python3-pycuda - Python 3 module to access Nvidia‘s CUDA parallel computation API

Property Value
Distribution Debian 8 (Jessie)
Repository Debian Contrib i386
Package filename python3-pycuda_2014.1-3_i386.deb
Package name python3-pycuda
Package version 2014.1
Package release 3
Package architecture i386
Package type deb
Category contrib/python
License -
Maintainer Tomasz Rybak <>
Download size 288.95 KB
Installed size 1.66 MB
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
* 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.
* 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.
This package contains Python 3 modules.


Package Version Architecture Repository
python3-pycuda_2014.1-3_amd64.deb 2014.1 amd64 Debian Contrib
python3-pycuda - - -


Name Value
libboost-python1.55.0 -
libboost-thread1.55.0 -
libc6 >= 2.4
libcuda1 >= 331.38
libcurand6.0 >= 4.0
libgcc1 >= 1:4.1.1
libstdc++6 >= 4.4.0
nvidia-cuda-toolkit -
python3 >= 3.4~
python3 << 3.5
python3-appdirs -
python3-decorator -
python3-numpy >= 1:1.8.0
python3-numpy-abi9 -
python3-pytools -


Name Value
python-pycuda-headers -


Type URL
Binary Package python3-pycuda_2014.1-3_i386.deb
Source Package pycuda

Install Howto

  1. Update the package index:
    # sudo apt-get update
  2. Install python3-pycuda deb package:
    # sudo apt-get install python3-pycuda




2014-11-29 - Tomasz Rybak <>
pycuda (2014.1-3) unstable; urgency=medium
* Add dependency to appdirs and use it to create cache directories
(Closes: #770689).
2014-10-05 - Tomasz Rybak <>
pycuda (2014.1-2) unstable; urgency=medium
* Install OpenGL and sparse modules files for both Python 2 and 3.
* Fix problems with debian/copyright pointed by lintian.
* Update Standards-Version to 3.9.6; no changes necessary.
2014-08-24 - Tomasz Rybak <>
pycuda (2014.1-1) unstable; urgency=medium
* New upstream release.
* Remove dependency on libcuda1, leaving nvidia-cuda-toolkit dependency,
which depend on libcuda1 (Closes: #749428).
* Change d/rules get-orig-source to pack *.orig.tar with xz, not with gzip.
* Change d/control to note that upstream stopped supporting Python 2.5.
2014-06-08 - Tomasz Rybak <>
pycuda (2013.1.1+git20140527-1) unstable; urgency=medium
* New upstream version.
* Fix Vcs-Browser link (Closes: #749489).
* Rebuild to remove Python 3.3 dependency (Closes: #750886).
2014-03-22 - Tomasz Rybak <>
pycuda (2013.1.1+git20140310-1) unstable; urgency=low
* New upstream version.
* Change python-pycuda-doc recommendation of Python and Python 3 packages
to suggestions to avoid installing unwanted packages, like in #739173.
* Rebuild with Python 3.4 support.
* Update Standards-Version to 3.9.5; no changes necessary.
2013-11-28 - Tomasz Rybak <>
pycuda (2013.1.1+git20131128-1) unstable; urgency=low
* New upstream version.
* Rebuild against CUDA 5.5 (Closes: #730263).
2013-07-17 - Tomasz Rybak <>
pycuda (2013.1.1-1) unstable; urgency=low
* New upstream release.
2013-06-30 - Tomasz Rybak <>
pycuda (2013.1~git20130626-1) unstable; urgency=low
* New upstream release, adding new CURAND support.
2013-06-20 - Tomasz Rybak <>
pycuda (2013.1~git20130616-1) unstable; urgency=low
* New upstream release.
* Add Python 3 package.
* Add *-dbg packages.
* Remove *-headers package; upstream now ships headers in python directory.
* Upstream switched to setuptools, update patches to deal with it.
* Do not call dh_sphinxdoc for architecture dependent builds, build
documentation in separate step in debian/rules.
* Fix CUDA library build-dependencies, remove old libcuda1-dev.
* Add README.source describing get-orig-source in debian/rules.
* Disable package optimisation option as it might clash with hardening
compile options.
* Include tests (all files in test/) in documentation package.
* Fix VCS-* fields to contain canonical URIs.
* Update Standards-Version to 3.9.4; no changes necessary.
2012-06-21 - Tomasz Rybak <>
pycuda (2012.1-1) unstable; urgency=low
* New upstream release.
* Change my email.

See Also

Package Description
python3-vertica_0.2.3-1_all.deb native Python client for the Vertica database (Python 3)
qmhandle_1.3.2-1_all.deb tool to manage the qmail message queue
quake-server_8_all.deb classic first person shooter - server and init script
quake2-server_8_all.deb dedicated server for Quake II
quake2_8_all.deb classic science-fiction-themed first person shooter
quake3-server_8_all.deb Quake III Arena dedicated server launcher scripts
quake3_8_all.deb Quake III Arena menu entry and launcher scripts
quake_8_all.deb classic gothic/horror-themed first person shooter
raccoon_1.0b-1_all.deb preparation of in silico drug screening projects
reminiscence_0.2.1-2_i386.deb free implementation of Flashback game engine
residualvm-data_0.1.1+dfsg-3_all.deb interpreter to play LucasArts' Lua-based 3D adventures (data files)
residualvm_0.1.1+dfsg-3_i386.deb interpreter to play LucasArts' Lua-based 3D adventures
rocksndiamonds_3.3.0.1+dfsg1-2.2_i386.deb arcade-style game
rott_1.1.2-2_i386.deb SDL-Port of Rise of the Triad
ruby-googlecharts_1.6.8-2_all.deb Ruby library to generate charts using Google Chart API