Announcement

Collapse
No announcement yet.

AMD Releases Open-Source R600/700 3D Code

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

  • Originally posted by bridgman View Post
    Just to be clear, we're dealing with finite resources here so the question is not "would it be nice to have MPEG2 accel ?" (even I can answer that one ) it's "should the community work on MPEG2 accel instead of H.264/VC-1 accel ?", ie which should be worked on first ?



    The GPU programming is pretty much the same for the two families, so sequence of implementation would be the same for both. If you were combining the families, what would the sequence be ?



    Agreed, but that is another example of a workload which needs specialized hardware and is difficult to parallelize

    Oh,theres your problem , your thinking taking your workload and trying to parallelize it, you should be thinking micro-kernel and multitasking in the same vain as Carl Sassenrath of the original home mutitasking Amiga OS and rebol http://en.wikipedia.org/wiki/REBOL TCP/IP GUI scripting fame.

    actually, rebol GUI scripting would be a very good thing to use for any cross platform GUI HW assisted front end, and Carl's old AOS multitasking kernel Co-Processing ideas would work great for todays Gfx cards,plus a simple JIT backend/frontend perhaps, go ask him over on rebol.org
    Last edited by popper; 07 January 2009, 06:28 PM.

    Comment


    • Originally posted by popper View Post
      its been said that "the API is the least of the problems" and thats true to some degree, but Bridgeman has stated he beleaves theres enough data documentation out there right now.

      presumably that means theres enough information right now for someone here ? to take parts of the ATI/AMD API(s) and make an equivenent VDPAU ?
      The thread where I posted that was about XvMC. What I said was that there was enough information out there to implement XvMC. The information released is also sufficient to accelerate the back half of H.264/VC-1 decode (from inverse quantization onwards, with the rest being done on CPU. There's a good chance that frames using spatial prediction would need to have more of the work done on CPU.

      Originally posted by popper View Post
      The unfortunate thing is that they didn't also ship any development headers with the driver, with the result that the binary libraries were available, but there was no SDK or information available to media player developers to actually utilize the libraries. So XvBA currently remains a white elephant.
      No, it remains an unannounced feature which some clever people have started picking apart and talking about already.

      Originally posted by popper View Post
      remember also that its 4 months since the library(s) have been available, so alpha/beta test code at the very least must exist on the ATI devs machines to show off this new libray use, but still NO docs are available that i know of, to explain how you might use this library or its official API for hardware assist video decoding etc, WHY AS THAT?
      Because we haven't released it yet as far as I know.
      Last edited by bridgman; 07 January 2009, 06:38 PM.
      Test signature

      Comment


      • Originally posted by popper View Post
        people dont want a poll this late in the game, they want and NEED a real subset AVC decode and related libray ASAP, perhaps as a tempory stop gap measure until it all settles down later if needs be, PLUS development headers and DOCUMENTATION, and sample full working code showing anyone how to use it ASAP/TODAY.
        no, people want 3D perfornance first. As shown here:
        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


        so all that crying about video decoding is just a vocal minority - like always.

        Comment


        • hmm , perhaps , but non the less "a vocal minority" that puts in their picket every time theres a new use for their minority interest.

          and lets face it, theres far more people into video than any 3D linux use, sure linux 3D is cool, but it does nothing for your Co-Location ISP usage,HW assisted decoding/transcoding, and other related NON 3D processing etc will find its way into the data center for home/SOHO use to name but one example.

          Comment


          • > Just to be clear, we're dealing with finite resources here
            > so the question is not "would it be nice to have MPEG2 accel ?"
            > it's "should the community work on MPEG2 accel instead of
            > H.264/VC-1 accel ?".

            Mpeg2 accel is needed NOW. NTSC goes away in 6 weeks. ATSC is
            Mpeg2, and many people don't have CPUs fast enough to decode HD
            in real time, and some people probably don't have CPUs fast enough
            for SD.

