Note: This fork adds the missing X nvidia drivers. The Nvidia drivers are available in: /var/lib/nvidia/lib64/xorg/ on the host machine.

To make X work in a container, the following prerequisites are required:

• The container must be privileged

• The following devices must be made available to the container:

  • /dev/nvidia0
  • /dev/nvidiactl
  • /dev/nvidia-uvm
  • /dev/nvidia-uvm-tools

• The Nvidia driver must be made available to the container ie /var/lib/nvidia/lib64/ to /usr/local/nvidia/lib64/

• The Nvidia Xorg module path must be made known to the X server in xorg.conf:

Section "Files"
   ModulePath      "/usr/local/nvidia/lib64/xorg/modules"
   ModulePath      "/usr/lib/xorg/modules"
EndSection

• The X server must be started with -seat 1 to trick X into not trying to open a virtual terminal.

Note: This is not an official Google product.

This repository contains scripts to build Docker containers that can be used to download, compile and install GPU drivers on Container-Optimized OS images.

Example command:

gcloud compute instances create $USER-cos-gpu-test \
    --image-family cos-stable \
    --image-project cos-cloud \
    --accelerator=type=nvidia-tesla-k80 \
    --boot-disk-size=25GB \
    --maintenance-policy=TERMINATE \
    --metadata-from-file "cos-gpu-installer-env=scripts/gpu-installer-env,user-data=install-test-gpu.cfg,run-installer-script=scripts/run_installer.sh,run-cuda-test-script=scripts/run_cuda_test.sh"

The command above creates a GCE instance based on cos-stable image. Then it installs GPU driver on the instance by running a container 'cos-gpu-installer' which is implemented in this repository.

The GPU driver version and container image version are specified in scripts/gpu-installer-env. You can edit the file if you want to install GPU driver version or use container image other than the default.