LECTURE: Parallel Programming Models

Overview

Teaching: 30 min
Exercises: 0 min

Questions

• How is software actually written to run in parallel?

• Does this differ in the shared-memory vs distributed-memory context?

• What are the roles of processes and threads?

• How is software written to use accelerators (GPUs)?

• Why is this useful to know for running (if not writing) software on HPC?

Objectives

• Understand concretely how software can execute in parallel through MPI and OpenMP using processes and threads

• Understand the kinds of modifications needed to enable applications to use accelerators (GPUs)

• Appreciate some of the complexities that parallel programmers need to resolve and which can impact performance

• Gain an ability to assess whether a given software application of interest is able to run on a given machine (HPC or smaller) based on the programming models it supports

Slides

This part of the course will be delivered using a slide presentation. You can find the slides for this here.

Key Points

• To be determined