CuVec
=====
Unifying Python/C++/CUDA memory: Python buffered array ↔ C++11 ``std::vector`` ↔ CUDA managed memory.
|Version| |Downloads| |Py-Versions| |DOI| |Licence| |Tests| |Coverage|
.. contents:: Table of contents
:backlinks: top
:local:
Why
~~~
Data should be manipulated using the existing functionality and design paradigms of each programming language. Python code should be Pythonic. CUDA code should be... CUDActic? C code should be... er, Clean.
However, in practice converting between data formats across languages can be a pain.
Other libraries which expose functionality to convert/pass data formats between these different language spaces tend to be bloated, unnecessarily complex, and relatively unmaintainable. By comparison, ``cuvec`` uses the latest functionality of Python, C/C++11, and CUDA to keep its code (and yours) as succinct as possible. "Native" containers are exposed so your code follows the conventions of your language. Want something which works like a ``numpy.ndarray``? Not a problem. Want to convert it to a ``std::vector``? Or perhaps a raw ``float *`` to use in a CUDA kernel? Trivial.
- Less boilerplate code (fewer bugs, easier debugging, and faster prototyping)
- Fewer memory copies (faster execution)
- Lower memory usage (do more with less hardware)
Non objectives
--------------
Anything to do with mathematical functionality. The aim is to expose functionality, not (re)create it.
Even something as simple as setting element values is left to the user and/or pre-existing features - for example:
- Python: ``arr[:] = value``
- NumPy: ``arr.fill(value)``
- CuPy: ``cupy.asarray(arr).fill(value)``
- C++: ``std::fill(vec.begin(), vec.end(), value)``
- C & CUDA: ``memset(vec.data(), value, sizeof(T) * vec.size())``
Install
~~~~~~~
Requirements:
- Python 3.7 or greater (e.g. via `Anaconda or Miniconda <https://docs.conda.io/projects/conda/en/latest/user-guide/install/download.html#anaconda-or-miniconda>`_, or via ``python3-dev``)
- (optional) `CUDA SDK/Toolkit <https://developer.nvidia.com/cuda-downloads>`_ (including drivers for an NVIDIA GPU)
* note that if the CUDA SDK/Toolkit is installed *after* CuVec, then CuVec must be re-installed to enable CUDA support
.. code:: sh
pip install cuvec
Usage
~~~~~
See `the usage documentation <https://amypad.github.io/CuVec/#usage>`_ and `quick examples <https://amypad.github.io/CuVec/#examples>`_ of how to upgrade a Python ↔ C++ ↔ CUDA interface.
See also `NumCu <https://github.com/AMYPAD/NumCu>`_, a minimal stand-alone Python package built using CuVec.
External Projects
~~~~~~~~~~~~~~~~~
For integration into Python, C++, CUDA, CMake, pybind11, and general SWIG projects, see `the external project documentation <https://amypad.github.io/CuVec/#external-projects>`_.
Full and explicit example modules using the `CPython API <https://github.com/AMYPAD/CuVec/tree/main/cuvec/src/example_cpython>`_, `pybind11 API <https://github.com/AMYPAD/CuVec/tree/main/cuvec/src/example_pybind11>`_, and `SWIG <https://github.com/AMYPAD/CuVec/tree/main/cuvec/src/example_swig>`_ are also provided.
Contributing
~~~~~~~~~~~~
See `CONTRIBUTING.md <https://github.com/AMYPAD/CuVec/blob/main/CONTRIBUTING.md>`_.
