The only algorithm this pertains to is XGBoost, but in the general FAQ there is no mention of GPU support for XGBoost (it instead points users to H2O4GPU): http://docs.h2o.ai/h2o/latest-stable/h2o-docs/faq/general.html?highlight=gpu and there is not much info in the XGBoost page either.
We should specify the requirements (Nvidia GPUs and CUDA 8). We can suggest to the user that they can monitor the GPU utilization via the nvidia-smi command. https://developer.nvidia.com/nvidia-system-management-interface
It would also be nice to add a question to the AutoML FAQ: "Can we make use of GPUs with AutoML?" and the answer explains that if the requirements are met, then XGBoost will use GPUs (with a link to the other pages that explain in more detail).