Enable Intel GPUs¶

This guide describes how to configure Data Science Stack (DSS) to utilise the Intel GPUs on your machine.

You can do so done by enabling the Intel device plugin on your MicroK8s cluster.

Prerequisites¶

Ubuntu 22.04.
DSS is installed and initialised.
The kubectl snap package is installed.
Your machine includes an Intel GPU.

Note

After installing kubectl, configure MicroK8s to run kubectl commands as follows:

mkdir -p ~/.kube
microk8s config > ~/.kube/config

Verify the Intel GPU drivers¶

To confirm that your machine has the Intel GPU drivers set up, first install the intel-gpu-tools package:

sudo apt install intel-gpu-tools

Now list the Intel GPU devices on your machine as follows:

intel_gpu_top -L

If the drivers are correctly installed, you should see information about your GPU device such as the following:

card0                    8086:56a0
pci:vendor=8086,device=56A0,card=0
└─renderD128

Note

For Intel discrete GPUs on Ubuntu versions older than 24.04, you may need to perform additional steps such as installing a HWE kernel.

Enable the Intel GPU plugin¶

To ensure DSS can utilise Intel GPUs, you have to enable the Intel GPU plugin in your MicroK8s cluster.

Use kubectl kustomize to build the plugin YAML configuration files:

VERSION=v0.30.0
kubectl kustomize https://github.com/intel/intel-device-plugins-for-kubernetes/deployments/nfd?ref=${VERSION} > node_feature_discovery.yaml
kubectl kustomize https://github.com/intel/intel-device-plugins-for-kubernetes/deployments/nfd/overlays/node-feature-rules?ref=${VERSION} > node_feature_rules.yaml
kubectl kustomize https://github.com/intel/intel-device-plugins-for-kubernetes/deployments/gpu_plugin/overlays/nfd_labeled_nodes?ref=${VERSION} > gpu_plugin.yaml

To allow multiple containers to utilise the same GPU, run:

sed -i 's/enable-monitoring/enable-monitoring\n        - -shared-dev-num=10/' gpu_plugin.yaml

Apply the built YAML files to your MicroK8s cluster:

kubectl apply -f node_feature_discovery.yaml
kubectl apply -f node_feature_rules.yaml
kubectl apply -f gpu_plugin.yaml

The MicroK8s cluster is now configured to recognise and utilise your Intel GPU.

Note

After the YAML configuration files have been applied, they can be safely deleted.

Verify the Intel GPU plugin¶

To verify the Intel GPU plugin is installed and the MicroK8s cluster recognises your GPU, run:

kubectl get nodes --show-labels | grep intel

You should see an output with the cluster name such as the following:

kubectl get nodes --show-labels | grep intel
fluent-greenshank   Ready    <none>   18s   v1.30.3   beta.kubernetes.io/arch=amd64,beta.kubernetes.io/os=linux,intel.feature.node.kubernetes.io/gpu=true

Verify DSS detects the GPU¶

Verify DSS has detected the GPU by checking the DSS status. To do so, run the following command using the DSS CLI:

dss status

You should expect an output like this:

Output:
MLflow deployment: Ready
MLflow URL: http://10.152.183.68:5000
NVIDIA GPU acceleration: Disabled
Intel GPU acceleration: Enabled

Enable Intel GPUs¶

Prerequisites¶

Verify the Intel GPU drivers¶

Enable the Intel GPU plugin¶

Verify the Intel GPU plugin¶

Verify DSS detects the GPU¶

See also¶