1. Overview

If you are already familiar with PyTorch on the IPU then Section 2, Quick start for experts provides a summary of the steps, focussing on the essential commands. If you need more detail, this can be found in Section 3, Quick start for beginners.

New to IPUs?

We suggest watching the Fundamentals of the IPU and Poplar video, which introduces the IPU architecture and programming model.

You can also read the IPU Programmer’s Guide and Switching from GPUs to IPUs for Machine Learning Models for more details on these topics.

You can expect the setup steps to take no more than 30 minutes.

This document is applicable if you are accessing IPUs on the systems listed in Section 1.1, IPU systems.

Note

There are three ways you can run PyTorch applications:

  1. Directly on the system using the setup described in this document.

  2. In a PyTorch Docker container that has already been setup. Refer to Using IPUs from Docker for more information.

1.1. IPU systems

Before following the instructions in this Quick Start guide, you must be able to log into a system with access to IPUs. Details are given in the getting started guide for your system: