“Over the last few years we have been running a GPU kernel performance competition in MA4261. The students compete to write the fastest GPU kernel which computes a Mandelbrot fractal image. This simple compute-bound kernel is used to learn about fused multiply-adds and algorithmic optimizations. It also produces interesting pictures. This year there will be two winners, one for the students that used AI/LLMs to aid their development and one for those that didn't. Both winners not only get a framed print of their fractal, they will also be given the opportunity to attend NVIDIA GTC.

This year's winners are LCDR Matthew Mcintyre, ENS David Jepsen, and LT Alexander Wayne.