Announcement

Collapse
No announcement yet.

AMD Releases Open-Source UVD Video Support

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

  • Originally posted by Nille View Post
    catalyst. (10 Chars)
    Thanks!

    So why then you suggest
    Originally posted by Nille View Post
    No, to both questions.
    when the original question was about radeon(uvd), not catalyst(uvd)?
    Isnīt it possible to implement uvd+OpenCL(shaders) now for all codecs on radeon driver?

    Comment


    • It's simply not useful or even possible to do OpenCL/shaders for some codecs. They are not parallelizable.

      Comment


      • Originally posted by brosis View Post
        when the original question was about radeon(uvd), not catalyst(uvd)?
        Both are using the same Hardware an Firmware. 10bit is not possible with catalyst so i doubt that its possible with the radeon UVD code. The Hardware is not an FPGA with what you can map all possible codecs.

        Originally posted by brosis View Post
        Isnīt it possible to implement uvd+OpenCL(shaders) now for all codecs on radeon driver?
        Not maybe impossible but ineffective (a 50$ CPU can handle 1080p high10). Bridgman and Christian König has said something about this Topic in the past.

        Comment


        • @Nille, curaga:
          So its possible, but developers consider its efficiency to be inadequate.
          Which is pretty much blind-sighted, because CPU could do something else. The whole video acceleration is about "offloading". You canīt easily add another "50$" CPU into the system, but you CAN put another GPU into PCIe.

          OP should look into HSA, maybe.

          Comment


          • Originally posted by brosis View Post
            You canīt easily add another "50$" CPU into the system, but you CAN put another GPU into PCIe.
            And maybe its need an 200$ GPU only for Play an 1080p Stream and consumes 150W/h. Its easier and cheaper to replace the CPU or re-encode the video the asic in the gpu like quiksync or nvidias cuda encoder ( that has nothing todo with cuda its use only the name for marketing and use like everyone else an hardware chip for this )

            AMD is for encoding the worst case btw. its impossible for an normal user or developer to get the SDK for use the Hardware encoder in the radeon products.

            Comment


            • Originally posted by Nille View Post
              And maybe its need an 200$ GPU only for Play an 1080p Stream and consumes 150W/h. Its easier and cheaper to replace the CPU or re-encode the video the asic in the gpu like quiksync or nvidias cuda encoder ( that has nothing todo with cuda its use only the name for marketing and use like everyone else an hardware chip for this )

              AMD is for encoding the worst case btw. its impossible for an normal user or developer to get the SDK for use the Hardware encoder in the radeon products.
              Hardware encoder is only in recent HD7xxx series AFAIK.
              Used 57xx+ cost around 50$ now. If PM were right, they would use minimal wattage and there is already OpenCL work done.
              I am pretty sure one could easily implement video decode backend for OpenCL stack by replacing existant codecs codepaths and optimizing them.
              IMHO its all possible, but I donī t claim anyone is interested in that yet.

              Comment


              • That's completely wrong. Better look at:

                http://en.wikipedia.org/wiki/Unified_Video_Decoder

                It was never emulated via OpenCL. You can do that maybe for very simple codes but that would be never a good idea. xvba crashed very hard in the beginning with my hd 3450, then i got rid of it and got a hd 4550. the hd 5670 has basically a similar uvd part. Only hd 6000+ has got UVD 3.

                Comment


                • Originally posted by brosis View Post
                  Used 57xx+ cost around 50$ now.
                  And why you think its has enough compute capability do decode h264 complete with the ALUs?

                  Originally posted by brosis View Post
                  If PM were right, they would use minimal wattage and there is already OpenCL work done.
                  This cards need around 100W/h under Full Hardware usage. And node that UVD is an special ASIC for decoding and has nothing todo with an decoding an video on the ALUs.

                  Originally posted by brosis View Post
                  I am pretty sure one could easily implement video decode backend for OpenCL stack by replacing existant codecs codepaths and optimizing them.
                  IMHO its all possible, but I donī t claim anyone is interested in that yet.
                  Look at VP8 there was an Project that try to Decode VP8 with OpenCL. There is no result.
                  Last edited by Nille; 04-16-2013, 12:36 PM.

                  Comment


                  • Originally posted by brosis View Post
                    @Nille, curaga:
                    So its possible, but developers consider its efficiency to be inadequate.
                    Which is pretty much blind-sighted, because CPU could do something else. The whole video acceleration is about "offloading". You canīt easily add another "50$" CPU into the system, but you CAN put another GPU into PCIe.
                    If the codec is not parallelizable (= most of them), the GPU will use its full power and still be _slower_ than your CPU. Would you consider that a worthy offload?

                    Comment


                    • This 10-bit stuff was a dumb idea from the get-go.

                      Comment


                      • Originally posted by brosis View Post
                        You canīt easily add another "50$" CPU into the system, but you CAN put another GPU into PCIe.
                        Then buy an Broadcom CrystalHD card. Its cheaper and need much less W/h

                        http://www.amazon.com/Broadcom-Hardw...oadCom+Crystal

                        Comment


                        • Originally posted by johnc View Post
                          This 10-bit stuff was a dumb idea from the get-go.
                          10bit is an nice idea. the bump idea is still to use the 8bit profiles even in h265.

                          http://x264.nl/x264/10bit_01-ateme_p...422_10-bit.pdf
                          http://x264.nl/x264/10bit_02-ateme-w..._bandwidth.pdf
                          http://x264.nl/x264/10bit_03-422_10_...eo_quality.pdf

                          Comment


                          • Originally posted by Nille View Post
                            W/h
                            Please stop writing that . It’s “W”.

