I've found some discussion about it from early 2012, but besides that it doesn't seem it got implemented at all, is that correct?
-- Gabriel.
Jérôme Kieffer
unread,
Jan 20, 2013, 3:06:54 PM1/20/13
Reply to author
Sign in to reply to author
Forward
Sign in to forward
Delete
You do not have permission to delete messages in this group
Copy link
Report message
Show original message
Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message
to cython...@googlegroups.com
I was playing with OpenCL kernels + cython wrapped C++ code and translated it into plain Python + pyopencl.
No performance loss was observed. pyopencl is really great (even if bound with boost)
I know it does not answer your question but it may explain the lack of interest on the Cython/OpenCL part.
--
Jérôme Kieffer <goo...@terre-adelie.org>
Dag Sverre Seljebotn
unread,
Jan 20, 2013, 3:12:55 PM1/20/13
Reply to author
Sign in to reply to author
Forward
Sign in to forward
Delete
You do not have permission to delete messages in this group
Copy link
Report message
Show original message
Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message
to cython...@googlegroups.com
On 01/17/2013 12:27 PM, Gabriel Jacobo wrote:
> Hey everyone, I was wondering if someone could give me more information
> about the implementation of the OpenCL proposal:
> http://wiki.cython.org/enhancements/opencl
I think the closest you'll come in spirit is the CUDA support in Numba.
Dag Sverre
Gabriel Jacobo
unread,
Jan 20, 2013, 6:44:54 PM1/20/13
Reply to author
Sign in to reply to author
Forward
Sign in to forward
Delete
You do not have permission to delete messages in this group
Copy link
Report message
Show original message
Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message
Thanks, I was looking at something more along the lines of the prange/openmp implementation, but I guess a more "manual" approach won't be that difficult to implement.