CPU的多层次并行调度优化模型仿真

GPU Multi-Level Parallel Scheduling Optimization Model for Simulation

通过CPU多调度模式优化,提高CPU运行效率。由于海量数据进行运算的过程中,存在调试不均衡的问题,传统的CPU调度模型不能很好的均衡所有的调度任务关系,无法满足数据运算的实际需求,导致CPU负载不均衡,降低了调度效率。提出基于二叉树搜索算法的CPU多层次并行调度方法。针对CPU中不同层次的任务量进行预测,建立多层次并行调度模型,实现海量调度任务的多层次并行调度。在每个层次中,进行二叉树搜索,完成各个层次独立的CPU任务调度,将二叉树搜索方法运用到多层次并行调度模型中,完成CPU的多层次并行调度。实验结果表明,利用改进算法进行CPU多层次并行调度,能够提高调度效率,缩短调度时间,完成CPU合理调度,保证CPU的运算速率。

CPU scheduling model affects the operating efficiency more directly. For huge amounts of data in the process of operation,the traditional CPU scheduling model cannot balance all the relations of scheduling tasks very well, cannot satisfy the needs of data operation ,which causes the CPU load imbalance and reduces the scheduling effi- ciency. The paper proposed a multi - level parallel CPU scheduling method based on binary tree search algorithm. Aiming at different levels in CPU quota, the paper established a multi - level parallel scheduling model to realize the multilevel parallel scheduling of mass scheduling tasks. In each level, the binary tree search completed all levels in- dependent of CPU scheduling, and the binary tree search method was applied to the multi - level parallel scheduling model, to complete CPU multi -level parallel scheduling. Experimental results show that the improved algorithm for CPU multi - level parallel scheduling, can improve the efficiency of the scheduling, shorten operation time, complete the reasonable CPU scheduling, and ensure the operation rate of the CPU.