GPU Computing with F# and Jack Pappas, Part 2

Please join me (Jack Pappas) for “Part 2” of my talk on HPC and GPU computing in F#. My first talk in November ran well over the allotted time so I wasn’t able to cover much of the GPU material many of you came to see.

This time, I’m going to focus solely on GPU computing and how it ties into F#. I’ll provide a brief overview of GPU hardware and programming terminology, teach you how to design massively parallel algorithms for GPU computing, and discuss how functional languages like F# are particularly suited for parallel computing. We’ll spend time walking through the “fundamental” algorithms of GPU computing and comparing implementations in C++ and F#. Finally -- if time allows -- I’ll give some tips for finding out which parts of your code could benefit the most from being ported to the GPU.