Scribd
Upload
144 views

NVIDIA GPU CUDA Toolkit & Driver Installation in Linux OS

1. Install NVIDIA drivers and CUDA toolkit on a Linux system with NVIDIA H100 GPUs. This involves verifying system requirements, installing OS prerequisites like blacklisting Nouveau drivers, installing dependent packages, downloading the NVIDIA driver package and CUDA toolkit, and modifying environment variables.

Uploaded by

MWANAHAWA BAKARI
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
144 views

NVIDIA GPU CUDA Toolkit & Driver Installation in Linux OS

1. Install NVIDIA drivers and CUDA toolkit on a Linux system with NVIDIA H100 GPUs. This involves verifying system requirements, installing OS prerequisites like blacklisting Nouveau drivers, installing dependent packages, downloading the NVIDIA driver package and CUDA toolkit, and modifying environment variables.

Uploaded by

MWANAHAWA BAKARI
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

NVIDIA GPU CUDA Toolkit & Driver Installation in Linux OS

1. Verify System Requirements:


• Run the command : lspci grep | -i "nvidia"
• Verify you have Nvidia H100 GPUs installed.
• Ensure that you have Red Hat Enterprise Linux (RHEL) installed on your system.
2. OS Pre-requisites:
• Edit /etc/default/grub. Append the following to “GRUB_CMDLINE_LINUX”
o rd.driver.blacklist=nouveau nouveau.modeset=0
• Generate a new grub configuration to include the above changes.
o grub2-mkconfig -o /boot/efi/EFI/redhat/grub.cfg
• Next create a new “initramfs” file and taking backup of existing.
o # mv /boot/initramfs-$(uname -r).img /boot/initramfs-$(uname -r).img.bak
o # dracut -v /boot/initramfs-$(uname -r).img $(uname -r)
• Reboot the system.
3. Install the dependent packages required for NVDIA drivers
• yum install elfutils-libelf-devel numactl-devel numactl kernel-devel
• yum install https://dl.fedoraproject.org/pub/epel/epel-release-latest-8.noarch.rpm
4. Install Nvidia Drivers and CUDA toolkit.
• Note: CUDA toolkit includes Drivers. You could skip installing drivers if you are
installing CUDA toolkit.
• Drivers :
o Download drivers
from: https://www.nvidia.com/Download/driverResults.aspx/202359/en-
us/
o Once you accept the download please follow the steps listed below

i) `sudo rpm -i nvidia-driver-local-repo-rhel8-525.105.17-1.0-


1.x86_64.rpm`

ii) `sudo dnf clean all`

iii) `sudo dnf -y module install nvidia-driver:latest-dkms`

Cuda 12.0.1:
wget https://developer.download.nvidia.com/compute/cuda/12.0.1/local_in
stallers/cuda-repo-rhel8-12-0-local-12.0.1_525.85.12-1.x86_64.rpm

rpm -i cuda-repo-rhel8-12-0-local-12.0.1_525.85.12-1.x86_64.

rpmsudo dnf clean all

sudo dnf -y module install nvidia-driver:latest-dkmssudo dnf -y install cuda

Ubuntu 22.04

wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2
204/x86_64/cuda-ubuntu2204.pin

Sudo mv cuda-ubuntu2204.pin /etc/apt/preferences.d/cuda-repository-


pin-600

wget https://developer.download.nvidia.com/compute/cuda/12.0.1/local_in
stallers/cuda-repo-ubuntu2204-12-0-local_12.0.1-525.85.12-
1_amd64.deb

Sudo dpkg -i cuda-repo-ubuntu2204-12-0-local_12.0.1-525.85.12-


1_amd64.deb

sudo cp /var/cuda-repo-ubuntu2204-12-0-local/cuda-*-keyring.gpg
/usr/share/keyrings/

sudo apt-get update

sudo apt-get -y install cuda

5. Include The PATH and LD_LIBRARY_PATH variable needs to be included.

$ export PATH=/usr/local/cuda-11.2/bin${PATH:+:${PATH}}

$ export LD_LIBRARY_PATH=/usr/local/cuda-
11.7/lib64{LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}

You might also like