CS267: Lecture 3
Parallel Machine Models
September 5, 2001
Lecturer: Kathy Yelick
We begin with a continuation of Lecture 2, examining optimizations of matrix
multiplication (search-based blocking and Strassens' algorithm), as well as a
set of techniques for optimizing serial programs. We then move on to the
main topic of today's lecture: parallelism models. We discuss several
models of parallel machines and the programming models that most naturally fit
on those machines. We give an overview of shared memory, distributed
memory, and data parallelism. We note that one machine may support several different
models, however, and the models typically differ in how complicated they are to
use and how well they perform.
2001 Lecture Notes
(due 9/19/01). We have assigned "multidisciplinary" teams of 2-3 students
for this assignment. If you are not in a team, please contact David Bindel (dbindel@cs).