Make sure you checkout the k8s-gpu branch. This is the one with the driver preinstalled.
William Buchwalter
1

oops, thx!!! I tried it and it does work. But I am still stuck at running pod. After I used kubectl create -f nvidia-smi.yaml to deploy the container, I found the pod was always at pending state and I got the message was “no nodes available to schedule pods”. I was wondering if it was caused by VNET issue like this doc mentions https://github.com/Azure/acs-engine/blob/master/docs/kubernetes.md#deployment

do you have any idea how to solve this?

Thanks a lot :)

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.