Announcement

Collapse
No announcement yet.

AMD Radeon HD 6000 Gallium3D Attempts To Compete With Catalyst

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

  • #31
    Originally posted by crazycheese View Post
    Why don't you do this instead? No, seriously, I have a bit mess right now, so ... hard. But I have urt installed and I'm pretty sure its 125 fps capped - 60-125 fps are selectable, the engine uses this cap for sync and physics. Quake3 WAS able to remove the cap for local-only gameplay. The problem is.. I wanted to remove it myself some time ago, just to get info where gtx260 gets stuck - and I was not able to find the path.
    So yes, the cap *should be* there .. or I'm not getting something. The other thing that makes me suspect - is pretty same framerates on all cards... Don't ask how PTS managed to record 160 or 180 fps.. I don't know,.. yet.
    If my memory serves me, it's upped via com_maxfps.
    Michael Larabel
    http://www.michaellarabel.com/

    Comment


    • #32
      Originally posted by darkbasic View Post
      It does nothing? It goes from 70C to 45C with my HD5870...
      With the HD3870 it goes from 70C to 55C and even dynpm did work. Flawlessly.
      As I said, with my HD4870 and "default", it idles at 85C. With "low", 83C. What you get seems pretty much made up :-/

      Edit: AFAIK, dynpm is not supported with KMS.

      Comment


      • #33
        When people told me dynpm doesn't work I didn't trust them initially, it just depend on the card. And on the vendor. And maybe even on the bios version. Usually the low profile does work, sometimes not. Usually dynpm does not (!) work, sometimes it does. We surely need more testing and we should report bugs upstream.
        ## VGA ##
        AMD: X1950XTX, HD3870, HD5870
        Intel: GMA45, HD3000 (Core i5 2500K)

        Comment


        • #34
          Originally posted by RealNC View Post
          Edit: AFAIK, dynpm is not supported with KMS.
          Trust me, it does work (if you're lucky enough) with kms. I did test it very well, ssh-ing while playing openarena, while idling in the desktop, with and without compositing, etc...
          ## VGA ##
          AMD: X1950XTX, HD3870, HD5870
          Intel: GMA45, HD3000 (Core i5 2500K)

          Comment


          • #35
            Originally posted by Michael View Post
            If my memory serves me, it's upped via com_maxfps.
            Thanks, this is exactly what I was thinking about.
            com_maxfps will unblock, but ultimately only for viewing recorded demos.
            in online connection, it is ignored, because network sync is bound to it
            in offline connection (starting local, non-dedicated server), it is ignored too, because urt sees this as client-server connection, even in local scope.

            I have just tested gtx260sp216 at fullhd with com_maxfps set to 150 and then 180 - and in every single demo I have, on map of every complexity, it is always 131-132 fps, always. Vsync was off, fullhd 1920x1080, all settings to max.

            But, this can't be correct, since
            - I clearly remember fps going down sometimes when playing online (large maps, like sliema with 30 players)
            - setting 150 and 180 did not make difference, maps were of different complexity themself (dressing room tutorial with two players, ridyah with ~12 players in one screen, rivalry with aimbot cheater and 7 players...)...
            - (!!) I downclocked the 260gtx by setting it to max performance mode and then lowering the memory to 600 mhz from 999, and gpu to 400 from 585... No change in fps.
            It is definately capping framerate, even with 125fps limit off.
            I don't know to what? Very possibly it is CPU bottlenecking. I will try to downclock the athlon , lets see... That would actually mean correct thing, if true. In your test, Michael, all GPUs were roughtly having same performance in same driver group.

            ... this would ALSO mean, catalyst driver works MORE efficient in kernel-space than gallium, exactly by the difference between those two driver groups. If that was CPU bottleneck. Something gallium developers should know, if proven... And be able to patch, this giving a bit MORE performance to gallium (hopefully).

            Comment


            • #36
              Originally posted by RealNC View Post
              Hey, you're not much better, pal Changing the power profile with the radeon driver is also idiotic; it doesn't do much. At most it goes from 85C to 83C, which is just laughable.
              What are you talking about? The temperature drops like a rock when switching profiles.

