Install drivers on a cloud server with GPU

For your information

This is an instructional guide with an example of installing drivers on a cloud server created from a pre-made image of Ubuntu 24.04 LTS 64-bit.

For stable operation of NVIDIA® GPUs on a cloud server with GPU, you need to install drivers.

If you created a cloud server from a pre-made image with GPU optimization, drivers are already installed on it, and no additional installation is required. Pre-made images with GPU optimization:

Ubuntu 24.04 LTS 64-bit GPU Driver 535;
Ubuntu 24.04 LTS 64-bit GPU Driver 535 Docker;
Ubuntu 24.04 LTS 64-bit GPU Driver 580;
Ubuntu 24.04 LTS 64-bit GPU Driver 580 Docker;
Ubuntu 22.04 LTS 64-bit GPU Driver 535;
Ubuntu 22.04 LTS 64-bit GPU Driver 535 Docker;
Ubuntu 22.04 LTS 64-bit GPU Driver 580;
Ubuntu 22.04 LTS 64-bit GPU Driver 580 Docker;
Data Science VM (Ubuntu 22.04 LTS 64-bit);
Data Analytics VM (Ubuntu 22.04 LTS 64-bit).

Install drivers

Connect to the cloud server with GPU.

Install the ubuntu-drivers-common package:

sudo apt install -y ubuntu-drivers-common alsa-utils

View the recommended driver version:

sudo ubuntu-drivers devices

A list of versions will appear in the output. The recommended version will be marked as recommended. Copy the recommended version.

Example for NVIDIA® Tesla T4 GPU with the recommended version nvidia-driver-550 ⁠:

== /sys/devices/pci0000:00/0000:00:06.0 ==
modalias : pci:v000010DEd00001EB8sv000010DEsd000012A2bc03sc02i00
vendor   : NVIDIA Corporation
model    : TU104GL [Tesla T4]
manual_install: True
driver   : nvidia-driver-450-server - distro non-free
driver   : nvidia-driver-535-server - distro non-free
driver   : nvidia-driver-470-server - distro non-free
driver   : nvidia-driver-470 - distro non-free
driver   : nvidia-driver-550 - third-party non-free recommended
driver   : nvidia-driver-418-server - distro non-free
driver   : xserver-xorg-video-nouveau - distro free builtin

Optional: verify that the selected driver version is higher than the minimum compatible version for the cloud server's GPU architecture:
```
sudo apt-cache search nvidia-driver-*
```
A list of compatible driver versions will appear in the output. You can find the GPU architecture in the Graphics Processing Units (GPUs) instructions, and the correspondence between driver versions and architectures in the CUDA Compatibility section of NVIDIA® documentation.
If the GPU architecture is Pascal (for example, for an NVIDIA® GTX 1080), add the NVIDIA® Personal Package Archive repository to the cloud server:
```
sudo add-apt-repository ppa:graphics-drivers/ppa -y
```
Install the kernel headers:
```
sudo apt update
for kernel in $(linux-version list); do apt install -y "linux-headers-<kernel-version>"; done
```
Specify <kernel-version> — the kernel version. You can view the list of kernel versions using the command apt-cache search linux-image.
Install the driver:
```
sudo apt install -y <driver_version>
```
Specify <driver_version> — the driver version you copied in step 3.

Example of installing the recommended version nvidia-driver-550 for NVIDIA® Tesla T4 GPU:
```
sudo apt install -y nvidia-driver-550
```

Verify that the driver is installed and working:

nvidia-smi

The NVIDIA-SMI and driver versions will appear in the response. Example output:

+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 550.54.15              Driver Version: 550.54.15      CUDA Version: 12.4     |
|-----------------------------------------+------------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id          Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |           Memory-Usage | GPU-Util  Compute M. |
|                                         |                        |               MIG M. |
|=========================================+========================+======================|
|   0  Tesla T4                       Off |   00000000:00:06.0 Off |                    0 |
| N/A   41C    P8             10W /   70W |       0MiB /  15360MiB |      0%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+

+-----------------------------------------------------------------------------------------+
| Processes:                                                                              |
|  GPU   GI   CI        PID   Type   Process name                              GPU Memory |
|        ID   ID                                                               Usage      |
|=========================================================================================|
|  No running processes found                                                             |
+-----------------------------------------------------------------------------------------+

Open the configuration file for the unattended-upgrades package, which handles security updates:
```
nano /etc/apt/apt.conf.d/50unattended-upgrades
```
Disable package updates for NVIDIA®. To do this, add the following block to the file:
```
Unattended-Upgrade::Package-Blacklist {
    "linux-";
    "nvidia-";
};
```
Exit the nano text editor saving the changes: press Ctrl+X, and then Y+Enter.
Optional: pin the kernel version to disable kernel updates. Updating the kernel version may cause errors in GPU driver operation.

Pin the kernel version

For your information

In pre-built images with pre-installed drivers, except for Data Analytics VM (Ubuntu 22.04 LTS 64-bit) and Data Science VM (Ubuntu 22.04 LTS 64-bit), the kernel version is already pinned.

During installation, drivers are compiled against the source code of the current kernel version headers. Changing the kernel version causes the GPU driver to stop working. In this case, the following error may appear in the output of the nvidia-smi command:

NVIDIA-SMI has failed because it couldn't communicate with the NVIDIA driver. Make sure that the latest NVIDIA driver is installed and running.

To disable kernel updates, pin the kernel version in the apt package manager settings. After pinning, you will be able to update the kernel version.

Open the CLI.

Create the pin-linux-kernel-nvidia-dkms file in the /etc/apt/preferences.d directory to pin the version of the linux-headers and linux-image packages:

cat <<EOF > /etc/apt/preferences.d/pin-linux-kernel-nvidia-dkms
Package: linux-image-*
Pin: version *
Pin-Priority: -1

Package: linux-headers-*
pin: version *
Pin-Priority: -1
EOF

Update the kernel version after pinning

Once the kernel version is pinned, you cannot update it. To download security updates, performance improvements, and new features, remove the kernel version pinning file and update the version.

Open the CLI.
Delete the file that you created to pin the kernel version:
```
rm /etc/apt/preferences.d/pin-linux-kernel-nvidia-dkms
```
Update the kernel version:
```
apt install linux-image-<kernel-version>
```
Specify <kernel-version> — the kernel version. You can view the list of kernel versions using the command apt-cache search linux-image.
Reboot the cloud server.
Install the kernel headers:
```
apt install linux-headers-$(uname -r)
```
After installing the kernel headers, the dkms utility will run, automatically rebuilding NVIDIA modules for the new kernel version.