CuPy

NumFOCUS Sponsored Project since 2022

CuPy is a GPU-based Python library for data analysis in numerical computing. CuPy shares the same API as NumPy and SciPy, allowing the use of GPUs at a minimal learning cost and without the need for in-depth knowledge of low-level GPU programming.

Share This Project:



Industry

Business & Industry Applications, Higher Education Research & Teaching, Government



Language

Python, Cython, CUDA



Features

Data Wrangling, Numerical Computing, High Performance Computing, Big Data, Statistical Computing, Numerical Computing

CuPy is an open source library for GPU-accelerated computing with Python, providing support for N-dimensional (dense) arrays, sparse matrices, and a variety of numerical algorithms implemented on top of them. CuPy acts as a drop-in replacement to run existing NumPy/SciPy code on NVIDIA CUDA or AMD ROCm platforms.

Projects in a variety of fields are built upon CuPy, including spaCy (natural language processing library), pomegranate (library for probabilistic modelling), RAPIDS (GPU data analysis suite by NVIDIA), PyQtGraph (graphics library for Qt), and turboSETI (Python based SETI search algorithm by UC Berkeley). See https://github.com/cupy/cupy/wiki/Projects-using-CuPy for more information.

Contribute Code

Website

Donate



Be the First to Know



Be the First to Know

New developments and features from our sponsored projects, straight to your inbox, once a month.