From 4032470d34d17ba46f33db99668d0d7a431e8fcc Mon Sep 17 00:00:00 2001
From: Carlos Eduardo Arango Gutierrez <eduardoa@nvidia.com>
Date: Sat, 27 Jan 2024 17:39:36 +0100
Subject: [PATCH] Document NFD for GPU Labeling

Signed-off-by: Carlos Eduardo Arango Gutierrez <eduardoa@nvidia.com>
---
 .../docs/tasks/manage-gpus/scheduling-gpus.md | 42 +++++++++++++++----
 1 file changed, 35 insertions(+), 7 deletions(-)

diff --git a/content/en/docs/tasks/manage-gpus/scheduling-gpus.md b/content/en/docs/tasks/manage-gpus/scheduling-gpus.md
index d02a20577e7f4..4a967860fb4fa 100644
--- a/content/en/docs/tasks/manage-gpus/scheduling-gpus.md
+++ b/content/en/docs/tasks/manage-gpus/scheduling-gpus.md
@@ -64,7 +64,7 @@ spec:
           gpu-vendor.example/example-gpu: 1 # requesting 1 GPU
 ```
 
-## Clusters containing different types of GPUs
+## Manage clusters with different types of GPUs
 
 If different nodes in your cluster have different types of GPUs, then you
 can use [Node Labels and Node Selectors](/docs/tasks/configure-pod-container/assign-pods-nodes/)
@@ -81,12 +81,40 @@ kubectl label nodes node2 accelerator=other-gpu-k915
 That label key `accelerator` is just an example; you can use
 a different label key if you prefer.
 
-## Automatic node labelling {#node-labeller}
+### Automatically labeling nodes with Node Feature Discovery {#node-feature-discovery}
 
-If you're using AMD GPU devices, you can deploy
+As an administrator, you can automatically discover and label all your GPU enabled nodes
+by deploying the K8S-Sig project [Node Feature Discovery](https://github.com/kubernetes-sigs/node-feature-discovery) (NFD).
+NFD detects the hardware features that are available on each node in a Kubernetes cluster and advertises those features.
+Typically, NFD adds node labels to advertise the features, but NFD can also add extended resources, annotations, and node taints.
+NFD is compatible with any recent version of Kubernetes (v1.21+).
+
+Administrators can leverage NFD to also taint nodes with specific features, so that only pods that request those features can be scheduled on those nodes.
+After a cluster is labeled with the GPU feature, you can schedule pods on GPU nodes by adding the following to your pod spec:
+
+```yaml
+apiVersion: v1
+kind: Pod
+metadata:
+  name: example-vector-add
+spec:
+  restartPolicy: Never
+  containers:
+    - name: example-vector-add
+      image: "registry.example/example-vector-add:v42"
+      resources:
+        limits:
+          gpu-vendor.example/example-gpu: 1 # requesting 1 GPU
+      nodeSelector:
+        gpu-vendor.example/example-gpu: "true"
+```
+
+NFD exposes an API which allows vendors to leverage the automatic labeling functionality. 
+NVIDIA has implemented this API in the [GPU feature discovery](https://github.com/NVIDIA/gpu-feature-discovery/blob/main/README.md).
+
+### Using custom labellers
+
+For AMD GPUs, you can use the
 [Node Labeller](https://github.com/RadeonOpenCompute/k8s-device-plugin/tree/master/cmd/k8s-node-labeller).
 Node Labeller is a {{< glossary_tooltip text="controller" term_id="controller" >}} that automatically
-labels your nodes with GPU device properties.
-
-Similar functionality for NVIDIA is provided by
-[GPU feature discovery](https://github.com/NVIDIA/gpu-feature-discovery/blob/main/README.md).
+labels nodes in a Kubernetes cluster with AMD GPU device properties.