                            Comment


                            • Originally posted by stqn View Post
                              Please stop writing that . It’s “W”.
                              Lets be a little bit nicer and explain the units rather than just screaming "WRONG".

                              Energy is measured in a unit called JOULES (J).
                              A WATT (W) is the RATE of energy transfer (because energy can be neither created nor destroyed, you can do nothing besides transfer it).

                              1 W = 1 J/s <--- that is Joule per SECOND, clearly the rate of transfer of energy.
                              We can play with the time unit a bit to bring it to HOUR....
                              1 W = 60 J/m
                              1 W = 3600 J/h <--- now we can see what a watt is in terms of HOURS.

                              To picture this best, imagine a boat that has been sitting in the rain. It is partly full of water (Joules).
                              So you can think of JOULES in similar terms as "gallons of water".
                              Now you have a bucket, you need to empty that water out, so you hop in the boat and start bailing. You scoop up that water and pour it over the side of the boat.
                              How much time does it take you to bail out all of the water from the boat? That's WATTS. How fast you transfer that water out of the boat.

                              Moving those units around a little bit to give another familiar unit, we can see something like this;
                              1 W x 1 h = (3600 J / h) x 1h
                              1 W.h = 3600 J <--- 1 WATT-HOUR = 3600 JOULES.
                              When you pay for electricity delivered to your home, it is usually charged per KILOWATT-HOUR. kWh. This is just a convention to make the units look more familiar with respect to things like light bulbs and stove burners, which are described in watts, than it would be if measured in Joules.
                              1 kWh = 3,600,000 J <--- 1 kWh equals 3.6 million JOULES.

                              NOW, the inapplicable unit in question... W/h.....

                              W/h would be J/h/h <--- Joules per hour, per hour. This would be a RATE OF CHANGE OF RATE.
                              So for example, you have a rain barrel that drives a turbine. The deeper the water is in the rain barrel, the more pressure it outputs, and therefore the more energy you can transfer out of the turbine. Starting with an empty barrel, your transfer rate is 0 W. But it is raining, and filling your barrel at a rate that you measure to increase your energy transfer rate by 2 W/h.

                              After it has been raining for 1 h, the water is deep enough to transfer 0 W + (1 h x 2 W/h) = 2 W.
                              After it has been raining for 2 h, the water is deep enough to transfer 0 W + (2 h x 2 W/h) = 4 W, ****OR**** 2 W + (1 h x 2 W/h) = 4 W, depending on whether you are adding one incremental hour, or calculating from the start.
                              3 h --> 6 W,
                              4 h --> 8 W,
                              etc.
                              If you want to run a 60 watt light bulb off that, you're going to need a pretty tall barrel, and wait 30 hours for it to fill deep enough to generate enough pressure.
                              Last edited by droidhacker; 04-18-2013, 12:17 PM.

                              Comment


                              • Originally posted by droidhacker View Post
                                Lets be a little bit nicer and explain the units rather than just screaming "WRONG".

                                Energy is measured in a unit called JOULES (J).
                                A WATT (W) is the RATE of energy transfer (because energy can be neither created nor destroyed, you can do nothing besides transfer it).

                                1 W = 1 J/s <--- that is Joule per SECOND, clearly the rate of transfer of energy.
                                We can play with the time unit a bit to bring it to HOUR....
                                1 W = 60 J/m
                                1 W = 3600 J/h <--- now we can see what a watt is in terms of HOURS.

                                To picture this best, imagine a boat that has been sitting in the rain. It is partly full of water (Joules).
                                So you can think of JOULES in similar terms as "gallons of water".
                                Now you have a bucket, you need to empty that water out, so you hop in the boat and start bailing. You scoop up that water and pour it over the side of the boat.
                                How much time does it take you to bail out all of the water from the boat? That's WATTS. How fast you transfer that water out of the boat.

                                Moving those units around a little bit to give another familiar unit, we can see something like this;
                                1 W x 1 h = (3600 J / h) x 1h
                                1 W.h = 3600 J <--- 1 WATT-HOUR = 3600 JOULES.
                                When you pay for electricity delivered to your home, it is usually charged per KILOWATT-HOUR. kWh. This is just a convention to make the units look more familiar with respect to things like light bulbs and stove burners, which are described in watts, than it would be if measured in Joules.
                                1 kWh = 3,600,000 J <--- 1 kWh equals 3.6 million JOULES.

                                NOW, the inapplicable unit in question... W/h.....

                                W/h would be J/h/h <--- Joules per hour, per hour. This would be a RATE OF CHANGE OF RATE.
                                So for example, you have a rain barrel that drives a turbine. The deeper the water is in the rain barrel, the more pressure it outputs, and therefore the more energy you can transfer out of the turbine. Starting with an empty barrel, your transfer rate is 0 W. But it is raining, and filling your barrel at a rate that you measure to increase your energy transfer rate by 2 W/h.

                                After it has been raining for 1 h, the water is deep enough to transfer 0 W + (1 h x 2 W/h) = 2 W.
                                After it has been raining for 2 h, the water is deep enough to transfer 0 W + (2 h x 2 W/h) = 4 W, ****OR**** 2 W + (1 h x 2 W/h) = 4 W, depending on whether you are adding one incremental hour, or calculating from the start.
                                3 h --> 6 W,
                                4 h --> 8 W,
                                etc.
                                If you want to run a 60 watt light bulb off that, you're going to need a pretty tall barrel, and wait 30 hours for it to fill deep enough to generate enough pressure.
                                Just for my own edification, um How then would you describe voltage and current. wattage is equal to voltage x current if 1W = 1J/s then how does currrent and voltage work its way in? Is voltage x current just simply a different way of stating J/s?

                                Comment

                                Working...
                                X