PyOpenCL is an open source, multiplatform and completely free command-line software project implemented in Python and designed from the ground up to provide a cross-platform wrapper for the OpenCL (Open Computing Language) framework.
Tested with the Nvidia, AMD and Apple CL implementations
The software is engineered in such a way that it lets users access GPUs (Graphics Processing Units), as well as other densely parallel computing devices directly from the Python programming language. It has been successfully tested with the Nvidia, AMD and Apple CL implementations.
Provides completeness, automatic error checking
Being as well written as the PyCUDA project, PyOpenCL provides completeness, automatic error checking, speed, object cleanup, broad support, as well as complete and helpful documentation. A Wiki with comprehensive documentation will help developers get started with the PyOpenCL project in no time.
It’s a very fast application that provides automatic error checking, translating all OpenCL errors into Python exceptions. It helps you write leak-free and crash-free code, thanks to its object cleanup functionality that ties to lifetime of objects.
Getting started with PyOpenCL
The program is distributed as a universal source package for all supported operating systems. To install it on your GNU/Linux distribution, you must download the latest release from either Softpedia or the project’s official website (see the homepage link at the end of the article), save the archive somewhere on your computer and extract its contents using an archive manager tool.
Then, open a Terminal app, move to the location where you’ve extracted the archive file (e.g. cd /home/softpedia/pyopencl-2014.1) and run the ‘sudo python setup.py install’ command to install PyOpenCL system wide and make it available to all users on your operating system. The software can be installed on computers supporting either of the 32 or 64-bit CPU architectures.