Parallel Programming - Projects

HW1 : Odd - Even Sort

• Spec

• Use MPI to complete odd-even sort across different processes.
  Github repository - OddEvenSort

HW2 : Mandelbrot Set

• Spec
• Increase the computation time of sequential Mandelbrot Set using Dynamic Load Balancing and Vectorization.

• Complete with two version: 2a : Pthread only 2b : Hybrid with MPI and OpenMP

• Ranked 5th out of 64 students in the class
• Handcrafted dynamic task assignment with Master and Slave workpool using MPI Send() and Recv().
  Github repository - Mandelbrot Set

HW3 : All-Pair Shortest Path (CPU version)

• Spec
• Number of Vertices: from 2 to 6000
• Number of Edges: from 0 to V*(V-1)
  Github repository - hw3-PP

HW4-1: Block All-Pair Shortest Path (GPU CUDA version)

• Spec
• Number of Vertices: from 2 to 40000
• Number of Edges: from 0 to V*(V-1)

HW4-2: Block All-Pair Shortest Path (Multi-GPU version)

• Spec
• Number of Vertices: from 2 to 60000
• Number of Edges: from 0 to V*(V-1)
  Github repository - hw4

Final Project: Parallelize RRT*

• Parallelize sequential version RRT and RRT* algorithms.

Presentation Slide

Github repository - RRT*-PPFinal