Forgot your password?
typodupeerror
AMD Bug Graphics Hardware

AMD Tweaking Radeon Drivers To Reduce Frame Latency Spikes 105

Posted by Soulskill
from the stomping-bugs-with-extreme-prejudice dept.
crookedvulture writes "Slashdot has previously covered The Tech Report's exposure of frame latency issues with recent AMD graphics processors. Both desktop and notebook Radeons exhibit frame latency spikes that interrupt the smoothness of in-game animation but don't show up in the FPS averages typically used to benchmark performance. AMD has been looking into the problem and may have discovered the culprit. The Graphics Core Next architecture underpinning recent Radeons is quite different from previous designs, and AMD has been rewriting the memory management portion of its driver to properly take advantage. This new code improves frame latencies, according to AMD's David Baumann, and the firm has accelerated the process of rolling it into the official Catalyst drivers available to end users. Radeon owners can take some comfort in the fact that a driver update may soon alleviate the frame latency problems associated with AMD's latest GPUs. However, they might also be disappointed that it's taken AMD this long to optimize its drivers for the now year-old GCN architecture."
This discussion has been archived. No new comments can be posted.

AMD Tweaking Radeon Drivers To Reduce Frame Latency Spikes

Comments Filter:
  • Give them credit (Score:5, Insightful)

    by Anonymous Coward on Wednesday January 02, 2013 @06:14PM (#42455729)

    taken AMD this long to optimize its drivers for the now year-old GCN architecture.

    Give them some credit... they've acknowledged the problem and this isn't a simple tweaking/bugfix, this is a complete redesign and rewrite of the entire driver architecture.

    • by hattig (47930) on Wednesday January 02, 2013 @06:38PM (#42455977) Journal

      Well, it's in the memory manager portion of the driver. Memory management isn't easy at the best of times, and when you're dealing with a GPU that has thousands of cores, and each of those cores has its own local memory, and shared memory with a local cluster group, and then there are software controllable caches further up the hierarchy, I can see how writing this code could be fraught with difficulty.

      And as many of us here have worked in professional software environments, I'm sure we can all see how something that was pretty hard to pin down like these latency spikes might not have been a top priority for development, even if they were aware of it at all - after all the FPS figures were great. You'd end up with a driver kernel that had some magic that nobody would want to touch, and most of the work would be game specific optimisations and higher level optimisations. A year sounds about right really.

      • One would think an automated test that looked at timestamped frame production would show the jumpy nature of it.

      • by hairyfeet (841228)

        This is true and on top of that they have completely switched designs, previous designs were VLIW and the new design is vector based GCN, so naturally when doing something as massive as a complete change to the architecture of the chip itself its gonna take some time to get all the bugs worked out.

        Now that said this right here is why I always stay a couple of gens behind when it comes to GPUs, because frankly the higher end GPUs (the x8xx and x9xx chips especially) have been so badass that by staying a coup

    • Re:Give them credit (Score:5, Informative)

      by Anonymous Coward on Wednesday January 02, 2013 @06:50PM (#42456119)

      More than that, this frame latency issue is new to the recent drivers - possibly even restricted to the 12.11 beta drivers, which one might expect to have the occasional issue. Further, other review sites that investigate the smoothness of gameplay, even if not by quite the same method (ex: Hard|OCP [hardocp.com]), have not found the issues that Tech Report has, which leads me to suspect that it might not be a universal issue.

      I think that Tech Report has been very irresponsible in their handling of this issue. They've been quick to condemn, ignoring that the driver is _not_ a final version, and apparently unwilling to investigate the disparity between their older results with the same cards/older drivers (which had no such frame latency issues) and their new results. Frankly, I'm not willing to give them any regard as a review site any longer - their frame latency methodology certainly has merit, but they seem to be utterly incompetent in actually analyzing or investigating the results.

      • by nanoflower (1077145) on Wednesday January 02, 2013 @09:00PM (#42457009)
        Only one problem with what you suggest is that it is based on bad information. TR made the effort to look at different versions of the drivers and they've tested it on Win7 and Win8. Also only a couple of other sites have done the same level of testing frame rates that TR has been doing and they've found the same issues. Then you add in that AMD has looked into the issue and acknowledged there is a real issue that they need to address. So you are doing a disservice to Tech Report by misstating the situation and ignoring the other sites that have agreed with their findings.
        • Re: (Score:2, Insightful)

          by Anonymous Coward

          I don't doubt their findings, merely that the latency spikes are a universal problem, and that they weren't recently introduced in a _beta_ driver (or at the least, a newer one - we know the problem didn't exist back in earlier versions).

          I wouldn't be surprised to find out that this is a result of the chipset/disk controller driver (which they updated in the interval between their older stutter-free results and their new stutter-heavy ones) interacting poorly with the newer graphics drivers, and possibly th

      • Also, having just looked at the videos [techreport.com], I actually find the AMD version to be more visually pleasing - yes, there's a big "jump" at a few spots in the video, but it's in no way regular, while the nVidia jumps (albeit a smaller distance) periodically - roughly once a second. It's like a rhythmic pulsing... really weird.

        And I can't say I've seen this issue on my HD7750 (awesome little card - no PEG connector needed, doesn't break 50C running Starcraft 2 on high, low power consumption)... although that might b

    • by smash (1351)
      Given that they wrote both the software and designed the hardware, this smacks of insufficient testing to me. It's not like they're attempting to make somebody else's hardware work.
    • by Cinder6 (894572)

      I have 2x R7950s in my machine, and have been reasonably satisfied with them. The microstutter, as people call it, is pretty annoying at times, and had I known it was a particular issue with AMD cards, I would have gone with nVidia. The second issue is damned capacitor squeal, another thing I didn't know would be an issue. I've always run EVGA cards before this rig, and never encountered it (it's a buzz that happens in some game menus and, oddly enough, when there's a lot of white on the screen--meaning

      • The second issue is damned capacitor squeal

        AFAIK (CMIIW) capacitors cannot make noise, but rather inductors, coils...

