水动力学数学模型并行计算技术研究及实现

Research and application of parallel computing technology on hydrodynamic model

基于MPI(Message Passing Interface)消息传递实现了黄河二维水动力学数学模型的并行编程。以数据的分布存储作为区域划分的依据,实现了计算量的负载平衡;在全局网格和局部区域之间建立映射关系,并且在临界单元、进出口单元、共用节点等通过规约等进行特殊处理,一方面尽可能的减少通讯量,另一方面也避免消息的阻塞。采用曙光4000A系统的8个CPU进行计算的加速比达0.8,多CPU并行计算极大的提高了计算任务的容量,使得无法在单机上完成的巨量计算成为可能。

Parallel program design of Yellow River flow-sediment model was realized based on MPI（ Message Passing Interface）. Load balance of computing quantity was realized through regional division according to distributive storage of data. The mapping relation between whole grids and part-region was established. Special treatments were made at critical units through protocols, input ＆ output units and shared nodes. These treatments can decrease communication as much as possible and avoid information-jam. The parallel program can meet the demands of computing capacity and speed in the Yellow River flow-sediment mathematical model.