A Hierarchical Distributed Shared‐Memory Parallel Branch & Bound Application with PVM and OpenMP for Multiprocessor Clusters
R. Aversa, Beniamino Di Martino, Nicola Mazzocca, Salvatore Venticinque · 2005
This chapter contains sections titled: Introduction The Branch&Bound Parallel Application The OPENMP Extension Experimental Results Conclusions Reference Branch & Bound applications represent a typical example of irregularly structured problems whose parallelization using hierarchical computational architectures (e.g. clusters of SMPs) involves several issues. In this paper we show how the combined use of PVM and OpenMP libraries allow us to develop hybrid code in order to introduce an additional dynamic distribution among the shared memory nodes of the system. Both coarse-grain and fine-grain parallelization are based on the coordinator/workers paradigm.