GPU-based Parallel R-tree Construction and Querying

被引:20
|
作者
Prasad, Sushil K. [1 ]
McDermott, Michael [1 ]
He, Xi [1 ]
Puri, Satish [1 ]
机构
[1] Georgia State Univ, Dept Comp Sci, Atlanta, GA 30303 USA
关键词
R-Tree; GPGPU; Recursion; Hierarchical Data Structure; Dynamic Parallelism; GIS; CUDA;
D O I
10.1109/IPDPSW.2015.127
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
An R-tree is a data structure for organizing and querying multi-dimensional non-uniform and overlapping data. Efficient parallelization of R-tree is an important problem due to societal applications such as geographic information systems (GIS), spatial database management systems, and VLSI layout which employ R-trees for spatial analysis tasks such as mapoverlay. As graphics processing units (GPUs) have emerged as powerful computing platforms, these R-tree related applications demand efficient R-tree construction and search algorithms on GPUs. This problem is hard both due to (i) non-linear tree topology of the data structure itself and (ii) the unconventional single-instruction multiple-thread (SIMT) architecture of modern GPUs requiring careful engineering of a host of issues. Therefore, the current best parallelizations of R-tree on GPU has limited speedup of only about 20-fold. We present a space-efficient data structure design and a non-trivial bottom-up construction algorithm for R-tree on GPUs. This has yielded the first demonstrated 226-fold speedup in parallel construction of an R-tree on a GPU compared to one-core execution on a CPU. We also present innovative R-tree search algorithms that are designed to overcome GPU's architectural and resource limitations. The best of these algorithms gives a speed up of 91-fold to 180-fold on an R-tree with 16384 base objects for query sizes ranging from 2k to 16k.(1)
引用
收藏
页码:618 / 627
页数:10
相关论文
共 50 条
  • [11] 2-Level R-tree Index Based on Spatial Grids and Hilbert R-tree
    GUO Jing LIU Guangjun DONG Xurong GUO Lei
    Geo-Spatial Information Science, 2006, (02) : 135 - 141
  • [12] 2-Level R-tree Index Based on Spatial Grids and Hilbert R-tree
    Guo Jing
    Liu Guangjun
    Dong Xurong
    Guo Lei
    GEO-SPATIAL INFORMATION SCIENCE, 2006, 9 (02) : 135 - 141
  • [13] 2-level R-tree index based on spatial grids and Hilbert R-tree
    Beijing Institute of Command and Technology of Equipment, 1 Jingjia Road, Huairou District, Beijing 101416, China
    Geo-Spatial Inf. Sci., 2006, 2 (135-141):
  • [14] A GPU-based Parallel Fireworks Algorithm for Optimization
    Ding, Ke
    Zheng, Shaoqiu
    Tan, Ying
    GECCO'13: PROCEEDINGS OF THE 2013 GENETIC AND EVOLUTIONARY COMPUTATION CONFERENCE, 2013, : 9 - 16
  • [15] A GPU-Based Parallel Algorithm for Landscape Metrics
    Zhong A.
    Chang L.
    Ma Y.
    Kang M.
    Mao Z.
    Wuhan Daxue Xuebao (Xinxi Kexue Ban)/Geomatics and Information Science of Wuhan University, 2020, 45 (06): : 941 - 948
  • [16] GPU-Based Parallel Processing Technology in DPI
    Zhong, Zhimin
    Zhang, Yuliang
    Yang, Guanglong
    Kong, Yongping
    WEB TECHNOLOGIES AND APPLICATIONS, APWEB 2015 WORKSHOPS, 2015, 9461 : 44 - 53
  • [17] GPU-based Parallel Implementation of SAR Imaging
    Jin, Xingxing
    Ko, Seok-Bum
    2012 INTERNATIONAL SYMPOSIUM ON ELECTRONIC SYSTEM DESIGN (ISED 2012), 2012, : 125 - 129
  • [18] GPU-based Parallel Particle Swarm Optimization
    Zhou, You
    Tan, Ying
    2009 IEEE CONGRESS ON EVOLUTIONARY COMPUTATION, VOLS 1-5, 2009, : 1493 - +
  • [19] The GPU-based parallel Ant Colony System
    Skinderowicz, Rafal
    JOURNAL OF PARALLEL AND DISTRIBUTED COMPUTING, 2016, 98 : 48 - 60
  • [20] Parallel GPU-based Plane-Sweep Algorithm for Construction of iCPI-Trees
    Andrzejewski, Witold
    Boinski, Pawel
    JOURNAL OF DATABASE MANAGEMENT, 2015, 26 (03) : 1 - 20