RAPIDS | NVIDIA Setup on Linux
Introduction
Now that you got a basic idea about GPUs and their need in data science and why technologies like RAPIDS are essential, we will show you how to set up RAPIDS on Linux (Ubuntu 20.4). Let’s start.
Installation of CUDA Toolkit
· To download the CUDA toolkit click here. Next, choose options as follows:
Operating System: Linux | Architecture: x86_64 | Distribution: Ubuntu |
Version: 20.4 |Installer: deb (network)
· You will get all the commands to install the CUDA on Ubuntu 20.4 as shown below:
Installation of Anaconda Distribution
· To download the anaconda toolkit click here.
· In case, you are facing issues while downloading the toolkit refer to alternative source instructions.
· Follow the prompts on the installer screens.
· If you are unsure about any setting, accept the defaults. You can change them later.
· To make the changes take effect, close and then re-open your terminal window.
· Test your installation in your terminal window or Anaconda Prompt, run the command conda list.
· A list of installed packages appears if it has been installed correctly.
Installation of MySQL
To install MySQL on Ubuntu run the following commands:
sudo apt install libmysqlclient-devsudo apt install mysql-server
Installation of RAPIDS
Firstly, create a new virtual environment using the following commands:
cd rapidsconda env create -f environment.yml -n MY_ENV_NAMEconda activate MY_ENV_NAME
To install RAPIDS, run the following command on the terminal:
conda install -c rapidsai rapidsconda install -c conda-forge blazeconda install -c rapidsai rapids-blazing
This step could take several minutes to complete, especially if you have less than 3Gb of RAM or packages that need to be compiled from the source. Please be patient and let it run until completion. You can grab a cup of coffee/tea in the meanwhile.
· Note: cuDF is supported only on Linux and with Python versions 3.7 and later.
· Finally, import any RAPIDS library or start a Jupyter notebook.
· And that’s it, fellas!