摘要
OpenMP作为共享存储并行编程标准,以其良好的易用性、支持增量并行等优点,成为并行程序设计的主流模型之一。本文主要探讨了铸造数值模拟温度场在重力补缩与非重力补缩条件下的并行计算。实例表明,在支持超线程的硬件环境中,使用OpenMP技术可以取得良好的并行性,大大提高运算效率。
As an industrial standard for shared-memory programming, with excellent programmable ability and portability, OpenMP becomes one of the mainstream parallel programming models. This parallel computations of temperature field in the casting simulation with and without the gravity feeding were investigated. The results show a fair parallel performance can be achieved by using the OpenMP technology in the hype-threading hardware environment, thus improves the efficiency greatly.
出处
《铸造技术》
CAS
北大核心
2007年第9期1230-1234,共5页
Foundry Technology
基金
教育部新世纪优秀人才支持计划(2005年)
宁波市科技合作计划项目(2005B10037)
