A Cluster Based Load-Balancing Method for Multi-Computer Based Solution-Adaptive Finite Element Graphs

摘要

Load imbalance can degrade the performance of a solution-adaptive finite element application program on a distributed memory multicomputer. To solve the load imbalance problem load of a refined finite element graph can be redistributed based on the current load of each processor. In this study, a load-balancing algorithm is applied to balance the computational load of each processor. A distributed method for load balancing is proposed which is based on the global load balancing information and current load distribution of the system. A simulation model is developed to compare the performance of the proposed method with the existing methods such as MCSTLB, BTLB and CBTLB methods. The execution time and the number of process migration required by different load balancing methods are used for performance evaluation. The experimental result shows that the proposed method is more efficient than that of the existing methods.