Parallel Computing: Lecture 3

Learning objectives

After this class, you should be able to:

1. Describe the following: (i) the task/channel model, (ii) the four steps of parallel algorithm design methodology (partitioning, communication, agglomeration, and mapping), and (iii) domain decomposition and functional decomposition.
2. Show how the boundary value problem of section 3.4.1 can be parallelized using the four-step process mentioned above, and derive an expression for the time taken by it.
3. Given a simple problem, parallelize it using the four-step process mentioned above, and derive an expression for the time taken by the parallel algorithm.

Reading assignment

1. Chapter 3, up to (and including) section 3.4.1.
2. Chapter 3, page 83-84 (especially figure 3.18).

Exercises and review questions

• Questions on current lecture's material
  1. Exercise 3.12. Also give an expression for the time taken by the parallel algorithm (after agglomeration) on an input of size n with p processors.

• Questions on next lecture's material
  1. (Post your solution on the discussion board) Give an example where six processors contain one integer each initially, and show the data on each processor after an all-gather operation.