Licence
~~~~~~~
|Licence| |DOI|
Copyright:
- `Casper O. da Costa-Luis <https://github.com/casperdcl>`_
- `Contributors <https://github.com/AMYPAD/cuvec/graphs/contributors>`_
.. |DOI| image:: https://zenodo.org/badge/DOI/10.5281/zenodo.4446211.svg
:target: https://doi.org/10.5281/zenodo.4446211
.. |Licence| image:: https://img.shields.io/pypi/l/cuvec.svg?label=licence
:target: https://github.com/AMYPAD/CuVec/blob/main/LICENCE
.. |Tests| image:: https://img.shields.io/github/actions/workflow/status/AMYPAD/CuVec/test.yml?branch=main&logo=GitHub
:target: https://github.com/AMYPAD/CuVec/actions
.. |Downloads| image:: https://img.shields.io/pypi/dm/cuvec?logo=pypi&logoColor=white
:target: https://pypi.org/project/cuvec
.. |Coverage| image:: https://codecov.io/gh/AMYPAD/CuVec/branch/main/graph/badge.svg
:target: https://codecov.io/gh/AMYPAD/CuVec
.. |Version| image:: https://img.shields.io/pypi/v/cuvec.svg?logo=python&logoColor=white
:target: https://github.com/AMYPAD/CuVec/releases
.. |Py-Versions| image:: https://img.shields.io/pypi/pyversions/cuvec.svg?logo=python&logoColor=white
:target: https://pypi.org/project/cuvec
Raw data
{
"_id": null,
"home_page": "",
"name": "cuvec",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.7",
"maintainer_email": "",
"keywords": "Python C C++ buffer vector array CUDA CPython SWIG pybind11 extensions API",
"author": "",
"author_email": "Casper da Costa-Luis <casper.dcl@physics.org>",
"download_url": "https://files.pythonhosted.org/packages/21/85/261081078829f2458642f9c271254da4aaf0033e1b1081d849b50f950475/cuvec-6.0.0.tar.gz",
"platform": null,
"description": "CuVec\n=====\n\nUnifying Python/C++/CUDA memory: Python buffered array \u2194 C++11 ``std::vector`` \u2194 CUDA managed memory.\n\n|Version| |Downloads| |Py-Versions| |DOI| |Licence| |Tests| |Coverage|\n\n.. contents:: Table of contents\n :backlinks: top\n :local:\n\nWhy\n~~~\n\nData should be manipulated using the existing functionality and design paradigms of each programming language. Python code should be Pythonic. CUDA code should be... CUDActic? C code should be... er, Clean.\n\nHowever, in practice converting between data formats across languages can be a pain.\n\nOther libraries which expose functionality to convert/pass data formats between these different language spaces tend to be bloated, unnecessarily complex, and relatively unmaintainable. By comparison, ``cuvec`` uses the latest functionality of Python, C/C++11, and CUDA to keep its code (and yours) as succinct as possible. \"Native\" containers are exposed so your code follows the conventions of your language. Want something which works like a ``numpy.ndarray``? Not a problem. Want to convert it to a ``std::vector``? Or perhaps a raw ``float *`` to use in a CUDA kernel? Trivial.\n\n- Less boilerplate code (fewer bugs, easier debugging, and faster prototyping)\n- Fewer memory copies (faster execution)\n- Lower memory usage (do more with less hardware)\n\nNon objectives\n--------------\n\nAnything to do with mathematical functionality. The aim is to expose functionality, not (re)create it.\n\nEven something as simple as setting element values is left to the user and/or pre-existing features - for example:\n\n- Python: ``arr[:] = value``\n- NumPy: ``arr.fill(value)``\n- CuPy: ``cupy.asarray(arr).fill(value)``\n- C++: ``std::fill(vec.begin(), vec.end(), value)``\n- C & CUDA: ``memset(vec.data(), value, sizeof(T) * vec.size())``\n\nInstall\n~~~~~~~\n\nRequirements:\n\n- Python 3.7 or greater (e.g. via `Anaconda or Miniconda <https://docs.conda.io/projects/conda/en/latest/user-guide/install/download.html#anaconda-or-miniconda>`_, or via ``python3-dev``)\n- (optional) `CUDA SDK/Toolkit <https://developer.nvidia.com/cuda-downloads>`_ (including drivers for an NVIDIA GPU)\n\n * note that if the CUDA SDK/Toolkit is installed *after* CuVec, then CuVec must be re-installed to enable CUDA support\n\n.. code:: sh\n\n pip install cuvec\n\nUsage\n~~~~~\n\nSee `the usage documentation <https://amypad.github.io/CuVec/#usage>`_ and `quick examples <https://amypad.github.io/CuVec/#examples>`_ of how to upgrade a Python \u2194 C++ \u2194 CUDA interface.