针对油气勘探地震资料处理 台式协同并行机研制.pdfVIP

针对油气勘探地震资料处理的 台式协同并行机研制 1 2 1 2 1 刘国峰 刘 钦 李 博 佟小龙 刘 洪 1 中国科学院地质与地球物理研究所,北京,100029 2 北京吉星吉达科技有限公司,北京,100029 摘 要 随着图形处理器（Graphic Processing Unit: GPU ）在通用计算领域的日趋成熟，使台式协同并 行计算机的研制成为可能，并将在油气勘探地震资料处理中，对诸多大规模计算的关键性环节得以成功应用。 本文阐明协同并行计算机研制的思路、架构及编程环境, 着重分析其计算效率得以大幅度提升的关键所在。 文中以地震资料处理中的叠前时间偏移和 Gazdag 深度偏移为切入点，展示所研制样机测试结果的图像显示。 显而易见，生产实践中，时常面临对诸多算法进行算法精度和计算速度之间的折中选择。本文阐明的研制样 机具有高并行度，进而却可在算法精度与计算速度的优化配置协调上获得广阔空间。笔者认为，本文的台式 协同并行机研制思路及架构，或可作为地球物理配置高性能计算机全新选择的一项依据。 关键词 CUDA ／GPU 协同并行计算，叠前时间偏移，Gazdag 深度偏移，台式协同并行机研制 The Development of parallel computer on the desk for oil and gas exploration seismic data processing 1 2 1 2 1 Liu Guo-feng Liu Qin Li Bo Tong Xiaolong Liu Hong ( 1Institute of Geology and Geophysics ，Chinese Academy of Sciences，Beijing ，100029) (2 Beijing Geo-star Science and Technology Ltd.Beijing,100029 ) Abstract: With the development of (GPU) Graphic Processing Unit in general purpose ,it makes the parallel computer on the desk come to true , and it also will play a import role in many massive computing module in oil and gas seismic data processing .In this paper ,we introduce the idea ,architecture and coding environment for GPU and CPU co-processing with CUDA, at the same time we pay more attention on point out why this method can greatly improve the efficiency. we use the pre-stack time migration and Gazdag depth migration in seismic data processing as the first point of this paper, and the figures can tell us the result of these tests. Obviously ,we often front the situation that we must abandon some good method because of some problem about comput