              It is certainly possible that it doesn't work correctly on some hardware, in which case you should file a bug.

              I only called him names because he's been consistently and persistently spreading false information, even though he should know better because people have been correcting him on the same topics for years now. It's not the first time.

              And yes, dynpm needs improvement. No, low profile will not fry your card, or eat your children, or help bin Laden.

              Comment


              • #37
                Originally posted by crazycheese View Post
                Thanks, this is exactly what I was thinking about.
                com_maxfps will unblock, but ultimately only for viewing recorded demos.
                in online connection, it is ignored, because network sync is bound to it
                in offline connection (starting local, non-dedicated server), it is ignored too, because urt sees this as client-server connection, even in local scope.
                So what? Remember, this is graphics benchmark. We benchmark GPU capability, no matter demo,real gameplay, whatever. The point here is that Michael did things right.

                Comment


                • #38
                  Originally posted by pingufunkybeat View Post
                  What are you talking about? The temperature drops like a rock when switching profiles.

                  It is certainly possible that it doesn't work correctly on some hardware, in which case you should file a bug.

                  I only called him names because he's been consistently and persistently spreading false information, even though he should know better because people have been correcting him on the same topics for years now. It's not the first time.

                  And yes, dynpm needs improvement. No, low profile will not fry your card, or eat your children, or help bin Laden.
                  I think radeon profiles, are video BIOS configured profiles. If his card has broken BIOS, then PM will not work the way it should. IIRC, radeons, have some additional internal PM logic, which is not documented for FOSS, so only catalyst takes advantage of it.

                  Comment


                  • #39
                    I do agree, otherwise there is other explanation why does two 3870 from different vendors do behave differently...
                    ## VGA ##
                    AMD: X1950XTX, HD3870, HD5870
                    Intel: GMA45, HD3000 (Core i5 2500K)

                    Comment


                    • #40
                      Originally posted by Drago View Post
                      So what? Remember, this is graphics benchmark. We benchmark GPU capability, no matter demo,real gameplay, whatever. The point here is that Michael did things right.
                      Dude, what the fck did you chew on? The framerate is PINNED - there is NO graphics benchmarking there. And according to the stuff I found it is PINNED to CPU even in rare case when fps *should* be unblocked. Im gonna test this now, but I did not build-in the powersave governor in kernel for the reason I though I ain't gonna need it. You can retest this yourself:

                      Get x64 urban terror client.
                      Launch game
                      Go to options->system and set gfx to max and resolution to native
                      Go to options->misc and set Show FPS to on

                      1)Type ^ (or AltGr+~ on german keyboard) to open console
                      /com_maxfps 200
                      /demo [Tab] type there any one, client 4.2 has several of them present
                      Note the fps
                      2) Go to any control center you use, or use profile to force card to lower performance level
                      Launch urt and repeat step 1) - I get SAME fps there, it is not bound to 125, BUT it is topping at SAME value, which means this is either CPU BENCHMARK or game still enforces this FPS somehow(may be it depends on framerate of machine that RECORDED the original demo).

                      It needs to be tested - I can already say that that ain't GPU benchmark for sure. The framerates currently do not scale with GPU performance.
                      Last edited by crazycheese; 07-15-2011, 05:54 AM.

                      Comment


                      • #41
                        Originally posted by Drago View Post
                        I think radeon profiles, are video BIOS configured profiles. If his card has broken BIOS, then PM will not work the way it should. IIRC, radeons, have some additional internal PM logic, which is not documented for FOSS, so only catalyst takes advantage of it.
                        Then we are in desperate need of white, grey and black list on correct BIOSen. Please guys do this.

                        Comment


                        • #42
                          I have downclocked the processor to 2.2Ghz from 2.8
                          There is no change in fps - 131-132 fps
                          On any demo

                          I have further again downclocked nvidia driver (prefer maximum performance, GPU585->300, MEM999->600) and retested
                          There is no change - 131-132 fps

                          I guess I should recompile kernel, include the powersave demon and force it (2.8->0.8).
                          But but I don't think CPU is barrier anymore.

                          It seems that this game has some barrier that starts limiting max fps to curtain level - this is for sure and is independent of CPU or GPU clock rate.