\n\nSee also `NumCu <https://github.com/AMYPAD/NumCu>`_, a minimal stand-alone Python package built using CuVec.\n\nExternal Projects\n~~~~~~~~~~~~~~~~~\n\nFor integration into Python, C++, CUDA, CMake, pybind11, and general SWIG projects, see `the external project documentation <https://amypad.github.io/CuVec/#external-projects>`_.\nFull and explicit example modules using the `CPython API <https://github.com/AMYPAD/CuVec/tree/main/cuvec/src/example_cpython>`_, `pybind11 API <https://github.com/AMYPAD/CuVec/tree/main/cuvec/src/example_pybind11>`_, and `SWIG <https://github.com/AMYPAD/CuVec/tree/main/cuvec/src/example_swig>`_ are also provided.\n\nContributing\n~~~~~~~~~~~~\n\nSee `CONTRIBUTING.md <https://github.com/AMYPAD/CuVec/blob/main/CONTRIBUTING.md>`_.\n\nLicence\n~~~~~~~\n\n|Licence| |DOI|\n\nCopyright:\n\n- `Casper O. da Costa-Luis <https://github.com/casperdcl>`_\n- `Contributors <https://github.com/AMYPAD/cuvec/graphs/contributors>`_\n\n.. |DOI| image:: https://zenodo.org/badge/DOI/10.5281/zenodo.4446211.svg\n :target: https://doi.org/10.5281/zenodo.4446211\n.. |Licence| image:: https://img.shields.io/pypi/l/cuvec.svg?label=licence\n :target: https://github.com/AMYPAD/CuVec/blob/main/LICENCE\n.. |Tests| image:: https://img.shields.io/github/actions/workflow/status/AMYPAD/CuVec/test.yml?branch=main&logo=GitHub\n :target: https://github.com/AMYPAD/CuVec/actions\n.. |Downloads| image:: https://img.shields.io/pypi/dm/cuvec?logo=pypi&logoColor=white\n :target: https://pypi.org/project/cuvec\n.. |Coverage| image:: https://codecov.io/gh/AMYPAD/CuVec/branch/main/graph/badge.svg\n :target: https://codecov.io/gh/AMYPAD/CuVec\n.. |Version| image:: https://img.shields.io/pypi/v/cuvec.svg?logo=python&logoColor=white\n :target: https://github.com/AMYPAD/CuVec/releases\n.. |Py-Versions| image:: https://img.shields.io/pypi/pyversions/cuvec.svg?logo=python&logoColor=white\n :target: https://pypi.org/project/cuvec\n",
"bugtrack_url": null,
"license": "MPL-2.0",
"summary": "Unifying Python/C++/CUDA memory: Python buffered array -> C++11 `std::vector` -> CUDA managed memory",
"version": "6.0.0",
"project_urls": {
"Changelog": "https://github.com/AMYPAD/CuVec/releases",
"Documentation": "https://amypad.github.io/CuVec",
"Repository": "https://github.com/AMYPAD/CuVec"
},
"split_keywords": [
"python",
"c",
"c++",
"buffer",
"vector",
"array",
"cuda",
"cpython",
"swig",
"pybind11",
"extensions",
"api"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "2185261081078829f2458642f9c271254da4aaf0033e1b1081d849b50f950475",
"md5": "a219dc34556ee7f7612558180dd92bda",
"sha256": "5f0471d6a17cb653e452383bede361866065234b40fa29ae477315139e936677"
},
"downloads": -1,
"filename": "cuvec-6.0.0.tar.gz",
"has_sig": false,
"md5_digest": "a219dc34556ee7f7612558180dd92bda",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.7",
"size": 27585,
"upload_time": "2024-03-13T20:08:56",
"upload_time_iso_8601": "2024-03-13T20:08:56.784146Z",
"url": "https://files.pythonhosted.org/packages/21/85/261081078829f2458642f9c271254da4aaf0033e1b1081d849b50f950475/cuvec-6.0.0.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2024-03-13 20:08:56",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "AMYPAD",
"github_project": "CuVec",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"lcname": "cuvec"
}
JSON
