首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于CUDA技术的矿产储量计算
引用本文:章 浩,庞振山,姚长利,张明华.基于CUDA技术的矿产储量计算[J].地质通报,2009,28(203):216-223.
作者姓名:章 浩  庞振山  姚长利  张明华
作者单位:1. 地质过程与矿产资源国家重点实验室和地下信息探测技术与仪器教育部重点实验室,2. 中国地质调查局发展研究中心
摘    要:插值计算是矿产储量计算中的重要环节,但由于计算量大,经常需要在插值密度和计算时间之间进行权衡。本项研究通过最新的CUDA技术,将插值运算转换到显卡上进行并行化计算,极大地提高了计算速度。以紫金山铜矿数据为研究对象,进行了方法技术的实验,与基于CPU的计算相比,基于GPU下CUDA的计算速度提高了67倍。

关 键 词:储量计算  GPU  CUDA  距离平方反比法  插值  并行处理

Mineral reserve calculation based on CUDA technology
ZHANG Hao,PANG Zhen-shan,YAO Chang-li,ZHANG Ming-hua.Mineral reserve calculation based on CUDA technology[J].Geologcal Bulletin OF China,2009,28(203):216-223.
Authors:ZHANG Hao  PANG Zhen-shan  YAO Chang-li  ZHANG Ming-hua
Institution:State Key Laboratory of Geological Process and Mineral Resources, Key Laboratory of Geo-detection, Ministry of Education,2. Development and Research Center of China Geological Survey
Abstract:Interpolation is playing an important role in evaluation of mineral reserves, and is also an overwhelming computation that makes it needed to find a tradeoff between the density of interpolating and the time of calculating. The research is taken to move the interpolation from CPU to graphic card by means of CUDA, and has improved much the speed of the calculation. An experiment on evaluation the Zijinshan copper mine data showed the merit, and a result of acceleration of 67 times was achieved by compared the calculations of CPU based and GPU based.
Keywords:interpolation calculation  GPU  CUDA  inverse distance squares method (IDS)  Interpolation  parallel processing
点击此处可从《地质通报》浏览原始摘要信息
点击此处可从《地质通报》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号