Readme.md

About this tutorial

In this tutorial, you will learn how to use idefix on various architectures. Here, we will do everything through a Jupyter notebook opened on a GPU-hosting machine, but you can also do most of the tutorial on the CPU of your laptop.

This tutorial was prepared by Geoffroy Lesur (geoffroy.lesur@univ-grenoble-alpes.fr) and Sebastien Valat (sebastien.valat@univ-grenoble-alpes.fr).

Deployment

You can play around with this tutorial on a distant machine or on your laptop (for the asnum school, we recommend deploying on either the dgx, h200x4 or procamd machines). In any cases, you should clone this tutorial on the target machine, so that you can directly use these source files and test what you are doing.

We first connect to the target machine (here dgx, but change it as needed to balance the charge between users) and make a tunnel to a unix socket (which will be used by our jupyter notebook)

ssh dgx.cbp.ens-lyon.fr -L 8888:/home/<your_login>/mysock.sock

In the directory of your choice, you then clone the tutorial to the directory idefix-tutorial:

git clone --branch asnum2026 https://github.com/idefix-code/tutorial.git ./idefix-tutorial
export IDEFIX_TUTORIAL_DIR="$PWD/idefix-tutorial"
cd "$IDEFIX_TUTORIAL_DIR"

We then create a python environement, install what we need, and launch a jupyter lab session

cd "$IDEFIX_TUTORIAL_DIR"
python3 -m venv ./idefix.env
source idefix.env/bin/activate
pip install -r ./python_requirements.txt
jupyter lab --sock=~/mysock.sock

you're now good to go. On your laptop, using your favourite browser, open the link

http://localhost:8888/lab?token=<your_token>

Which should open your jupyter lab session running on the distant machine. From there, open start.ipynb.