docs: write about GPU support

README.md

@@ -38,6 +38,12 @@ A remotely accessible Kubernetes home lab with OIDC authentication. Build a mode
 - **[Longhorn](https://longhorn.io/)**: Distributed block storage
 - **[MinIO](https://min.io/)**: S3-compatible object storage
 
+### GPU Support (Optional)
+
+- **[NVIDIA Device Plugin](https://github.com/NVIDIA/k8s-device-plugin)**: GPU resource management for Kubernetes
+    - Exposes NVIDIA GPUs to Kubernetes as schedulable resources
+    - Required for GPU-accelerated workloads in JupyterHub and other applications
+
 ### Data & Analytics (Optional)
 
 - **[JupyterHub](https://jupyter.org/hub)**: Interactive computing with collaborative notebooks
@@ -178,6 +184,7 @@ Multi-user platform for interactive computing:
 - **Keycloak Authentication**: OAuth2 integration with SSO
 - **Persistent Storage**: User notebooks stored in Longhorn volumes
 - **Collaborative**: Shared computing environment for teams
+- **GPU Support**: CUDA-enabled notebooks with nvidia-device-plugin integration
 
 [📖 See JupyterHub Documentation](./jupyterhub/README.md)
 

nvidia-device-plugin/README.md

@@ -193,30 +193,6 @@ spec:
         nvidia.com/gpu: 1
 ```
 
-### Using GPUs in JupyterHub
-
-Configure JupyterHub to allow GPU access for notebook servers:
-
-```yaml
-# jupyterhub values.yaml
-singleuser:
-  runtimeClassName: nvidia
-  extraResource:
-    limits:
-      nvidia.com/gpu: "1"
-```
-
-After deploying JupyterHub with this configuration, users can access GPUs in their notebooks:
-
-```python
-import torch
-
-# Check GPU availability
-print(torch.cuda.is_available())  # True
-print(torch.cuda.device_count())   # 1
-print(torch.cuda.get_device_name(0))  # NVIDIA GeForce RTX 4070 Ti
-```
-
 ### Multiple GPUs
 
 To request multiple GPUs: