Announcement

Collapse
No announcement yet.

AMD Opens Up XvBA! Their Catalyst Linux Video API

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

  • AMD Opens Up XvBA! Their Catalyst Linux Video API

    Phoronix: AMD Opens Up XvBA! Their Catalyst Linux Video API

    Back in 2008 we were the first to thoroughly talk about AMD's X-Video Bitstream Acceleration (XvBA) API found in their Catalyst Linux driver to expose their UVD2 video engine now under non-Windows operating systems. However, when the XvBA library was made available, it was next to useless since they hadn't published the documentation or any header files describing this video playback acceleration interface. A year later, in November of 2009, AMD and Splitted Desktop Systems released a VA-API front-end to XvBA so that VA-API multi-media applications could seamlessly use XvBA with the Catalyst driver...

    Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite

  • #2
    Nice move, but I can't see many video players supporting this... VA-API directly would be the best (VLC).

    Comment


    • #3
      See, this is why I didn't even look at Nvidia when I was looking for a new graphics card. (And, I'm encouraging others to buy AMD, too.)

      Although I must admit, I'm buying it used, unless someone wants to give me the additional $30 for a new card.

      Comment


      • #4
        Originally posted by d2kx View Post
        Nice move, but I can't see many video players supporting this... VA-API directly would be the best (VLC).
        Wouldn't VA-API just wrap XvBA? I think that's what they're doing for VDPAU on Nvidia.

        Comment


        • #5
          at least their still behind
          at least they are still behind
          It's nice to hear good news about AMD Linux development too once in a while., although it's just partially good news because DRM is ruining the fun again.

          Comment


          • #6
            At least now everyone can see for themselves if the api is as horrible as gbeauche has implied.

            Comment


            • #7
              This feels more like a move from AMD to address a key deficiency on Linux for their Fusion lines...Both the E-/C-series (Zacate/Ontario) and the upcoming A-series (Llano) are pretty much ideal for their respective markets; given their reasonably capable GPU-based IGP.

              If they can get their A-series working with accelerated video playback on Linux; I'm pretty much sold for the K10.5-based "Llano" (2011) and Bulldozer-based "Trinity" (2012).

              ...I'm not all that enthusiastic about buying a discrete video card when I don't have to. I've seen leaked slides that indicate that the A-series (Llano) IGP performs like a Radeon HD 5550 GPU. (It meets my computing needs.)

              Comment


              • #8
                Originally posted by curaga View Post
                At least now everyone can see for themselves if the api is as horrible as gbeauche has implied.
                And I will let people try to write a player from scratch with only the header file and doc file, without cheating and looking at demo code and realize the particular things to adjust...

                Comment


                • #9
                  Originally posted by gbeauche View Post
                  And I will let people try to write a player from scratch with only the header file and doc file, without cheating and looking at demo code and realize the particular things to adjust...
                  not fair ROTFL

                  does this mean that you are released from the NDA restrictions up to a point, and can now tell us were the bugs are in the catalyst driver and the header errors you mentioned that feels like a millennia ago ?

                  i noticed you posting on the ffmpeg mailing list BTW , perhaps you can send a post off there mention there that they have released this header and basic lib etc , perhaps someone there will slap something together using this info or even use it for the GSOC 2011 idea's and you can mentor them if that's allowed

                  Comment


                  • #10
                    Originally posted by aussiebear View Post
                    This feels more like a move from AMD to address a key deficiency on Linux for their Fusion lines...Both the E-/C-series (Zacate/Ontario) and the upcoming A-series (Llano) are pretty much ideal for their respective markets; given their reasonably capable GPU-based IGP.

                    If they can get their A-series working with accelerated video playback on Linux; I'm pretty much sold for the K10.5-based "Llano" (2011) and Bulldozer-based "Trinity" (2012).

                    ...I'm not all that enthusiastic about buying a discrete video card when I don't have to. I've seen leaked slides that indicate that the A-series (Llano) IGP performs like a Radeon HD 5550 GPU. (It meets my computing needs.)
                    it's more than just 'This feels like', it probably is a key part of 'the man with the plan' ,
                    however in this case it doesn't seem to be bridgman , rather Tim Writer ECSD Liaison Engineer at AMD, Owner at SaaS 44 Inc. whoever he is ?


                    "...
                    there's the fact that AMD made us do it. You see, back when the firm revealed the first details about the Bobcat CPU architecture used in Brazos, it predicted Bobcat would achieve an "estimated 90% of today's mainstream performance in less than half the silicon area." Now, such statements are famously slippery, but those words did conjure up some rather dreamy expectations inside of our heads. We likened Bobcat's potential performance to 90% of the Athlon X2 255's...."

                    and they clearly expect a lot from the UVD3 inside that latest AMD E-350 Zacate even though they are windows only testers, no one seemed to have the heart to tell techreport that UVD (other than gbeauche's valiant but limited success with a few UVD combinations OC) doesnt work to well for your basic high profile decode assistance in linux OSS...

                    Comment

                    Working...
                    X