A New Threads Scheduling Algorithm For Multi-core Architecture And Its Simulation

As multi-core processor begin to emerge in every area of computing, and multi-core processors with tens of hundreds of cores begin to proliferate, operating system scheduling that take the peculiarities of such architectures into account will become mandatory. Due to architectural differences to traditional multi-processors, such as shared L2 cache, memory controllers and smaller cache size available per computa-tional unit, it does not suffice to simply schedule tasks on multi-core processor in the same way as on SMP systems.To make the best use of the computational power available, it is essential to assign the tasks dynamically to the CPUs and cores, and balance the load after a thread or a task has finished. Here we develop a model to allocate the task to the CPUs and cores and balance the load between the CPUs and cores to maximum the throughput. Running cost(RC),communicating cost(CC) and migrate cost(MC) have been taken into account, besides, simulation programs in Python have been developed for the algorithm on Ubuntu platform. this model is suitable for arbitrary number of CPUs and cores in a CPU.