Gray-Scott with Rust : an introduction to Rust for numerical computing

I am a software engineer in a French physics lab. This year, with a bunch of colleagues across the country, we've been running a 2 weeks school on various technologies for numerical computing (CUDA, SYCL, Python...), using the optimization of a finite-difference Euler solver of the Gray-Scott model as a leitmotiv since it's a computational problem that is just tricky enough to break most optimizing compilers, yet simple enough to be explained quickly.

As part of this, one of my main contributions there was to run some Rust practical work : 1 day of introduction to the language + CPU computing, and 1/2 day of GPU computing using Vulkan.

The practical's material is written in English as a preparation for the next edition of the school (which will be run in English with an EU-wide audience), so I figured out it might be of interest to other people here.