MPAGS: High Performance Computing in Julia In this lecture, we talk about the concept of GPU programming, including the differences between GPU and CPU hardware. We discuss some models of how to compute on a GPU, with particular focus on CUDA and the CUDA.jl library. We cover some examples of the high-level array based programming mechanism provided by CUDA.jl to avoid the need to write one's own kernels. This is module designed for the Midlands Physics Alliance Graduate School (MPAGS). More information can be found on the website.