Authors : Karanjeet Singh Kahlon O.P. Gupta and Rakesh Jindal
Abstract: GPCC is a Grid enabled PC Cluster system developed for harnessing computational resources available over LAN for parallel computing. Parallel computing is essential for solving very large scientific and engineering problems. An effective parallel computing solution requires an appropriate parallel machine and a well-optimized parallel program. Most parallel matrix multiplication algorithms use matrix decomposition based on square root of the number of processors available. In this reserch, a grid enabled PC Cluster based on Windows OS integrating the DeskGrid API to communicate among the processors is u The tasks are allocated to the processors based on the log of the processor status and errors committed. sed. The decomposition of the matrix is independent of the no. of processors available for the computation. Dynamic allocation of the matrix pointer not only proved to be distributed memory efficient but also time efficient. This reserch is an attempt to highlight the ease and stable nature of the Grid enabled PC Cluster (GPCC) based on Windows OS.
Karanjeet Singh Kahlon O.P. Gupta and Rakesh Jindal , 2006. Parallel Matrix Multiplication on Grid Enabled PC Cluster. Asian Journal of Information Technology, 5: 1210-1212.