Having multiple cores available only speeds your code when you know how to use them. If you want to discover how to enable CPU parallelism in your code, and have a familiarity with C, C++ or Fortran, then this hands-on workshop is for you.

You can join in-person OR virtually. This is a free workshop. Please register only if you plan to attend.

Please join at 8:30am (AWST) to ensure you can access the system. The class starts promptly at 9:00am.

What will I learn in this 3-hour, hands-on workshop?
You will learn about OpenMP: the reference in the High-Performance Computing world for shared memory programming over the last 25+ years. This workshop will address how to:

• spawn multiple threads to enable CPU parallelism.
• coordinate threads to have a certain piece of code executed by certain threads.
• handle data between threads, depending on whether you want them to have their own copy or share it.
• parallelise the execution of loops across threads.
• make sure that different threads end with similar amounts of work.
• synchronise threads to ensure that the program correctness is preserved.

You must be familiar with at least one of the following languages:

C/C++ (Refresher videos) or
Fortran

Motivation: reason for using OpenMP with CPUs
Under the hood: How OpenMP works (directive-based, fork-join pattern)
Starting parallelism (parallel construct)
Data-sharing attribute clauses (default(none), shared, private and firstprivate clauses)
Worksharing constructs (single and master constructs)
Synchronisation constructs (barrier and critical constructs)
Loops time (for / do constructs)
Reduction (reduction clause / atomic construct)
Load-balancing techniques (Schedule clause)