                          Comment


                          • #43
                            Who knows maybe my next GFX card will be an AMD?

                            Comment


                            • #44
                              Originally posted by Yezu View Post
                              Who knows maybe my next GFX card will be an AMD?
                              Is it due to the Radeon vs Nouveau driver benchmarks?

                              Comment


                              • #45
                                Power profiles

                                Originally posted by Drago View Post
                                I think radeon profiles, are video BIOS configured profiles. If his card has broken BIOS, then PM will not work the way it should.
                                This is correct, but there are also bugs in the kernel code.

                                I have made an almost proper patch for debugging and manually fixing these problems. If there is any demand I might complete it. The problem is it already works good enough for me, so I have almost stopped development.

                                Basically the patch allows the user to see AND modify the power states manually. It can be used for underclocking, overclocking and for some other related tasks. It can be used like this:

                                Dump the table:
                                # cat /sys/class/drm/card0/device/power_table > power_table

                                Fix any problems:
                                # vim power_table

                                Commit changes (This could be added to boot scripts for more permanent fix):
                                # cat power_table > /sys/class/drm/card0/device/power_table

                                Currently on my box the file looks like this (there is perhaps too much information):
                                Code:
                                # power_state,clock_mode  dynpm_skip(boolean) engine(kHz) mem(kHz) vddc(mV) vddci(?)  [extra info]
                                0,0 1 100000 157000 900 0   [Boot]
                                0,1 0 100000 157000 900 0   [Boot]
                                0,2 0 100000 157000 900 0[*] [Boot]
                                1,0 1 100000 150000 900 0   [HIGH_SH_DPMS] [Performance] [Single display only]
                                1,1 0 450000 800000 950 0   [Performance] [Single display only]
                                1,2 0 550000 800000 1000 0  [HIGH_SH] [Performance] [Single display only]
                                2,0 1 550000 800000 1000 0
                                2,1 0 550000 800000 1000 0
                                2,2 0 550000 800000 1000 0
                                3,0 1 400000 800000 950 0   [HIGH_MH_DPMS] [Performance]
                                3,1 0 450000 800000 950 0   [Performance]
                                3,2 0 550000 800000 1000 0  [HIGH_MH] [Performance]
                                4,0 1 100000 150000 900 0   [LOW_SH] [LOW_SH_DPMS] [MID_SH_DPMS] [Battery] [Single display only]
                                4,1 0 100000 150000 900 0   [MID_SH] [Battery] [Single display only]
                                4,2 0 300000 300000 900 0   [Battery] [Single display only]
                                5,0 1 100000 150000 900 0   [LOW_MH] [LOW_MH_DPMS] [MID_MH_DPMS] [Battery]
                                5,1 0 100000 150000 900 0   [MID_MH] [Battery]
                                5,2 0 300000 300000 900 0   [Battery]
                                6,0 1 300000 300000 900 0   [Battery]
                                6,1 0 300000 300000 900 0   [Battery]
                                6,2 0 300000 300000 900 0   [Battery]
                                7,0 1 375000 400000 900 0   [Battery]
                                7,1 0 375000 400000 900 0   [Battery]
                                7,2 0 375000 400000 900 0   [Battery]
                                8,0 1 105000 150000 900 0
                                8,1 0 105000 150000 900 0
                                8,2 0 300000 300000 900 0
                                *_SH = Single Head modes
                                *_MH = Multi Head modes
                                *_DPMS = Modes used when all monitors are off in powersaving

                                I have only 1 monitor so in my case the important lines are the ones which are flagged with LOW_SH, MID_SH and HIGH_SH. As can be seen from the list the power_profiles "low" (4,0) and "mid" (4,1) are actually identical on my system.

                                Let's say I want to fix that. I want the GPU and memories to run at 300 MHz when in "mid" mode. This one command fixes that:
                                # echo "4,1 0 300000 300000 900 0" > /sys/class/drm/card0/device/power_table

                                Now that was only one simple example. I actually use this to fix the modes dynpm uses, but this post will get too long if I start explaining that.

                                Comment

                                Working...
                                X