MPP Optimization

• Programming

  • Message passing (MPI, PVM, Shmem)
  • Shared memory (HPF or MP directive based)

• Algorithms

  • Data or Functional Parallelism
  • SIMD, MIMD
  • Granularity (fine, medium, coarse)
  • Master/Worker or Hostless

Previous slide

Next slide

Back to first slide

View graphic version