Parallel performance optimization of large-scale unstructured data visualization for the earth simulator

This paper describes some efficient parallel performance optimization strategies for large-scale unstructured data visualization on SMP cluster machines including the Earth Simulator in Japan. The three-level hybrid parallelization is employed in our implementation, consisting of message passing for inter-SMP node communication, loop directives by OpenMP for intra-SMP node parallelization, and vectorization for each processing element (PE). In order to improve the speedup performance for the hybrid parallelization, some techniques, such as multi-coloring for removing data race and dynamic load repartition for load balancing, are considered. Good visualization images and high parallel performance have been achieved on Hitachi SR8000 for large-scale unstructured datasets, which shows the feasibility and effectiveness of our strategies.

[1]  Brian Cabral,et al.  Imaging vector fields using line integral convolution , 1993, SIGGRAPH.

[2]  C. R. Ramakrishnan,et al.  Optimal processor allocation for sort-last compositing under BSP-tree ordering , 1999, Electronic Imaging.

[3]  Li Chen,et al.  Parallel visualization of gigabyte datasets in GeoFEM , 2002, Concurr. Comput. Pract. Exp..

[4]  I. Fujishiro,et al.  Volumetric Data Exploration Using Interval Volume , 1996, IEEE Trans. Vis. Comput. Graph..

[5]  Kwan-Liu Ma,et al.  A scalable parallel cell-projection volume rendering algorithm for three-dimensional unstructured data , 1997, Proceedings IEEE Symposium on Parallel Rendering (PRS'97).

[6]  Lambertus Hesselink,et al.  Visualizing vector field topology in fluid flows , 1991, IEEE Computer Graphics and Applications.

[7]  Marc Levoy,et al.  Display of surfaces from volume data , 1988, IEEE Computer Graphics and Applications.

[8]  Kwan-Liu Ma,et al.  Parallel volume rendering using binary-swap compositing , 1994, IEEE Computer Graphics and Applications.

[9]  Jane Wilhelms,et al.  Octrees for faster isosurface generation , 1992, TOGS.

[10]  Li Chen,et al.  Comprehensible volume LIC rendering based on 3D significance map , 2002, IS&T/SPIE Electronic Imaging.

[11]  Hiroshi Okuda,et al.  Parallel Iterative Solvers for Unstructured Grids Using an OpenMP/MPI Hybrid Programming Model for the GeoFEM Platform on SMP Cluster Architectures , 2002, ISHPC.

[12]  Arie E. Kaufman,et al.  Parallel volume rendering of irregular grids , 1996 .

[13]  Arun K. Somani,et al.  Time and Space Optimal Data Parallel Volume Rendering Using Permutation Warping , 1997, J. Parallel Distributed Comput..

[14]  Yuriko Takeshima,et al.  Volume Data Mining Using 3D Field Topology Analysis , 2000, IEEE Computer Graphics and Applications.

[15]  Rolf Rabenseifner,et al.  Communication Bandwidth of Parallel Programming Models on Hybrid Architectures , 2009, ISHPC.

[16]  Marc Levoy,et al.  Efficient ray tracing of volume data , 1990, TOGS.

[17]  Lambertus Hesselink,et al.  Visualizing second-order tensor fields with hyperstreamlines , 1993, IEEE Computer Graphics and Applications.