Skip to main content

Install NVIDIA Driver

This guide covers installing the NVIDIA proprietary driver on a Lambda Tensorbook running Ubuntu 26.04.

System Specs

ComponentDetails
CPUIntel Core i7-10875H
GPUNVIDIA GeForce RTX 2080 Super Mobile / Max-Q (8GB)
iGPUIntel UHD Graphics (CometLake-H)
RAM64 GB
BIOSInsyde H2O (Clevo/Sager-based)

Step 1: Install the NVIDIA apt repo

Follow the standard Install CUDA guide to add the keyring and pin a driver version:

os=ubuntu2604
wget https://developer.download.nvidia.com/compute/cuda/repos/${os}/x86_64/cuda-keyring_1.1-1_all.deb
sudo apt install ./cuda-keyring_1.1-1_all.deb
sudo apt update

Step 2: Blacklist the Nouveau Driver

The Nouveau open-source driver must be disabled before rebooting into the NVIDIA driver:

sudo tee /etc/modprobe.d/blacklist-nouveau.conf > /dev/null << EOF
blacklist nouveau
options nouveau modeset=0
EOF

then update the initramfs image and reboot

sudo update-initramfs -u
sudo reboot

Step 3: Switch to TTY and Stop the Display Manager

At the login screen, press Ctrl + Alt + F1 to switch to a TTY session, then stop the display manager:

# Unity
# sudo systemctl stop lightdm

# Ubuntu GNOME
# sudo systemctl stop gdm3

# KDE
# sudo systemctl stop sddm

Step 4: Install the NVIDIA driver

Check the latest available NVIDIA driver, check the version first with apt policy and and pinning the specific version

sudo apt policy nvidia-driver-open
sudo apt install nvidia-driver-pinning-610
sudo apt install nvidia-driver-open

Step 5: Install the Xorg NVIDIA Video Driver

The CUDA repo package does not pull in xserver-xorg-video-nvidia automatically. Without it, X falls back to nouveau and fails to start:

sudo apt install xserver-xorg-video-nvidia
sudo reboot

Validation

nvidia-smi

You should see the GPU listed with driver and CUDA version, and a working desktop session.