pycuda

CUDA integration for Python, plus shiny features

https://github.com/inducer/pycuda

Science Score: 64.0%

This score indicates how likely this project is to be science-related based on various indicators:

✓
CITATION.cff file
Found CITATION.cff file
✓
codemeta.json file
Found codemeta.json file
✓
.zenodo.json file
Found .zenodo.json file
○
DOI references
✓
Academic publication links
Links to: zenodo.org
✓
Committers with academic emails
21 of 82 committers (25.6%) from academic institutions
○
Institutional organization owner
○
JOSS paper metadata
○
Scientific vocabulary similarity
Low similarity (9.0%) to scientific vocabulary

Keywords

array cuda gpu gpu-computing multidimensional-arrays pycuda python scientific-computing

Keywords from Contributors

gtk qt tk wx ansible closember multiprocessing parallel-computing physics software-carpentry

Last synced: 4 months ago · JSON representation ·

Repository

CUDA integration for Python, plus shiny features

Basic Info

Host: GitHub
Owner: inducer
License: other
Language: Python
Default Branch: main
Homepage: http://mathema.tician.de/software/pycuda
Size: 2.95 MB

Statistics

Stars: 1,980
Watchers: 57
Forks: 293
Open Issues: 87
Releases: 9

Topics

array cuda gpu gpu-computing multidimensional-arrays pycuda python scientific-computing

Created over 14 years ago · Last pushed 7 months ago

Metadata Files

Readme License Citation

README.rst

PyCUDA: Pythonic Access to CUDA, with Arrays and Algorithms
=============================================================

.. image:: https://gitlab.tiker.net/inducer/pycuda/badges/main/pipeline.svg
    :alt: Gitlab Build Status
    :target: https://gitlab.tiker.net/inducer/pycuda/commits/main
.. image:: https://badge.fury.io/py/pycuda.png
    :target: https://pypi.org/project/pycuda
.. image:: https://zenodo.org/badge/1575319.svg
    :alt: Zenodo DOI for latest release
    :target: https://zenodo.org/badge/latestdoi/1575319

PyCUDA lets you access `Nvidia `_'s `CUDA
`_ parallel computation API from Python.
Several wrappers of the CUDA API already exist-so what's so special
about PyCUDA?

* Object cleanup tied to lifetime of objects. This idiom, often
  called
  `RAII `_
  in C++, makes it much easier to write correct, leak- and
  crash-free code. PyCUDA knows about dependencies, too, so (for
  example) it won't detach from a context before all memory
  allocated in it is also freed.

* Convenience. Abstractions like pycuda.driver.SourceModule and
  pycuda.gpuarray.GPUArray make CUDA programming even more
  convenient than with Nvidia's C-based runtime.

* Completeness. PyCUDA puts the full power of CUDA's driver API at
  your disposal, if you wish. It also includes code for
  interoperability with OpenGL.

* Automatic Error Checking. All CUDA errors are automatically
  translated into Python exceptions.

* Speed. PyCUDA's base layer is written in C++, so all the niceties
  above are virtually free.

* Helpful `Documentation `_.

Relatedly, like-minded computing goodness for `OpenCL `_
is provided by PyCUDA's sister project `PyOpenCL `_.

Owner

Name: Andreas Klöckner
Login: inducer
Kind: user
Location: Champaign, IL
Company: Scientific Computing, CS@UIUC

Website: http://andreask.cs.illinois.edu/aboutme
Repositories: 225
Profile: https://github.com/inducer

Citation (CITATION.cff)

