Collaborative Computing: Networking, Applications and Worksharing. 4th International Conference, CollaborateCom 2008, Orlando, FL, USA, November 13-16, 2008, Revised Selected Papers

Research Article

Collaborative Graphic Rendering for Improving Visual Experience

Download
451 downloads
  • @INPROCEEDINGS{10.1007/978-3-642-03354-4_45,
        author={Xiaoxin Wu and Guodong Pei},
        title={Collaborative Graphic Rendering for Improving Visual Experience},
        proceedings={Collaborative Computing: Networking, Applications and Worksharing. 4th International Conference, CollaborateCom 2008, Orlando, FL, USA, November 13-16, 2008, Revised Selected Papers},
        proceedings_a={COLLABORATECOM},
        year={2012},
        month={5},
        keywords={},
        doi={10.1007/978-3-642-03354-4_45}
    }
    
  • Xiaoxin Wu
    Guodong Pei
    Year: 2012
    Collaborative Graphic Rendering for Improving Visual Experience
    COLLABORATECOM
    Springer
    DOI: 10.1007/978-3-642-03354-4_45
Xiaoxin Wu1,*, Guodong Pei1,*
  • 1: Intel China Research Center Ltd
*Contact email: xiaoxin.wu@intel.com, guodong.pei@intel.com

Abstract

Handheld devices such as UMPC, though convenient, bear weakness of size constraint for display. To mitigate such a problem and enhance user experience for owners of small devices, in this paper we design a collaborative rendering platform. When running game graphic applications at a handheld, the generated OpenGL graphic commands are intercepted and then delivered to a device with a larger display. The graphics are rendered and displayed at that device. The performance of the collaborative rendering platform is determined by graphic computing resources and network bandwidth. Analysis and simulation prove that other than providing a better display, the collaborative system can improve game experience also by increasing frame rates. In particular, at a low computing cost, a further collaboration between GPUs of collaborators can improve frame rate by eliminating the negative impact from network delay on applications that require GPU feedback.