Introduction: serial vs parallel

• Oskar Klimas (ACK Cyfronet AGH)

Room: https://cyfronet.zoom.us/j/94176138049 Basic information about tasks, threads, MPI, OpenMP and more.

Embarrassingly parallel (so parallel it's embarrassing!)

• Oskar Klimas (ACK Cyfronet AGH)

Room: https://cyfronet.zoom.us/j/94176138049 The examples of algorithms that are trivial to make parallel.

Intranode standards (1)

• Oskar Klimas (ACK Cyfronet AGH)

Room: https://cyfronet.zoom.us/j/94176138049 Meeting MPI, OpenMP and srun head on. Learn when and how to use them.

Intranode standards (2)

• Oskar Klimas (ACK Cyfronet AGH)

Room: https://cyfronet.zoom.us/j/94176138049 Meeting MPI, OpenMP and srun head on. Learn when and how to use them.

Going multinode

• Oskar Klimas (ACK Cyfronet AGH)

Room: https://cyfronet.zoom.us/j/94176138049 Spreading a calculation over more than 1 node

Accelerating with a GPU card (1)

• Oskar Klimas (ACK Cyfronet AGH)

Room: https://cyfronet.zoom.us/j/94176138049 Running calculations using an accelerator, in this case a GPU card.

Accelerating with a GPU card (2)

• Oskar Klimas (ACK Cyfronet AGH)

Room: https://cyfronet.zoom.us/j/94176138049 Running calculations using an accelerator, in this case a GPU card.

When HPC goes HTC (array jobs)

• Oskar Klimas (ACK Cyfronet AGH)

Room: https://cyfronet.zoom.us/j/94176138049 Running multiple small jobs of the same template.

Chaining calculations with job dependencies

• Oskar Klimas (ACK Cyfronet AGH)

Room: https://cyfronet.zoom.us/j/94176138049 Using dependency option in Slurm to run calculations based on the status of another.