CuPy-Xarray: Xarray on GPUs!

Overview

CuPy-Xarray is a Python library that leverages CuPy, a GPU array library, and Xarray, a library for multi-dimensional labeled array computations, to enable fast and efficient data processing on GPUs. By combining the capabilities of CuPy and Xarray, CuPy-Xarray provides a convenient interface for performing accelerated computations and analysis on large multidimensional datasets.

Installation

cupy-xarray will use an existing cupy installation, hence cupy needs to be installed manually! Please follow cupy’s install instructions at https://docs.cupy.dev/en/stable/install.html.

CuPy-Xarray can be installed using conda or pip:

From Conda Forge:

conda install -c conda-forge cupy-xarray

From PyPI:

pip install cupy-xarray

The latest version from Github:

pip install git+https://github.com/xarray-contrib/cupy-xarray.git

Extras

To enable the cog3pio backend for reading TIFFs (only available for linux-x86_64 and linux-aarch64), do:

From Conda Forge:

conda install -c conda-forge cupy-xarray cog3pio

From PyPI:

pip install cupy-xarray[tiff]

Acknowledgements

Large parts of this documentations comes from SciPy 2023 Xarray on GPUs tutorial and this NCAR tutorial to GPUs.

Contents

User Guide:

User Guide

• Basics of CuPy
• Introduction
• Basic Computations
• High-level Computation
• Custom Kernels with apply_ufunc
• A real world example
• Reading TIFFs to CUDA memory
• Summary

Tutorials & Presentations:

Tutorials & Presentations

• Tutorials and Presentations

Contributing:

Contributing

• Contributing guide

Reference:

Reference

• API Reference
• Changelog