利用CUDA的地图代数局部算子优化
|
摘要
本文针对地图代数局部算子的传统实现方法应用于海量栅格数据计算时效率低下的问题,从串行算法的并行化映射、计算机图形处理器资源的自适应参数调整等多角度来研究地图代数空间并行算法的实现机制,总结出地图代数局部算子在GPU并行处理架构上的通用求解步骤。实验结果表明,该方法在大数据量处理时较CPU加速效果明显。
Aiming at the low efficiency when traditional realization methods of local arithmetic operators of map algebra apply to calculations for gigantic raster data,this paper discussed the realization mechanism from surrounding parallel mapping of serial algorithms and adaptive parameter adjustments on computer graphic processor resources.Then it summarized the common solving procedures of implementing those local operators on GPU parallel processing architecture.The experimental results demonstrated that the proposed method could achieve better accelerating effect than that of CPU.
Aiming at the low efficiency when traditional realization methods of local arithmetic operators of map algebra apply to calculations for gigantic raster data,this paper discussed the realization mechanism from surrounding parallel mapping of serial algorithms and adaptive parameter adjustments on computer graphic processor resources.Then it summarized the common solving procedures of implementing those local operators on GPU parallel processing architecture.The experimental results demonstrated that the proposed method could achieve better accelerating effect than that of CPU.
引文
[1]胡鹏,游涟,胡海.地图代数概论[M].北京:测绘出版社,2008.
[2]Tomlin C D,Cartographic Modeling.n Maguire D,Good-child M F,and Rhind D.(Eds.)Geographic Informa-tion Systems:Principles and Applications[M].Lon-don:Longman:361-374.
[3]PULLAR D,MapScript:A Map Algebra ProgrammingLanguage Incorporating Neighborhood Analysis[J].GeoInformatica,2001,5(2):145-163.
[4]Gilberto C,Danilo P and Ricardo C.Towards a general-ized map algebra:principles and data types[M].2007.
[5]Mennis J,Viger R and Tomlin C D.Cubic Map AlgebraFunctions for Spatio-Temporal Analysis[J].Cartogra-phy and GIS,2005,32(1):17-32.
[6]Yacine Amara,Xavier Marsault.A GPU Tile-Load-Maparchitecture for terrain rendering:theory and applica-tions[J].Vis Comput,2009,25:805-824.
[7]Antonio Ruiz,Manuel Ujaldon,Lee Cooper,etc.Non-rigid Registration for Large Sets of Microscopic Images onGraphics Processors[J].Sign Process Syst,2009,55:229-250.
[8]Adrian Nistor,Wei-Ngan Chin,Tiow-Seng Tan,etc.Optimizing the Parallel computation of linear recur-rences using compact matrix representations[J].Par-alle Distrib Comput,2009,69:373-381.
[9]Weiguo Liu,Bertil Schmidt,Gerrit Voss,etc.Acceleratingmolecular Dynamics simulations using Graphics Processing U-nits with CUDA[J].Computer Physics Communications,2008,179:634-641.
[10]吴连贵,易瑜,李肯立.基于CUDA的地震数据相干体并行算法[J].计算机应用,2009,29(3).
[2]Tomlin C D,Cartographic Modeling.n Maguire D,Good-child M F,and Rhind D.(Eds.)Geographic Informa-tion Systems:Principles and Applications[M].Lon-don:Longman:361-374.
[3]PULLAR D,MapScript:A Map Algebra ProgrammingLanguage Incorporating Neighborhood Analysis[J].GeoInformatica,2001,5(2):145-163.
[4]Gilberto C,Danilo P and Ricardo C.Towards a general-ized map algebra:principles and data types[M].2007.
[5]Mennis J,Viger R and Tomlin C D.Cubic Map AlgebraFunctions for Spatio-Temporal Analysis[J].Cartogra-phy and GIS,2005,32(1):17-32.
[6]Yacine Amara,Xavier Marsault.A GPU Tile-Load-Maparchitecture for terrain rendering:theory and applica-tions[J].Vis Comput,2009,25:805-824.
[7]Antonio Ruiz,Manuel Ujaldon,Lee Cooper,etc.Non-rigid Registration for Large Sets of Microscopic Images onGraphics Processors[J].Sign Process Syst,2009,55:229-250.
[8]Adrian Nistor,Wei-Ngan Chin,Tiow-Seng Tan,etc.Optimizing the Parallel computation of linear recur-rences using compact matrix representations[J].Par-alle Distrib Comput,2009,69:373-381.
[9]Weiguo Liu,Bertil Schmidt,Gerrit Voss,etc.Acceleratingmolecular Dynamics simulations using Graphics Processing U-nits with CUDA[J].Computer Physics Communications,2008,179:634-641.
[10]吴连贵,易瑜,李肯立.基于CUDA的地震数据相干体并行算法[J].计算机应用,2009,29(3).
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心