High Performance Distributed Interactive Visualization - PowerPoint PPT Presentation

1 / 11
About This Presentation
Title:

High Performance Distributed Interactive Visualization

Description:

Enhanced ipaq via high performance clusters via wireless network. ... pov desc generator. How about the performance. Improvements with Remote Visualization: ... – PowerPoint PPT presentation

Number of Views:49
Avg rating:3.0/5.0
Slides: 12
Provided by: ccGa
Category:

less

Transcript and Presenter's Notes

Title: High Performance Distributed Interactive Visualization


1
High Performance DistributedInteractive
Visualization
  • Advisors
  • Dr. Karsten Schwan
  • Dr. Matthew Wolf
  • Zhongtang Cai
  • Weiyun Huang

www.cercs.gatech.edu
2
Research Application Molecular Dynamics
  • Mechanical Engineering
  • Physics
  • Chemistry
  • Aerospace Engineering

3
Introduction
  • Collaborate between highly heterogeneous
    platforms
  • Jedi clusters
  • SGI graphics servers
  • Desktop workstations
  • Handhelds like ipaq
  • iPaqs ipaq as a control portal of other
    large-scale heterogeneous platform, e.g.
    Immersadesk.
  • Remote Visualization and interaction for ipaq.
    Enhanced ipaq via high performance clusters via
    wireless network.
  • ECho high performance event channels of large
    volume data sets
  • Proactive Directory Service. Service directory
    and shared data consistency automatism.

4
ECho Event-based Data Exchange
  • Represent information flows as event streams
    (publish/subscribe model)
  • Transparency of local and remote receivers via
    event channels, implemented with underlying
    peer-to-peer communications over TCP, UDP,
    Wireless, multicast, etc.
  • Efficient, fully typed binary data transmission,
    based on dynamically defined event formats (PBIO)
  • Dynamic extension of existing formats and
    discovery/operation on format contents
    (reflection)
  • Interoperate with CORBA and Java (JECho) via IDL
    and XML

5
ECho Source-based Filtering Achieving Data
Personalization via Dynamic Code Generation
Sink
F( )
Source
Filter Function
if ((data inside viewing area) and
((data values inside useful ranges))
return 1 / submit into channel /
return 0 / suppress data /
6
Overview of the Smart PointerVisualization System
Filters
7
Remote Visualization
8
How about the performance
  • Improvements with Remote Visualization
  • Response Time to render a 3D scenario
  • On Jedi 1s -2s
  • Atoms 1000
  • Bonds 100000
  • Rendering algorithm Ray-tracing
  • On iPaq 5s-7s
  • Atoms 10
  • Bonds 0
  • Rendering algorithm Simple shading model
  • If Atoms 1000, Bonds 100000?

- Ipaq will keep silent
9
A Real Demo in this Room(Everyone is welcome to
play with!)
10
References
  • Event Services for High Performance Computing,
    Greg Eisenhauer, Fabian Bustamente and Karsten
    Schwan, Proceedings of High Performance
    Distributed Computing (HPDC-2000)
  • Collaborative distributed virtual
    environmentTerrence.Disz et al
  • Daniel A. Reed and C. E. Catlett. Distributed
    data and immersive collaboration. Communications
    of the ACM, 40(11)3849, November 1997.
  • GridPort https//gridport.npaci.edu
  • Access Grid https//www.accessgrid.org

11
Thanks
Write a Comment
User Comments (0)
About PowerShow.com