cff-version: 1.2.0
message: "If you use this software, please cite it as below."
authors:
- family-names: "Kloeckner"
  given-names: "Andreas"
  orcid: "https://orcid.org/0000-0003-1228-519X"
- family-names: Wohlgemuth
  given-names: Gert
- family-names: Lee
  given-names: Gregory
- family-names: Rybak
  given-names: Tomasz
- family-names: Nitz
  given-names: Alex
- family-names: Chiang
  given-names: David
- family-names: Seibert
  given-names: Stan
- family-names: Bergtholdt
  given-names: Martin
- family-names: Unterthiner
  given-names: Thomas
- family-names: Markall
  given-names: Graham
- family-names: Kotak
  given-names: Mit
- family-names: Favre-Nicolin
  given-names: Vincent
- family-names: Opanchuk
  given-names: Bogdan
- family-names: Merry
  given-names: Bruce
- family-names: Pinto
  given-names: Nicolas
- family-names: Milo
  given-names: Fabrizio
- family-names: Collignon
  given-names: Thomas
- family-names: Rathgeber
  given-names: Florian
- family-names: Perkins
  given-names: Simon
- family-names: Rutsky
  given-names: Vladimir
- family-names: Catanzaro
  given-names: Bryan
- family-names: Park
  given-names: Alex
- family-names: Witherden
  given-names: Freddie
- family-names: E. Givon
  given-names: Lev
- family-names: Pfister
  given-names: Luke
- family-names: Brubaker
  given-names: Marcus
- family-names: ZA
  given-names: RA
- family-names: Hausammann
  given-names: Loic
- family-names: Gohlke
  given-names: Christoph

title: "PyCUDA"
version: 2022.1
doi: 10.5281/zenodo.6728391
date-released: 2022-06-10
url: "https://github.com/inducer/pycuda"
license: MIT

GitHub Events

Total

Create event: 2
Release event: 2
Issues event: 13
Watch event: 140
Delete event: 1
Issue comment event: 21
Push event: 7
Pull request event: 4
Fork event: 11

Last Year

Create event: 2
Release event: 2
Issues event: 13
Watch event: 140
Delete event: 1
Issue comment event: 21
Push event: 7
Pull request event: 4
Fork event: 11

Committers

Last synced: 8 months ago

All Time

Total Commits: 1,356
Total Committers: 82
Avg Commits per committer: 16.537
Development Distribution Score (DDS): 0.355

Past Year

Commits: 23
Committers: 4
Avg Commits per committer: 5.75
Development Distribution Score (DDS): 0.217

Top Committers

Name	Email	Commits
Andreas Kloeckner	i**m@t**t	875
wohlgemuth	w**h@u**u	65
Andreas Kloeckner	k**r@h**)	56
Andreas Kloeckner	k**r@h**u	44
HendrikRiedmann	r**n@d**u	35
Lurch	L****h	15
Pierre Paleo	p**o@e**r	15
Tomasz Rybak	t**k@p**l	14
Gregory R. Lee	g**e@c**g	14
Kaushik Kulkarni	k**d@g**m	13
Mit Kotak	m**2@i**u	11
Vincent Favre-Nicolin	f**e@e**r	11
Alexander Harvey Nitz	a**z@s**u	10
Andreas Kloeckner	k**r@t**)	9
Stan Seibert	s**n@m**g	9
Roberto Zamora-Zamora	r**z@f**x	8
Thomas Unterthiner	t**r@g**t	8
Martin Bergtholdt	m**t@p**m	8
davidweichiang	d**g@n**u	8
Bogdan Opanchuk	m**r@g**m	7
Graham Markall	g**l@n**m	7
Bruce Merry	b**y@s**a	6
Nicolas Pinto	p**o@m**u	5
Nicolas Pinto	p**o@a**u	5
Bogdan Opanchuk	b**n@o**t	5
Guillaume THOMAS-COLLIGNON	g**t@n**m	4
Fabrizio Milo (misto)	m**n@g**m	4
Dzhelil Rufat	d**t@c**u	4
Hannes Bretschneider	h**s@p**a	4
Bryan Catanzaro	c**r@c**u	3
and 52 more...

Committer Domains (Top 20 + Academic)

nvidia.com: 3 esrf.fr: 2 illinois.edu: 2 cs.berkeley.edu: 1 psi.utoronto.ca: 1 caltech.edu: 1 opanchuk.net: 1 alum.mit.edu: 1 mit.edu: 1 ska.ac.za: 1 nd.edu: 1 philips.com: 1 gmx.net: 1 fisica.unam.mx: 1 mtrr.org: 1 teramite.(none): 1 syr.edu: 1 cchmc.org: 1 post.pl: 1 dam.brown.edu: 1 hpcgeek.ccv.brown.edu: 1 ucdavis.edu: 1 flatironinstitute.org: 1 mat.ethz.ch: 1 ee.columbia.edu: 1 cuda.es.its.nyu.edu: 1 walrus.dam.brown.edu: 1 columbia.edu: 1 uci.edu: 1 biac.duke.edu: 1

