摘要
采用基于列交换的Gauss-Jordan并行算法来解决空气动力学中超音速高阶面元法的稠密矩阵求逆问题,该方法采取了块循环数据分配方式,尤其对超立方体结构的并行机系统来说具有通讯优势。在4台SGI工作站构成的2×2网格上进行的实验表明,对秩为1000左右的矩阵可得到57%~64%的效率。
In this paper we propose GaussJordan algorithm using column interchanges for computing a highorder dense matrix inverse produced by the highorder panel method in aerodynamics. The algorithm implements the block cyclic data distribution,which especially has communication advantages for the hypercube network.Results are achieved when running this program on the 4 SGI workstations configured as a 2×2 processor grid. It shows that 57%~64% efficiency can be gained when solving a matrix is about 1000 order.
出处
《国防科技大学学报》
EI
CAS
CSCD
1997年第4期1-4,共4页
Journal of National University of Defense Technology
基金
国家自然科学基金
863计划基金
关键词
超音速流
高阶面元法
矩阵求逆
G-J并行算法
supersonic flow, highorder panel method,matrix inversion,GaussJordan algorithm,SGI