Parallel Programming for HPC at UNSW Kensington

Why do this course?

You have written, compiled and run functioning programs in C and/or Fortran. You know how HPC works and you've submitted batch jobs.

Now you want to move from writing single-threaded programs into the parallel programming paradigm, so you can truly harness the full power of High Performance Computing.

Who is this course for?

This course is targeted at C and Fortran programmers. It gives an introduction into different parallel programming methods. OpenMP as an wide-spread method for a shared memory programming model and MPI as the standard for a distributed memory programming model are discussed.

The course includes examples and exercises. Attendees can choose between C and Fortran for the exercises. All exercises will be done on our local HPC machine. Student accounts will be provided. Students must be fluent in C or Fortran and be familiar with a Linux environment.

You'll learn how to program with:

• OpenMP (Open Multi-Processing): a widespread method for shared memory programming

• MPI (Message Passing Interface): a leading distributed memory programming model

Prerequisites

To do this course you need to have:

• A good working knowledge of HPC. Consider taking our Unix Shell and Command Line Basics course to come up to speed beforehand.

• Prior experience of writing programs in either C or Fortran.

The Intersect approach to training

At Intersect, we work closely with our member universities to develop and deliver training that targets the day-to-day software and technology problems that researchers face. We deliver hands-on courses in a relaxed setting with knowledgeable, helpful trainers who are themselves researchers and who know how researchers work.

For more information about this course and others, see our course catalogue, or visit Learn.intersect.org.au

Header image credit: flickr.com/photos/be808/8494700566