Issues and Pull Requests

Last synced: 4 months ago

All Time

Total issues: 110
Total pull requests: 40
Average time to close issues: 4 months
Average time to close pull requests: 27 days
Total issue authors: 94
Total pull request authors: 22
Average comments per issue: 2.58
Average comments per pull request: 1.98
Merged pull requests: 30
Bot issues: 0
Bot pull requests: 0

Past Year

Issues: 13
Pull requests: 3
Average time to close issues: 7 days
Average time to close pull requests: about 6 hours
Issue authors: 12
Pull request authors: 2
Average comments per issue: 1.23
Average comments per pull request: 0.0
Merged pull requests: 2
Bot issues: 0
Bot pull requests: 0

View more stats

Top Authors

Issue Authors

mitkotak (6)
anbe42 (3)
pierrepaleo (3)
inducer (3)
vincefn (3)
ccarmatic (2)
bmerry (2)
jmpolinski (2)
ishandutta0098 (2)
timgates42 (1)
thuzhf (1)
ymmx2 (1)
ilciavo (1)
denniswittich (1)
pwalsh0 (1)

Pull Request Authors

inducer (12)
totaam (5)
vincefn (5)
zzjjbb (3)
matthiasdiener (2)
namoshizun (2)
guoci (2)
mitkotak (2)
bmerry (2)
isuruf (1)
bastidest (1)
kaushikcfd (1)
brumen (1)
hyperfraise (1)
ByLamacq (1)

Top Labels

Issue Labels

bug (53) enhancement (7)

Pull Request Labels

Packages

Total packages: 3
Total downloads:
- pypi 49,390 last-month
Total docker downloads: 14,044

Total dependent packages: 58
(may contain duplicates)
Total dependent repositories: 547
(may contain duplicates)
Total versions: 61
Total maintainers: 2

pypi.org: pycuda

Python wrapper for Nvidia CUDA

Homepage: http://mathema.tician.de/software/pycuda
Documentation: https://pycuda.readthedocs.io/
License: MIT
Latest release: 2025.1.1
published 7 months ago

Versions: 57
Dependent Packages: 57
Dependent Repositories: 544
Downloads: 48,664 Last month
Docker Downloads: 14,044

Rankings

Dependent packages count: 0.3%

Dependent repos count: 0.6%

Docker downloads count: 1.3%

Average: 1.4%

Downloads: 1.6%

Stargazers count: 1.7%

Forks count: 3.1%

Maintainers (1)

inducer77

Last synced: 4 months ago

conda-forge.org: pycuda

PyCUDA lets you access GPUs from Python, through the CUDA parallel compute interface.

Homepage: http://pypi.python.org/pypi/pycuda
License: MIT
Latest release: 2022.1
published over 3 years ago

Versions: 3
Dependent Packages: 1
Dependent Repositories: 3

Rankings

Stargazers count: 10.4%

Forks count: 10.7%

Average: 17.1%

Dependent repos count: 18.1%

Dependent packages count: 29.0%

Last synced: 4 months ago

pypi.org: pycuda-gml

Python wrapper for Nvidia CUDA

Homepage: http://mathema.tician.de/software/pycuda
Documentation: https://pycuda-gml.readthedocs.io/
License: MIT
Latest release: 2024.1.2
published over 1 year ago

Versions: 1
Dependent Packages: 0
Dependent Repositories: 0
Downloads: 726 Last month

Rankings

Dependent packages count: 10.4%

Average: 34.5%

Dependent repos count: 58.7%

Maintainers (1)

gml-buildbot

Last synced: 4 months ago

Dependencies

.github/workflows/autopush.yml actions

actions/checkout v3 composite

.github/workflows/ci.yml actions

actions/checkout v3 composite
actions/setup-python v4 composite

setup.py pypi

appdirs >=1.4.0
mako *
pytools >=2011.2

pyproject.toml pypi