论文部分内容阅读
首次提出了采用基于并行计算技术的高性能计算方法进行群库优化调度的方法，并采用客户／服务器模式在曙光１０００并行机上设计实现了淮河群库优化调度系统。结果表明：其调度的准确性大大优于常规调度，并缩短了整个决策时间，由此产生的经济效益和社会效益是难以估量的。
For the first time, this paper proposes a high performance computing method based on parallel computing technology to optimize the scheduling of group database. The client / server model is used to design and optimize Huaihe cluster optimization dispatching system based on Shuguang 1000 parallel machine. The results show that the accuracy of its scheduling is much better than that of the conventional scheduling and the whole decision-making time is shortened. The economic and social benefits arising therefrom are immeasurable.