Moin,
On Sat, 17 Feb 2007 16:35:24 +0000
Dieter <[EMAIL PROTECTED]> wrote:
> > All hardware solutions
> > for decoding of video sold these days are made for set top boxes
>
> And that is a significant market for the TRV (sp?) chip.
Yes, would be, if we could offer a complete solution.
Well, actually, if we could offer a mpeg1/2/4, h.264
decoding solution that could output S-Video+HDMI (inclduding HDCP)
with everything openly documented, then we could break the
whole marked of settop box chipsets. But that's a big if
and i have no idea how we could provide HDCP while still
having open specs (the HDCP license requries that you
do not give out any specs of HDCP chips or anything
related to anyone who does not have a HDCP license)
> > > Undocumented hw features aren't very useful.
> > > Documented hw that doesn't do what you need isn't very useful.
> > > We need both the feature and the documentation.
> >
> > First we need something that works at all.
>
> We already have things that work "at all". But static images
> are not sufficient.
Believe me, we (ie people from MPlayer) are watching over
this already. You will have proper video support.
> > Athlon 64 3700+. MPEG2/4 1080p and 1080i content, no problem, with
> > both Xv and mga_vid drivers (using MPlayer, what else ;).
> > CPU usage varies between 30 and 70%
>
> Requiring 70% of a AMD64 3700+ is not even close to acceptable.
> And probably even higher with other material. Maybe in ten years.
> assuming CPUs get faster.
Well, yes, but 70% is hardly ever reached and by having a proper
design of the graphics card driver and video player application
this could be reduced to <50% (making it feasible to watch high res
content on a 1GHz machine)
> > slow AGP bus
>
> OGC will have an even slower bus.
Yes and no. OGD1 will be PCI, but equiping TRV10 with a
PCI-E to PCI bridge shouldn't be a problem. But what i actually
meant by that line was, that i am not using a state of the
art AGP bus, but AGP1 (yes, it's a 7y old G550) which does
not perform much faster than PCI-64 would.
Attila Kinali
--
Linux ist... wenn man einfache Dinge auch mit einer kryptischen
post-fix Sprache loesen kann
-- Daniel Hottinger
_______________________________________________
Open-graphics mailing list
[email protected]
http://lists.duskglow.com/mailman/listinfo/open-graphics
List service provided by Duskglow Consulting, LLC (www.duskglow.com)
