[FFmpeg-devel] GPU Hardware Acceleration [was Re: openCL support]

[Patrick Shirkey]

> On Thu, Mar 15, 2012 at 01:08:40AM +0100, Patrick Shirkey wrote:
>>
>> > On Wed, Mar 14, 2012 at 11:56:15PM +0100, Patrick Shirkey wrote:
>> >>
>> >> > On Wed, Mar 14, 2012 at 03:23:39AM +0100, Patrick Shirkey wrote:
>> >> >> Hi,
>> >> >>
>> >> >> Maybe it would help if I rephrased the question.
>> >> >>
>> >> >> "Are there any parts of FFMPEG that would benefit from Hardware
>> >> >> Acceleration that do not already have development in process?"
>> >> >
>> >> > libavfilter, high quality deinterlace, noise reduce, temporal
>> >> > interpolation, postprocess (spp in the highest quality setting
>> comes
>> >> > to mind as something thats relatively simple, very computationally
>> >> > intensive and highly parallizeable)
>> >> >
>> >>
>> >> Thanks Michael for your valuable feedback.
>> >>
>> >> Of these items can we prioritise them for development? They are all
>> >> useful
>> >> but which would have the most benefit?
>> >
>> > maybe spp as its too slow in its highest quality setting
>> >
>>
>> But you're not sure ;-)  Do other devs want to add their 2c?
>
> If you really care about this: Get something started (preferably even
> something simpler that spp), get the infrastructure in and then look
> for something useful.
> People trying to do "something useful" usually end up frustrated and
> give up before anything came out of it at all, so if you just manage to
> do something completely useless that would be great progress compared
> to that.


Thank Reimar,

I know I'm new to the list but I have been around in Linux Audio Dev for
about 14 years.

I'm trying to piece together a plan of attack for a larger project that I
am working on. FFMPEG is a core component so I am trying to get a real
idea of what I am up against.

At the moment It looks like it would take around 2 years of sustained
effort to get FFMPEG to the point where it is fully leveraging GPU
Hardware Acceleration. So I would rather focus on the priorities first
than play around with some simple code.

However if you have a suggestion for a simple place to start that would be
helpful too.

Is there anything simple that would have a good overall purpose and lead
in the right direction?



--
Patrick Shirkey
Boost Hardware Ltd