Abstract: This paper analyzes and quantifies the impact of context switching on CPU performance when operating system task scheduling, it concludes that the relation of the cost of mission context switching to the CPU efficiency. On the basis of the study on the influence relation, this paper proposes a feasible Hyper-Scheduling engineering method. By setting up a special task switching dedicated data path line in the CPU, and insert the parasitic logic in each key parts of CPU will be used to monitor the running status of the hardware resources, this method realizes the CPU hardware resources idle state for data preparation in advance or task. Exchange the special data path and main CPU data path line when task switching, which makes the context switch time can be close to zero, task switching cost can be eliminated. This means even in the situation of frequency context switching or time slice length is very short, CPU can remain in the maximized efficiency.