Optimal Control of Reference Streams in Parallel Processing Systems

 The optimal control of memory space to raise speed of parallel processing systems is a scientific and practical issue, which is very important in many applied branches and fields. One of the major factors which reduces the efficiency of the parallel processing system is the conflict in accessing to the common memory. The article studies the proposal of a shared memory supplemented with a buffer structure at the input. This model allows the optimum of the shared memory via the motive supply method. It is easy to use the FPGA technology to re-build the queuing strcuture according to the parameter with size m to optimize the memory structure for problems as a solution to raise the efficiency and the speed. The model allows the use of memories with big sizes but unlimited speeds in order to increase the reliability for the parallel processing systems