            IIRC you wrote that you don't watch tv and perhaps you don't think it
            is important. But TV is more than just entertainment. It is an
            important part of the modern communications infrastructure. Many
            people depend on TV to get news about time critical problems, both
            natural (snow, ice, wind, earthquake, tsunami, volcano, tornado,
            hurricane, land slides, forest fires, floods, etc.) and man made
            (9-11, etc.). Is the tap water safe to drink? Am I supposed to
            try and get to school/work? Is transportation working, or have
            they closed the highways, and canceled the trains and planes?

            > If you were combining the families, what would the sequence be ?

            Mpeg2 accel is needed NOW.

            I've been asking about Mpeg2 accel for months (years in some forums).
            It is H.264/VC-1 accel that would be "nice to have", but that is
            entertainment, not critical news. Same for 3D. 3D is "nice to have",
            but is not needed to get critical news.

            Comment


            • OK, now *that* is a useful answer.

              BTW I mentioned that I don't watch TV only because someone posted that "obviously you watch TV so...". My personal habits should not be driving development decisions anyways

              Serious question though -- do you really believe that people who rely on their analog TV today for critical information are going to go out and buy a tuner card for their too-slow-to-decode-MPEG2 Linux PC and have that as their only option, ie they would not pick up a cheap set-top box to convert from ATSC to NTSC ? I can imagine maybe 10 people falling into that scenario, not tens of thousands.

              Anyone who already had an ATSC tuner card but could not decode the signals would have upgraded their PCs already or would not be relying on it. I have a tough time believing that they are sitting at home hoping that we will implement MPEG2 acceleration before Feb 11, or am I being too cynical here ?
              Last edited by bridgman; 07 January 2009, 07:22 PM.
              Test signature

              Comment


              • Originally posted by Dieter View Post
                > Just to be clear, we're dealing with finite resources here
                > so the question is not "would it be nice to have MPEG2 accel ?"
                > it's "should the community work on MPEG2 accel instead of
                > H.264/VC-1 accel ?".

                Mpeg2 accel is needed NOW. NTSC goes away in 6 weeks. ATSC is
                Mpeg2, and many people don't have CPUs fast enough to decode HD
                in real time, and some people probably don't have CPUs fast enough
                for SD.
                mpeg2 decoding can be easily done by a K6-2 400 - that is a 10 year old cpu....

                Comment


                • Nitpick old Mpeg2 SD and CIF type low bitrate encodes can be decoded by......K6-2 400 etc.

                  a K6-2 400 or 500MHz that i have cant decode PAL SD 50i thats broadcast at greater than 1.3Mbit/s on the DVB-T mpeg2 streams.

                  20Mbit HD 720P/1080i/P is totally another matter....
                  Last edited by popper; 07 January 2009, 07:48 PM.

                  Comment


                  • so what? are people watching blueray disks on their low level linux boxes? when you have the money to spent on a blueray player, you also have the money to buy a cpu with more than a 1ghz.

                    And tv?
                    'normal' digital tv is SD anyway - and handled easily by every cpu produced in the last 10 years.
                    And hd tv? Well - what about a decoder card? You need it anyway, so you can spent the few bucks more for one with hardware decoding, or on a better cpu.

                    In the mean time - 3d is something you can't go around with a better cpu. And 2d is far from perfect either. 2d is certainly more used than hd video watching.

                    So again, vocal minority.

                    Comment


                    • Originally posted by bridgman View Post
                      The thread where I posted that was about XvMC. What I said was that there was enough information out there to implement XvMC. The information released is also sufficient to accelerate the back half of H.264/VC-1 decode (from inverse quantization onwards, with the rest being done on CPU. There's a good chance that frames using spatial prediction would need to have more of the work done on CPU.



                      No, it remains an unannounced feature which some clever people have started picking apart and talking about already.



                      Because we haven't released it yet as far as I know.
                      the question then seems to become, after 4 months of releaseing the XvBA librays, why didnt/hasnt someone in ATI/AMD release (at least some of) the related docs to encurage developers to start using it/them!

                      Comment

                      Working...
                      X