Announcement

Collapse
No announcement yet.

Forthcoming VDR Release To Support VDPAU

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Forthcoming VDR Release To Support VDPAU

    Phoronix: Forthcoming VDR Release To Support VDPAU

    NVIDIA's VDPAU feature for providing greater GPU acceleration during video playback on Linux desktops has experienced great adoption since its November launch. This NVIDIA-spawned video API has already worked its way into open-source projects like MPlayer / FFmpeg, MythTV, Xine, and XBMC...

    http://www.phoronix.com/vr.php?view=NzEzMA

  • #2
    ATI really should be regretting not having released their implementation first.

    It's probably too late now. They'll have to copy nvidia and use VDPAU, cause devs aren't gonna wanna implement two dirfferent APIs

    Comment


    • #3
      I am quite sure that someone has misunderstood something here... VDR itself will definitely not support VDPAU, since it is based on functionality of so called full-featured DVB cards, which contain an integrated mpeg2 decoder. On the other hand there are already plugins for VDR that use xine-lib for decoding with cards that don't have integrated decoders (so called budget cards). Those plugins also already support the vdpau fork of xine-lib (xine-vdpau) and therefore it has been possible to use VDR with VDPAU for some time now. The news, I believe, are that someone is releasing a distro that is dedicated to VDR use and has support for xine-vdpau with the aforementioned plugins by default.

      Comment


      • #4
        I'de like to remember everyone that INTEL hates NVIDIA. They are actually going under court because INTEL says their "old" contract isn't valid for newer processors. INTEL wants to kill NVIDIA because NVIDIA wants to show the world everything can be done through GPU and INTEL of course says the opposite.

        Comment


        • #5
          Originally posted by bulletxt View Post
          I'de like to remember everyone that INTEL hates NVIDIA. They are actually going under court because INTEL says their "old" contract isn't valid for newer processors. INTEL wants to kill NVIDIA because NVIDIA wants to show the world everything can be done through GPU and INTEL of course says the opposite.
          Who isn't intel in court with? Intel will probably loose this case too. Intel has always taken a "how dare you compete against us" and it's deeply embedded in their company culture. There have been many examples of this such as their battle with Via, their anti-trust case with AMD, and being pushed into giving Nvidia a chipset license so the XBoX wouldn't use a AMD proc instead of a intel proc. No doubt having lost one of their biggest OEM's switching to nvidia chipsets has pissed them off (Apple).

          Just one personal example for seeing this in action was back around when the athlons just got released. AMD and Intel were both present at this symposium. Both had their flagship CPU's present running in pretty much identical configs. Both claimed they were the fastest. When a bunch of us asked to have a head to head to prove each others claim AMD said "sure" and intel declined. We further prodded to get this to happen and intel then promptly and I quote "This is BS and nothing but a lynch mob. We are intel and don't have to prove anything." . And with that they packed up their stuff and left the week long symposium on the first day.
          Last edited by deanjo; 03-09-2009, 02:43 PM.

          Comment


          • #6
            I wonder why is everyone jumping onto the bandwagon so fast. Is it because of the features it provides, or good documentation?

            Comment


            • #7
              Originally posted by Vadi View Post
              I wonder why is everyone jumping onto the bandwagon so fast. Is it because of the features it provides, or good documentation?
              I imagine documentation and sample code must have been good and clear or at least simple to understand. It is definitely something interesting at hand and I believe coding and hacking is fun . Moreover, they had a working implementation and also example code of mplayer-vdpau. Meanwhile, if CUDA was a little easier and more useful, I would have tried my evil hands on it too .

              Comment


              • #8
                Originally posted by hdas View Post
                I imagine documentation and sample code must have been good and clear or at least simple to understand. It is definitely something interesting at hand and I believe coding and hacking is fun . Moreover, they had a working implementation and also example code of mplayer-vdpau. Meanwhile, if CUDA was a little easier and more useful, I would have tried my evil hands on it too .
                Let's not forget one huge reason, it's a feature that has been wanted in linux for years and largely ignored until vdpau came out.

                Comment


                • #9
                  I think the question was why VDPAU over, say, VA-API which had been around longer and met the same needs. I agree that the combination of a spec *and* a readily available implementation would be the key.

                  Comment


                  • #10
                    http://www.linuxtv.org/pipermail/vdr...ch/019812.html

                    comments from VDR's author about that news

                    Comment

                    Working...
                    X