• DocumentCode
    3278579
  • Title

    MPI and POV-Ray parallel solution to improve performance of photo-realistic rendering utilizing Albacore and Sequoia

  • Author

    O´Keeffe, Shawn E. ; Breland, Jason ; Shiratuddin, Mohd Fairuz

  • Author_Institution
    Sch. of Constr., Univ. of Southern Mississippi, Hattiesburg, MS, USA
  • Volume
    2
  • fYear
    2012
  • fDate
    12-14 June 2012
  • Firstpage
    973
  • Lastpage
    978
  • Abstract
    The purpose of this research is to improve current photo-realistic rendering techniques of 3D computer generated images. Recent serial program techniques include `rendering-farms´ that consist of numerous topologically linked computers to improve performance, cost, and rendering time. The contribution of this research is to utilize the rendering farm philosophy, by utilizing parallel computing techniques and multi-thread processing. This research correlates the new parallel performance with traditional serial approach. This paper presents the cost, parallel overhead, and efficiency of rendering on Sequoia. The implementation of higher-level languages and how a message can be passed efficiently are described. A comparison between both serial and parallel code has also been made.
  • Keywords
    image processing; message passing; multi-threading; rendering (computer graphics); 3D computer generated image; Albacore; MPI-POVRAY parallel solution; Sequoia; cost; higher-level language; message passing; multithread processing; parallel code; parallel computing; parallel overhead; parallel performance; photo-realistic rendering technique; rendering efficiency; rendering farm philosophy; rendering time; serial code; serial program technique; Rendering (computer graphics);
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Computer & Information Science (ICCIS), 2012 International Conference on
  • Conference_Location
    Kuala Lumpeu
  • Print_ISBN
    978-1-4673-1937-9
  • Type

    conf

  • DOI
    10.1109/ICCISci.2012.6297167
  • Filename
    6297167