In this part, we introduce the Diffusion 2D example we will use throughout the workshop to exemplify various HPC concepts in Julia, namely:
The script diffusion_d2.jl
provides the starting point; a vectorised 2D linear diffusion solver computing diffusive fluxes and their divergence in a compute_flux!
and diffusion_step!
function, respectively.
The follow-up script, diffusion_d2_loop.jl
, implements a serial loop version of the previous script that we will use as a starting point for all our further experiments.
Your very first task is to get familiar with the script structure and output generated. Run the diffusion_2d-jl
script, verifying that plotting works and assess the reported effective memory throughput T_eff
(in the REPL).
Repeat the same for the diffusion_2d_loop.jl
script.