Designing Parallel Loop Self-Scheduling Schemes by the Hybrid MPI and OpenMP Model for Grid Systems with Multi-Core Computational Nodes

Abstract

Loop scheduling on parallel and distributed systems has been thoroughly investigated in the past. However, none of them considers the feature of multicore architecture dominating the current markets of desktop computers, laptop computers, servers, etc. On the other hand, although there have been many studies proposed to employ the hybrid MPI and OpenMP programming model to exploit different levels of parallelisms for the distributed system with multicore computers, none of them aimed at the design of parallel loop self-scheduling. Therefore, this paper investigates how to employ the hybrid MPI and OpenMP model to design parallel loop self-scheduling scheme to adapt to the feature of multicore architecture for emerging grid systems. The proposed scheduling approach is based on our previous work adopting the pure MPI model. Preliminary experimental results show that the proposed approach outperforms the previous work with the average speedup of 3.39.