What is a GPU? How can it be used for scientific computing? What GPU resources does Princeton provide? This workshop will answer these questions and have participants run code written in Python, R, Julia or MATLAB on the GPU nodes of the Adroit Research Computing cluster. Participants will also gain hands-on experience with getting started with CUDA, a low-level GPU programming model.

Learning objectives: Attendees will learn about GPU hardware and come away with the ability to write and execute simple, compiled programs that use GPUs.

Knowledge prerequisites: Basic Linux and some exposure to a compiled programming language (e.g., C/C++, Fortran, Java).

Hardware/software prerequisites: This workshop will use reserved nodes on the Adroit cluster. If you need an account then please request one at least an hour before the workshop.

Workshop format: Presentation (40%) and hands-on (60%)