Prerequisites
In order to compile and run the examples this week you will use:
- the shell command line (bash)
- the GNU C compiler (gcc)
- the Intel OneAPI (icc)
- the Nvidia HPC SDK (nvc)
Exercises involve compiling and running C code. Familiarity with this language will be helpful but is not required.
| Setup | Download files required for the lesson | |
| 00:00 | 1. Introduction |
How shared memory parallel programs work?
What is OpenMP? How to write and compile parallel programs in C? |
| 00:20 | 2. A Parallel Hello World Program |
How do you compile and run an OpenMP program?
What are OpenMP pragmas? How to identify threads? |
| 00:50 | 3. OpenMP Work Sharing Constructs | During parallel code execution, how is the work distributed between threads? |
| 01:20 | 4. Parallel Operations with Arrays |
How do I parallelize a loop?
How do I ensure that the parallel code gives correct results? |
| 02:00 | 5. Race Conditions with OpenMP |
How can we calculate integrals in parallel?
How to compile a program using math functions with gcc? |
| 02:30 | 6. OpenMP Tasks | How to recurse in parallel |
| 03:00 | 7. Calculating Many-Body Potentials | How to speed up calculation leveraging both OpenMP and SIMD instructions |
| 03:25 | 8. Programming GPUs with OpenMP | How to program GPU with OpenMP? |
| 04:05 | 9. Introduction to GPU Programming with OpenACC | How to program GPU with OpenACC? |
| 04:45 | Finish |
The actual schedule may vary slightly depending on the topics and exercises chosen by the instructor.