMPI 2012 - The 19th European MPI Users' Group Meeting, Sep 2012, Vienna, Austria. ⟨hal-00725477⟩