Introduction to Parallel Programming

This course will teach how to optimize and parallelize Fortran 77/90 and C programs, and how to run them on high-performance computing facilities available both:

on campus, such as the SGI Origin, the IBM multiprocessor machine, and Beowulf clusters, and
off campus, such as SHARCNET, which have a mandate to share a certain percentage of their resources.

Portable standards exist for both shared (SMP) and non-shared (MPI) memory parallel programming, as do portable mathematical libraries (NAG, etc.). These will be illustrated using computationally-intensive examples, such as fluid dynamics.

The course will be interspersed with hands-on exercises. Feel free to bring your own code to work on, too.

Course Outline

Introduction to Parallel Programming (.ppt)

Animation of MPI and SMP programs

Examples

MPI Information and MPI Function Calls and Talk on MPI-2 (Ed Chrzanowski, CSCF).

OpenMP Information and Talk on OpenMP (Ed Chrzanowski, CSCF)

IBM SP Course Notes

Fortran Tutorials

Scheduling Algorithms , an informative presentation by Ed Chrzanowski to CS 775 in Fall 2002.

Parallel Computing 2 (Course Notes) by Edward Chrzanowski, CSCF.

Parallel Programming from Concept to Compile (Course Notes) by Edward Chrzanowski, CSCF.

Books on Parallel Computing

Timing

Math & Statistical

Introduction to Parallel Programming