In the paper, we propose a generic parallel computing model for the distributed computing platforms. This model is built on the top of the popular CORBA middleware in order to ensure easy distribution of the software components along the participating computers. We use the concept of factory objects in the model to create slave objects dynamically to fulfil the master/slave parallel computing paradigm. The speedup of the experiment programs using our proposed model is measured and compared with the corresponding MPI implementation. The experiment results show that our model is not only a feasible model for designing parallel programs with CORBA but also an efficient way in comparison to the MPI implementation
Published in:
Parallel and Distributed Computing, Applications and Technologies, 2006. PDCAT '06. Seventh International Conference on
Date of Conference: Dec. 2006