Summary and Schedule
This lesson runs three kernels under Slurm. AXPY and the dot product have a hybrid MPI and OpenMP program. Matrix multiplication is sequential, and you parallelise the outer loop yourself.
The EuroCC2 project page is https://www.eurohpc-ju.europa.eu/research-innovation/our-projects/eurocc-2_en.
Credits
This lesson is adapted from Introducción a la
Supercomputación by José Luis Bosque, Borja Pérez, Ramón Beivide,
and Carmen Martínez, Universidad de Cantabria, by the EuroCC NL team at
SURF. License: CC-BY 4.0. The names are also in AUTHORS and
CITATION.cff.
| Setup Instructions | Download files required for the lesson | |
| Duration: 00h 00m | 1. Nodes, tasks, and threads |
Which Slurm option is a node, a task, a core, and a thread? How many hostname lines does one node print when four tasks run there? |
| Duration: 00h 40m | 2. AXPY |
Why does AXPY speed up and then flatten? What does the hybrid program print for y[0] and for the local length?
|
| Duration: 01h 20m | 3. Dot product |
Where is the reduction in the hybrid dot product? Why can the last digits differ when the printed value matches? |
| Duration: 02h 00m | 4. Matrix multiplication |
What has to stay the same when you parallelise the outer loop? Which compile flag makes the pragma real? |
| Duration: 02h 40m | Finish |
The actual schedule may vary slightly depending on the topics and exercises chosen by the instructor.
You need a terminal on Snellius or the course cluster. Copy
episodes/files/hands-on/ to the cluster and submit from the
exercise directory. The scripts compile the .c file in the
current directory.
The job scripts load foss/2025a after the 2025 stack.
That module provides gcc and mpicc.