|
Name |
|
Date |
Size |
#Lines |
LOC |
| .. | | 03-May-2022 | - |
| contrib/ | H | 14-Sep-2018 | - | 1,565 | 1,127 |
| doc/ | H | 14-Sep-2018 | - | 3,624 | 2,475 |
| examples/ | H | 14-Sep-2018 | - | 1,531 | 1,173 |
| pyopencl/ | H | 14-Sep-2018 | - | 25,235 | 19,528 |
| pyopencl.egg-info/ | H | 03-May-2022 | - | 82 | 66 |
| src/ | H | 14-Sep-2018 | - | 7,890 | 6,390 |
| test/ | H | 03-May-2022 | - | 4,254 | 3,056 |
| LICENSE | H A D | 06-Aug-2018 | 4.9 KiB | 104 | 78 |
| MANIFEST.in | H A D | 14-Sep-2018 | 527 | 29 | 24 |
| Makefile.in | H A D | 06-Aug-2018 | 323 | 22 | 15 |
| PKG-INFO | H A D | 14-Sep-2018 | 3.8 KiB | 82 | 66 |
| README.rst | H A D | 06-Aug-2018 | 2.4 KiB | 55 | 40 |
| aksetup_helper.py | H A D | 03-May-2022 | 26.3 KiB | 826 | 634 |
| configure.py | H A D | 06-Aug-2018 | 131 | 7 | 3 |
| setup.cfg | H A D | 14-Sep-2018 | 189 | 10 | 7 |
| setup.py | H A D | 21-Aug-2018 | 12.3 KiB | 353 | 275 |
README.rst
1PyOpenCL: Pythonic Access to OpenCL, with Arrays and Algorithms
2---------------------------------------------------------------
3
4.. image:: https://gitlab.tiker.net/inducer/pyopencl/badges/master/pipeline.svg
5 :target: https://gitlab.tiker.net/inducer/pyopencl/commits/master
6.. image:: https://badge.fury.io/py/pyopencl.png
7 :target: http://pypi.python.org/pypi/pyopencl
8
9PyOpenCL lets you access GPUs and other massively parallel compute
10devices from Python. It tries to offer computing goodness in the
11spirit of its sister project `PyCUDA <http://mathema.tician.de/software/pycuda>`_:
12
13* Object cleanup tied to lifetime of objects. This idiom, often
14 called
15 `RAII <http://en.wikipedia.org/wiki/Resource_Acquisition_Is_Initialization>`_
16 in C++, makes it much easier to write correct, leak- and
17 crash-free code.
18
19* Completeness. PyOpenCL puts the full power of OpenCL's API at
20 your disposal, if you wish. Every obscure `get_info()` query and
21 all CL calls are accessible.
22
23* Automatic Error Checking. All CL errors are automatically
24 translated into Python exceptions.
25
26* Speed. PyOpenCL's base layer is written in C++, so all the niceties
27 above are virtually free.
28
29* Helpful and complete `Documentation <http://documen.tician.de/pyopencl>`_
30 as well as a `Wiki <http://wiki.tiker.net/PyOpenCL>`_.
31
32* Liberal license. PyOpenCL is open-source under the
33 `MIT license <http://en.wikipedia.org/wiki/MIT_License>`_
34 and free for commercial, academic, and private use.
35
36* Broad support. PyOpenCL was tested and works with Apple's, AMD's, and Nvidia's
37 CL implementations.
38
39What you'll need:
40
41* gcc/g++ at or newer than version 4.8.2 and binutils at or newer than 2.23.52.0.1-10
42 (CentOS version number).
43 On Windows, use the `mingwpy <https://anaconda.org/carlkl/mingwpy>`_ compilers.
44* `numpy <http://numpy.org>`_, and
45* an OpenCL implementation. (See this `howto <http://wiki.tiker.net/OpenCLHowTo>`_ for how to get one.)
46
47Places on the web related to PyOpenCL:
48
49* `Python package index <http://pypi.python.org/pypi/pyopencl>`_ (download releases)
50
51* `C. Gohlke's Windows binaries <http://www.lfd.uci.edu/~gohlke/pythonlibs/#pyopencl>`_ (download Windows binaries)
52* `Github <http://github.com/inducer/pyopencl>`_ (get latest source code, file bugs)
53* `Documentation <http://documen.tician.de/pyopencl>`_ (read how things work)
54* `Wiki <http://wiki.tiker.net/PyOpenCL>`_ (read installation tips, get examples, read FAQ)
55