摘要
随着多核处理器的普及,并为了充分利用多核PC机的特性,计算机技术逐渐向多核架构及多核计算技术发展。为提高对湖南地区100m×100m小网格气温插值的速度,采用以OpenMP为标准的基于共享存储的并行编程模型对Kriging插值算法进行改进。在不同核的多核PC机中,采用100m×100m小网格和500m×500m小网格地形数据对平均气温进行插值,不仅有效减少了插值时间和提高了算法的加速比,而且集成到业务系统中大大提升了系统的反应时间及性能。
With the popularity of multi-core processors,full use of multi-core PC’s performance,computer technology is increasingly to multi-core architecture and multi-core computing technology development.To improve the small Hunan 100m × 100m grid interpolation rate of temperature,using the OpenMP standard for shared memory based parallel programming model to achieve the kriging interpolation algorithm based on OpenMP.Different multi-core PC,using a small grid 100m × 100m and 500m × 500m grid terrain data for a small average temperature for interpolation,not only effectively reduce the time and interpolation algorithms to improve the speedup,and significantly integrated into business systems enhance the system’s response time and performance.
出处
《计算机科学》
CSCD
北大核心
2012年第B06期392-395,共4页
Computer Science
