The deep learning AMIs for AWS make it easy to setup deep learning models (in particular, for using GPU instances), see: https://aws.amazon.com/de/machine-learning/amis/.
For Google Compute Engine many blog posts/tutorials show how to set up a instance for deep learning by installing all required libaries manually (Google search). However, this provides not the same convenience that the "one-click-solution" of an AMI offers.
I am curious if there is a functionality like Amazon's AMIs for Google Compute Engine (in particular, some which makes setting up a deep learning instances with GPU instances a breeze)?
GCE now has the deep learning images.