基于GPU计算平台的三维波动方程叠前深度偏移

英文篇名：3D wave equation prestack depth migration on Graphic Processing Unit
中文刊名：石油地球物理勘探
英文刊名：Oil Geophysical Prospecting
作者：匡斌 ; 杜继修 ; 王华忠 ; 郭见乐 ; 单联瑜
英文作者：Kuang Bin1 ; Du Ji-xiu1 ; Wang Hua-zhong2 ; Guo Jian-le 1 and Shan Lian-yu1.1.Geophysical Research Institute ; Shengli Oilfield Branch Co. ; Sinopec ; Dongying ; Shandong 257022 ; China2.School of Marine Geology and Earth Science ; Tongji University ; Shanghai 200092 ; China
中文关键词：图形处理器 ; 统一计算设备架构 ; 波动方程叠前深度偏移
英文关键词：Graphic Processing Unit ; Compute Unified Device Architecture ; wave equation prestack depth migration
出版日期：2011-10-15
机构：胜利油田分公司物探研究院;同济大学海洋地质与地球科学学院;
年：2011
期：05
出版单位：石油地球物理勘探

摘要

GPU/CPU协同并行计算是一项全新的、具有革命性的技术,将对地震资料处理行业产生深远的影响。本文在充分发挥GPU存储带宽宽、多寄存器和多处理器结构特点的基础上,通过改进地震成像并行计算模块的核心函数,形成一套波动方程地震成像GPU高效并行算法,并取得理想的加速比和可扩展性。所形成的一套GPU/CPU平台下波动方程叠前深度偏移软件,已投入地震资料处理实际应用。理论模型试算和实际资料的处理验证了本方法的有效性和实用性。
GPU/CPU Co-Processing Parallel Computing(GPPC) is a new revolutionary technology,which will make great contributions on seismic data processing.Based on the hardware structural characters of Graphic Processing Unit including wide memory band,multi-registers and multi-processors,this paper introduces a wave equation imaging parallel method with high efficiency by improving on the kernel function of seismic imaging parallel calculation modules,and produces ideal speed-up ratio and expansibility.The study develops one wave equation PSDM technology and software based on GPPC,which have been applied in seismic data processing.And the tests on the SEG/EAGE salt model and 3D real seismic datasets prove the usefulness and practicability of this method.

引文

[1]Francisco Ortigosa,Repsol-YPF,Mauricio Araya-Po-lo,F′elix Rubio.Evalution 3DRTM on HPC Plat-forms.SEG Technical Program Expanded Ab-stracts,2008,27:2879～2883
    [2]Martin Kaser,Josep de la Puente,Cristobal Castro.Seismic wave field modeling using high performancecomputing.SEG Technical Program Expanded Ab-stracts,2008,27:2884～2889
    [3]李博,刘国峰,刘洪.地震叠前时间偏移的一种图形处理器提速实现方法.地球物理学报,2009,52(1):245～252Li Buo,Liu Guo-feng,Liu Hong.A method of usingGPU to accelerate seismic pre-stack time migration.Chinese J Geophys,2009,52(1):245～252
    [4]夏健明,魏德敏.用GPU加速求解线性方程组的高斯消元法.计算机工程与设计,2009,30(19):4447～4450Xia Jian-ming,Wei De-min.Accelerating Gaussianelimination method for linear systems on GPU.Com-puter Engineering and Design,2009,30(19):4447～4450
    [5]夏健明,魏德敏.共轭梯度法的GPU实现.计算机工程,2009,35(17):274～276Xia Jian-ming,Wei De-min.GPU implementation ofconjugate gradient method.Computer engineering,2009,35(17):274～276
    [6]刘伟峰,王永胜,张天雷,张兵.使用GPU模拟地震波传播的性能研究.系统仿真学报,2009,21(增刊1):170～174Liu Wei-feng,Wang Yong-sheng,Zhang Tian-lei,Zhang Bing.Performance study of seismic wave propa-gation simulation using GPU.Journal of System Sim-ulation,2009,21(Suppl 1):170～174
    [7]石颖,刘洪.基于GPU的表面多次波预测技术.石油地球物理勘探,2010,45(4):540～544Shi Ying,Liu Hong.GPU-based surface multiple pre-diction technique.OGP,2010,45(4):540～544