High Performance Renderingon Massive Parallel Computers

• SPADE Project - Parallel Supercomputer

  • LSI - EPUSP - Brazil
  • cluster with 256 pentium-pro @ 200 MHz
    • quad-pentium node with
      • 512 MB memory
      • 8 GB disk
    • interconnection network
      • ATM, fast ethernet, myrinet

• Introduction Parallel Rendering DOMAIN Parallel Ray-Tracer Conclusion