--- orphan: true --- # Device Labeling (Node Feature Discovery) **Status: Supported** tt-operator deploys [Node Feature Discovery](https://kubernetes-sigs.github.io/node-feature-discovery/) (NFD) to detect Tenstorrent hardware and label the nodes that have it. Those labels are what make the Driver Manager, telemetry, and your own workloads schedule only where a device is present. ## What it does NFD inspects each node's PCI devices and applies the label: ``` feature.node.kubernetes.io/pci-1200_1e52.present=true ``` to every node that has a Tenstorrent device (PCI vendor `1e52`). The chart restricts NFD to the PCI source only, so it does not add the full set of CPU, kernel, and system labels. This keeps node labels focused and the footprint small. ## Verify ```bash kubectl get nodes -l feature.node.kubernetes.io/pci-1200_1e52.present=true ``` Every node with a Tenstorrent device should be listed. If a node you expect is missing, confirm the device is visible on the host with `lspci` and that the NFD worker pod is running on that node. ## Configuration NFD is enabled by default. To turn it off, for example if you already run NFD cluster-wide, set: ```bash --set node-feature-discovery.enabled=false ``` The feature and label sources are restricted to `pci` via `node-feature-discovery.worker.config.core.featureSources` and `labelSources`. See the [Configuration reference](../configuration.md) for the full set of NFD values.