Installing NVIDIA Drivers and CUDA on Linux Mint 21.3

3 min readJun 21, 2024

If you’re a Linux Mint 21.3 user with an NVIDIA GPU, you might want to install the NVIDIA drivers and CUDA toolkit to leverage the full potential of your graphics card. This guide will take you through the step-by-step process of installing NVIDIA drivers and CUDA on your Linux Mint 21.3 system.

Prerequisites

Before you begin, ensure you have:

Linux Mint 21.3 installed on your machine
An NVIDIA GPU (this guide uses GeForce RTX 3060 as an example)
Internet connection for downloading drivers and software packages

Step 1: Uninstall Existing NVIDIA Drivers

First, it’s a good idea to remove any existing NVIDIA drivers to avoid conflicts. Open a terminal and run the following commands:

sudo apt-get purge 'nvidia*'
sudo apt-get autoremove
sudo apt-get autoclean

Step 2: Add the Graphics Drivers PPA

To get the latest NVIDIA drivers, add the official graphics drivers PPA:

sudo add-apt-repository ppa:graphics-drivers/ppa
sudo apt-get update

Step 3: Install NVIDIA Drivers

Next, install the recommended NVIDIA drivers for your GPU. You can find the recommended driver version using the ubuntu-drivers tool:

ubuntu-drivers devices

Look for the recommended driver version (e.g., nvidia-driver-535). Install it by running:

sudo apt-get install nvidia-driver-535
sudo reboot

After rebooting, verify the installation with:

nvidia-smi

This command should display your GPU information and the installed driver version.

Step 4: Install CUDA Toolkit

With the NVIDIA drivers installed, you can now install the CUDA toolkit. Visit the NVIDIA CUDA Toolkit download page and download the appropriate version for your Linux Mint 21.3.

Here, we’ll use the terminal for installation. First, download the installer:

wget https://developer.download.nvidia.com/compute/cuda/12.5.0/local_installers/cuda_12.5.0_555.42.02_linux.run

Make the installer executable and run it:

chmod +x cuda_12.5.0_555.42.02_linux.run
sudo ./cuda_12.5.0_555.42.02_linux.run

Follow the on-screen instructions, ensuring you install both the driver and the toolkit.

Step 5: Set Up Environment Variables

To use CUDA, you need to set up the environment variables. Add the following lines to your ~/.bashrc file:

export PATH=/usr/local/cuda-11.6/bin${PATH:+:${PATH}}
export LD_LIBRARY_PATH=/usr/local/cuda-11.6/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}

Source the ~/.bashrc file to apply the changes:

source ~/.bashrc

Step 6: Verify CUDA Installation

Check the CUDA version to ensure it’s installed correctly:

nvcc -V

You should see output similar to:

nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2022 NVIDIA Corporation
Built on Wed_Feb_16_22:00:56_PST_2022
Cuda compilation tools, release 11.6, V11.6.124
Build cuda_11.6.r11.6/compiler.30978841_0

Step 7: Compile and Run a Sample CUDA Program

To ensure everything is working properly, compile and run a sample CUDA program. Create a file called sample.cu with the following content:

#include <iostream>

__global__ void helloFromGPU() {
    printf("Hello World from GPU!\n");
}

int main() {
    helloFromGPU<<<1, 10>>>();
    cudaDeviceSynchronize();
    return 0;
}

Compile the program using nvcc:

nvcc -o sample sample.cu

Run the compiled program:

./sample