Julia on HPC clusters

👋 Welcome to the notes about the Julia programming language on HPC clusters.

Objective

The goal of these notes is to offer good practices and tips and tricks about how to use and set up Julia on HPC clusters. The target audiences are therefore Julia users on clusters as well as HPC system administrators.

Acknowledgements

These notes don't come out of nowhere but stem from endless conversations with many people (see below for a certainly incomplete list) on the Julia discourse, over Slack and Discord, as well as in-person discussions at Julia HPC minisymposia, BoFs, and workshops.

They also build upon previous efforts to document best practices, most notably https://github.com/hlrs-tasc/julia-on-hpc-systems/, from which we took (among other things) the HPC systems with Julia support information.

People

Maintainers

Carsten Bauer - PC2 / NHR (creator)

Contributors

The following people have provided valuable contributions, either in the form of PRs or via public or private communication:

Alexander Bills (@abillscmu)
Johannes Blaschke (@jblaschke) - NERSC
Valentin Churavy (@vchuravy) - MIT
Steffen Fürst (@s-fuerst)
Mosè Giordano (@giordano) - UCL
William Godoy (@williamfgc) - ORNL
C. Brenhin Keller (@brenhinkeller)
Mirek Kratochvíl (@exaexa)
Pedro Ojeda (@pojeda)
Samuel Omlin (@omlins) - CSCS / ETHZ
Ludovic Räss (@luraess) - ETHZ
Michael Schlottke-Lakemper (@sloede) - HLRS / RWTH
Erik Schnetter (@eschnett) - PITP
Dinindu Senanayake (@DininduSenanayake)
Kjartan Thor Wikfeldt (@wikfeldt) - ENCCS

(You think we forgot your name here? Please make a PR!)