Skip to content

Custom cloud init for Intel OOT driver installation#288

Open
Ram-srini wants to merge 4 commits intomainfrom
custom_cloud_init
Open

Custom cloud init for Intel OOT driver installation#288
Ram-srini wants to merge 4 commits intomainfrom
custom_cloud_init

Conversation

@Ram-srini
Copy link
Contributor

PR Description

User guide to install the Intel OOT GPU drivers using custom cloud-init.

Changes

List the changes you have made.

Additional Information

Include any additional information, such as how to test your changes.

Checklist

  • Tests passed
  • Documentation updated

Signed-off-by: Srinivasamurthy, Ramakrishna <ramakrishna.srinivasamurthy@intel.com>
@Ram-srini Ram-srini added the EMF-2025.2 EMF 2025.2/3.2 release label Dec 9, 2025
@Ram-srini Ram-srini enabled auto-merge (squash) December 9, 2025 11:00

Here are the supported GPU drivers for Intel platforms:

#. Intel iGPU - ADL, RPL platforms.
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

these are code names we should change them like - Intel® Core™ Ultra Series 3 processors (formerly code-named Panther Lake)

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Done


#. Intel iGPU - ADL, RPL platforms.

#. Intel Battle Mage 580 GPU - Xeon platforms (Dell XR12 Icelake).
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

these are code names we should change them like - Intel® Core™ Ultra Series 3 processors (formerly code-named Panther Lake)

Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Battlemage is one word, the model is also Intel Arc B580.


Here are the supported GPU drivers for Intel platforms:

#. Intel iGPU - ADL, RPL platforms.
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

why do we need OOT for iGPU ON Ubuntu 24.04 does it not work out of box ?

Copy link
Contributor Author

@Ram-srini Ram-srini Mar 18, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Correct but Panther lake if iGPU gpu OOT drivers are not up streamed then we might need it. We don't need it for RPL

merge_how: 'dict(recurse_array,no_replace)+list(append)'
runcmd:
- |
wget https://af01p-png.devtools.intel.com/artifactory/hspe-edge-png-local/ubuntu/daily/panther-lake/20251106-1458/installer.sh --no-check-certificate --no-proxy -O /tmp/installer.sh
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This is inside Intel URL which customer can not access

Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

To add to what @krishnajs mentions, this is even more problematic as this script is written explicitly for internaly use and appears to statically define internal Intel proxies and discloses proxy names to external users if released as-is.


Replace the placeholders in the command below:

- ``<API_USER>``: Your API username.
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Should it not be export in the start?

Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We should link to existing orch-cli user guides rather than duplicating documents for logging in, this section should only refer to the specific verbs/commands being documented here.


Replace the placeholders below as per your orchestrator:

- ``<PROJECT_NAME>``: Project name.
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Should it not be export in the start?

Create a custom cloud-init YAML file called ``gpu_driver_installation.yaml`` with the content
below.

Replace ``<PLATFORM_TYPE>`` with the proper platform type:
Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The Platform Type should be defined somewhere as most users will not know our codename abbreviations, ideally this is documented within the script where this is released. User guide should suggest downloading the script to find correct parameters.

@hwindlas hwindlas removed the EMF-2025.2 EMF 2025.2/3.2 release label Dec 12, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants