In this article, we will guide you through the process of installing CUDA Toolkit version 11.6 on Ubuntu 18.04. The latest CUDA Toolkit version available in the Ubuntu repositories is 9.1, but for various reasons, you might need to install a newer version. Let’s get started!
What is CUDA?
CUDA, or Compute Unified Device Architecture, is a parallel computing platform and application programming interface (API) model created by NVIDIA. It allows developers to use a CUDA-enabled graphics processing unit (GPU) for general purpose processing – an approach termed GPGPU (General-Purpose computing on Graphics Processing Units).
Why Install CUDA Toolkit Version 11.6?
The CUDA Toolkit includes GPU-accelerated libraries, a compiler, development tools, and the CUDA runtime. It is designed to work with programming languages like C, C++, and Fortran. The latest version, 11.6, provides several new features and improvements over the older version available in the Ubuntu repositories.
Prerequisites
Before we begin, please ensure that you have a CUDA-capable GPU and a compatible version of Linux. You can check whether your GPU is CUDA-capable here.
Installation Steps
Step 1: Add the CUDA Repository
First, we will add the CUDA repository to our system. This will allow us to install the latest version of the CUDA Toolkit.
wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/x86_64/cuda-ubuntu1804.pin
sudo mv cuda-ubuntu1804.pin /etc/apt/preferences.d/cuda-repository-pin-600In these commands, wget is used to download the repository pin file, and mv is used to move the downloaded file to the /etc/apt/preferences.d/ directory.
Step 2: Download the CUDA Repository Package
Next, we will download the CUDA repository package for version 11.6.
wget https://developer.download.nvidia.com/compute/cuda/11.6.0/local_installers/cuda-repo-ubuntu1804-11-6-local_11.6.0-510.39.01-1_amd64.debStep 3: Install the CUDA Repository Package
Now, we will install the downloaded CUDA repository package using the dpkg command.
sudo dpkg -i cuda-repo-ubuntu1804-11-6-local_11.6.0-510.39.01-1_amd64.debStep 4: Add the CUDA Repository Key
We need to add the CUDA repository key to our system. This ensures that the packages we install are authentic and have not been tampered with.
sudo apt-key add /var/cuda-repo-ubuntu1804-11-6-local/7fa2af80.pubStep 5: Update the Package List
Before we can install the CUDA Toolkit, we need to update the package list on our system.
sudo apt-get updateStep 6: Install the CUDA Toolkit
Finally, we can install the CUDA Toolkit using the apt-get command.
sudo apt-get -y install cudaVerifying the Installation
After the installation is complete, you can verify that the correct version of the CUDA Toolkit is installed by checking the version of nvcc, the CUDA compiler.
nvcc --versionIf the installation was successful, you should see output similar to the following:
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2021 NVIDIA Corporation
Built on Fri_Dec_17_18:16:03_PST_2021
Cuda compilation tools, release 11.6, V11.6.55
Build cuda_11.6.r11.6/compiler.30794723_0Conclusion
In this article, we have guided you through the process of installing CUDA Toolkit version 11.6 on Ubuntu 18.04. We hope that this guide has been helpful to you. If you have any questions or run into any issues, please feel free to leave a comment below.
Yes, you can install CUDA Toolkit version 11.6 on Ubuntu 18.04 by following the steps outlined in this guide.
Before installing CUDA Toolkit version 11.6, you need to have a CUDA-capable GPU and a compatible version of Linux. You can check if your GPU is CUDA-capable by visiting the NVIDIA website provided in the prerequisites section of this guide.
To add the CUDA repository, you need to download the repository pin file using the wget command and then move the downloaded file to the /etc/apt/preferences.d/ directory using the mv command. The specific commands are provided in Step 1 of the installation steps.
Adding the CUDA repository key ensures that the packages you install from the repository are authentic and haven’t been tampered with. It is an important security measure.
After the installation is complete, you can verify it by checking the version of nvcc, the CUDA compiler, using the nvcc --version command. The expected output is provided in Step 6 of the installation steps.
