Announcement

Collapse
No announcement yet.

Linux 3.12 Can Change AMD's Catalyst Performance

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

  • Linux 3.12 Can Change AMD's Catalyst Performance

    Phoronix: Linux 3.12 Can Change AMD's Catalyst Performance

    Due to a change found in the Linux 3.12 kernel, the AMD Radeon GPU performance is a lot faster with the open-source driver for certain configurations and the Nouveau driver performance is also faster depending upon the CPU and GPU. Now the latest in comparing the Linux 3.11 and 3.12 kernel performance, I can confirm some small performance changes with the binary 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
    So does this only affect distros such as Ubuntu that forced the ondemand governor instead of the newer intel_pstate one ?

    Comment


    • #3
      Those games does not even taxes gpu enough. It's more obvious fps hit the cpu bottleneck, which have got better with kernel 3.12(fps does not change much from 1280x1024->1920x1020).

      Comment


      • #4
        Ignoring the issue of which governor is in use, I actually have a question for Marek and Alex... R600g saw jumps of anywhere between 7% and like 50% with this governor change. Catalyst didn't. All other things being equal, including the card being used and only changing the driver... why is R600g so CPU-tempermental? I know that Gallium has overhead, and that it was a point of concern originally, but I thought the developers found ways to make the overhead be negligible?

        If its not the Gallium-architecture, then it seems like the open source driver is leveraging the CPU in ways Catalyst isn't, in which case the question becomes: Whats being shoved to the CPU under R600g that Catalyst is having the GPU do?
        All opinions are my own not those of my employer if you know who they are.

        Comment


        • #5
          I wonder if it is still the case that the 7xxx ASICs perform worse than their 6xxx counterparts using Catalyst?
          At least they did in most of your previous benchmarks IIRC.

          Comment


          • #6
            I wonder if this affects nvidia as well?

            Comment


            • #7
              Originally posted by Pawlerson View Post
              I wonder if this affects nvidia as well?
              Nvidia have this threaded OpenGL implementation, so I should help a bit.

              Comment


              • #8
                Originally posted by Phoronix/Michael
                Due to the CPUfreq changes within Linux 3.12, depending upon the GPU, CPU, governor, and OpenGL workload, the performance can be improved compared to Linux 3.11 and prior. For the quick Catalyst testing done this morning, the Intel Core i7 4770K Haswell system was again used and an AMD Radeon HD 7950 graphics card
                The effect of the governor change on SI is interesting in its own right, but perhaps you should have stayed consistent with the earlier investigations by, say, using the same HD 6950 .... and then afterwards (or simultaneously) added on a look at the SI hardware.

                Comment


                • #9
                  Originally posted by tuke81 View Post
                  Those games does not even taxes gpu enough. It's more obvious fps hit the cpu bottleneck, which have got better with kernel 3.12(fps does not change much from 1280x1024->1920x1020).
                  The CPU used here is pretty good at single thread, there is no CPU limitation. I've played these games with a slightly slower CPU (2600K) so I know what I'm talking about.

                  Comment


                  • #10
                    Now if we can get the performance/install bugs fixed in the Catalyst this will be a win for AMD

                    Comment

                    Working...
                    X