Update "Using NVIDIA GPUs with minikube" (#19677)

Adding a note to delete existing minikube instance before starting one with the nvidia shim. This confused me for an good bit before I figured out why minikube couldn't see my GPUs.

Completely deleting the the instance and restarting from scratch seems to set everything up in a good state:
> 🌟  Enabled addons: nvidia-device-plugin, storage-provisioner, default-storageclass

site/content/en/docs/tutorials/nvidia.md

@@ -23,6 +23,7 @@ date: 2018-01-02
   ```shell
   sudo sysctl net.core.bpf_jit_harden
   ```
+
   - If it's not `0` run:
   ```shell
   echo "net.core.bpf_jit_harden=0" | sudo tee -a /etc/sysctl.conf
@@ -35,10 +36,20 @@ date: 2018-01-02
   ```shell
   sudo nvidia-ctk runtime configure --runtime=docker && sudo systemctl restart docker
   ```
+
+- Delete existing minikube (optional)
+
+  If you have an existing minikube instance, you may need to delete it if it was built before installing the nvidia runtime shim.
+  ```shell
+  minikube delete
+  ```
+  This will make sure minikube does any required setup or addon installs now that the nvidia runtime is available.
+  
 - Start minikube:
   ```shell
   minikube start --driver docker --container-runtime docker --gpus all
   ```
+
 {{% /tab %}}
 {{% tab none %}}
 ## Using the 'none' driver