Website | Docs | Install Guide | Tutorial | Examples (Official) | Forum (en, ja)

CuPy is an implementation of NumPy-compatible multi-dimensional array on CUDA. CuPy consists of the core multi-dimensional array class, cupy.ndarray, and many functions on it. It supports a subset of numpy.ndarray interface.

For detailed instructions on installing CuPy, see the installation guide.

You can install CuPy using pip:

$ pip install cupy

Note that if you want to enable CUDA, cuDNN, and/or NCCL, they need to be set up before installation of CuPy.

We provide the official Docker image. Use nvidia-docker command to run CuPy image with GPU. You can login to the environment with bash, and run the Python interpreter.

$ nvidia-docker run -it cupy/cupy /bin/bash

Please see the contribution guide.

• Release notes
• Projects using CuPy

MIT License (see LICENSE file).

Ryosuke Okuta, Yuya Unno, Daisuke Nishino, Shohei Hido and Crissman Loomis. CuPy: A NumPy-Compatible Library for NVIDIA GPU Calculations. Proceedings of Workshop on Machine Learning Systems (LearningSys) in The Thirty-first Annual Conference on Neural Information Processing Systems (NIPS), (2017). URL

@inproceedings{cupy_learningsys2017,
  author       = "Okuta, Ryosuke and Unno, Yuya and Nishino, Daisuke and Hido, Shohei and Loomis, Crissman",
  title        = "CuPy: A NumPy-Compatible Library for NVIDIA GPU Calculations",
  booktitle    = "Proceedings of Workshop on Machine Learning Systems (LearningSys) in The Thirty-first Annual Conference on Neural Information Processing Systems (NIPS)",
  year         = "2017",
  url          = "http://learningsys.org/nips17/assets/papers/paper_16.pdf"
}