Introduction
Pandas is an indispensable tool for many data scientists and analysts. Its intuitiveness and flexibility make manipulating and analyzing data a breeze. But with large datasets, Pandas’ familiar performance bottlenecks start to show. There are now many ways to speed up Pandas. This article looks at cuDF, which uses GPU acceleration, and walks you step by step through installing it on WSL2.
Installing cuDF
Step 1: Prepare the WSL2 environment
Before installing cuDF, make sure WSL2 is installed and configured on your Windows system. WSL2 not only provides a full Linux environment but also works directly with the Windows file system, making it an ideal platform for high-performance computing. If you haven’t installed WSL2 yet, follow Microsoft’s official guide.
Step 2: Install CUDA
cuDF’s performance depends on NVIDIA’s CUDA, so installing CUDA is a prerequisite for GPU acceleration. See my earlier article on installing CUDA and PyTorch on WSL2.
Step 3: Install cuDF
First, it has some requirements for your machine. Check whether yours meets them before continuing:
- OS: Windows 11 with Ubuntu 22.04 instance for WSL2. - WSL Version: WSL2 (WSL1 not supported). - GPU: GPUs with Compute Capability 7.0 or higher (16GB+ GPU RAM is recommended).
Next we need to install conda-libmamba-solver. It’s a Conda component that uses libmamba as the underlying solver for package dependencies and environment resolution. libmamba is the core of the Mamba project, an alternative to Conda that aims to provide faster package management and resolution. If you’re interested, see Mamba on GitHub.
conda install -n base conda-libmamba-solverOnce that’s installed, you can find the install command that matches your setup on the RAPIDS website:
You can check your CUDA version with nvidia-smi.
If you install with the commands above, a new conda env called rapids-23.12 will be created. Enter the env and you can check whether cudf installed successfully.
conda activate rapids-23.12
conda listSupport
If this article helped you, or you’d like to encourage me to keep writing, you can clap for it or buy me a coffee through the link below. Thank you for your support!