Discretization of 2-D Active Thermal Control Problem on Parallel Virtual Machine (PVM) with Stationary Iterative Methods

We present the discretization of 2-D Active Thermal Control Process (2D-ATCP) running on a Master-Slave platform. The parallel implementation was carried out using Parallel Virtual Machine (PVM). The 2-D ATCP was discretized using stationary iterative methods. Parallel compare of these methods ith parallel algorithms were experimentally evaluated and numerical results show their effectiveness. We show that the schemes have scalability performance with less information to schedule tasks. The results show that various block sizes on various domain of decomposition can significantly improve system throughput and the platform can be used for more sophisticated cluster scheduling on different dimensions of ATCP.