PARALLEL COMPUTING FOR THE 2-DIMENSION ENERGY EQUATIONS WITH THREE TEMPERATURES

摘要：

In this paper, we discuss the parallel computing questions for the 2-dimension energy equations with three temperatures. The parallel computing of the precon-ditioner is the bottleneck for the parallel computing of Krylov Subspace Methods. Based on the norm-minimization technique, a parallel sparse approximate inverse preconditioning method (SAIP method) is given for the unsymmetric sparse linear systems. The SAIP preconditioner has the fixed sparse pattern same as AT. Then we use SAIP-Krylov Subspace Methods to solve the large sparse linear systems from the 2-dimension energy equations with three temperatures, and the results show that the SAIPKrylov Subspace Methods have better convergence perfor-mance than the ParaSails toolkit. In the last of this paper, we design a parallel algorithm for 2dimension energy equations with three temperatures and discuss the questions of parallel implement, then we give the results of parallel computing.

关键词： 2-dimension energy equations with three temperatures precondi-tioner Krylov Subspace Methods parallel computing

作者: LIU JIE CHI LIHUA HU QINGFENG MO ZEYAO LI XIAOMEI

作者单位: Computer College, National University of Defense Technology, Changsha, China, 410073 Beijing Institute of Applied Physics and Computational Mathematics, 100088 Command Technology Institute, Beijing, 101416

会议类型: 国际会议

会议名称: The Joint Conference of ICCP6 and CCP2003(第6届国际计算伦理会议)

会议地点: 北京

会议语种:英文

页码: 343-346

在线出版日期: 2004-05-23（万方平台首次上网日期，不代表论文的发表时间）