  • Isn't it also an AMD CPU issue? Or is the problem simply much more pronounced with this particular line of Radeon GPUs?

    Or was the CPU problem long corrected?

    This article from TechReport in August convinced me not to go AMD for my next gaming PC:

    http://techreport.com/review/23246/inside-the-second-gaming-performance-with-today-cpus [techreport.com]

    • They make great dedicated video cards.

      I have noticed this issue on my ATI 5750 dedicated card last year.

      The CPUs are so-so unless you are looking for a cheap multicore system with virtual machines. Intels are better for single tasking and also multitasking if you have a large budget. Unfortunately, this is no market for cheap multicore heavy multitasking users except for a few geeks.

      Thew new beta drivers work for me and it is a big improvement.

      • by Pinhedd (1661735)

        The 5750 is still on the old VLIW5 architecture, this optimization only affects cards on the GCN architecture

        • Well I noticed it is smooth now in games with the beta last month when I have good FPS so AMD is doing something. I can see it with my own eyes.

          The 5xxx are the last cards supported with the latest drivers so it appears they are backporting the fixes to my card but no earlier ones as they are stuck on 12.3 I believe.

          • by Cinder6 (894572)

            I'm using the current betas and haven't noticed any change in microstutter, so far as I can tell (2x R7950s).

            • by Tynin (634655)

              I'm using the current betas and haven't noticed any change in microstutter, so far as I can tell (2x R7950s).

              I think I'm able to duplicate the micro-stuttering in counter strike global offensive. When I'm running BOINC (lowest priority, CPU only) occationally the game will seem to have a very brief, but noticable, halt... and go back to running smooth. Turned off BOINC and it went back to being only smooth. Might be an issue with the process scheduler... or perhaps like the articles suggest, the video card drivers. Either way, I'm all for better drivers :)

      • They're also "good enough" for pretty much any game on the market and they're still cheap... so there is a market for them.

  • I'd be surprised if any new drivers show up for any video cards out in the field until a couple years of sales of new cards (same old cards sporting new drivers).
    After all, they have no incentive to keep you using your old card if they can convince you to buy a newer one. AMD has a pretty long record of abandonware when it comes to video cards.

    • My el cheapo laptop from 2007 has WDDM 1.1 support for Windows 7 accelerated. I was impressed with that as this box was designed for Vista not 7. Also the driver version is from last March so ATI supported it for well over 5 years.

    • by antdude (79039)

      Agreed. I had to use v9.4 Catalyst driver for my old ATI Radeon 4870 video card (512 MB; PCIe) because of my very old, updated Windows XP Pro. SP3's clock/time DVI slowing down and rare random hard lock ups when playing videos. :(

      • Is an upgrade to Win7 possible?
        • by antdude (79039)

          Yes, but why when XP Pro. SP3 and old softwares work fine? I don't play computer games anymore. I will upgrade when I am forced (e.g, dropped support).

    • by Nemyst (1383049)

      I'm not sure what's implied here but AMD's got Windows 8 drivers for their HD2000 series, which was released in 2006. That's, I'd say, a fairly good legacy hardware support timespan.

    • by Cinder6 (894572)

      On the other hand (and I say this as an AMD user), if were to stop supporting my cards right now (R7950s), I sure as hell wouldn't buy from them again. AMD has a vested interest in fixing bugs and improving performance in cards already in customers' hands. Goodwill is important to a company, and I'm not aware of any consumer group that's as fickle or easily offended as gamers.

  • I thought AMD's last graphics chipset with the name "GCN" was its Flipper GPU in the Nintendo GameCube video game console from 2001, which was for some reason abbreviated GCN.
    • by Mr Z (6791)
      Game Cube Network, perhaps? At least that's what GCN means in the context of Mario Kart.
    • by dstyle5 (702493)
      GCN stands for "GameCube Nintendo" and as far as I recall did not describe any particular hardware component of the console.
  • I have the latest beta cataylst drivers from December 3rd and SWTOR is now fluid when I turn. Also video seems better too. Also worth mentioning is this and the last stable released fix the massive security hole by disabling protected mode in Vista/7 with ASLR.

    I highly recommend ATI users upgrade their drivers as I found the beta more stable than the stable one.

    • by Cinder6 (894572)

      I may have to look more carefully, then--I've not noticed any improvement to Skyrim since installing the betas last week, but that was also the same time I started heavily experimenting with graphics improvement mods.

  • by PhrostyMcByte (589271) <phrosty@gmail.com> on Wednesday January 02, 2013 @06:33PM (#42455915) Homepage

    This only went on so long because tech sites use such poor, useless benchmarking methods. Minimum/Average/Maximum FPS, or often just Average/Maximum FPS, are worthless!

    A game, or a video card, can average 100fps, but still have that one frame every second that performs some extra I/O and takes 3x longer than usual causing an annoying stutter effect.

    A good first step would be to use frame latency percentiles.. i.e. 90% of frames are at least 60 FPS, 95% of frames are at least 50 FPS, 99% of frames are at least 40 FPS.

    The next step is to measure spikes themselves -- low framerate sucks, but not nearly as much as a stuttering framerate. A sudden spike from constant 10ms/frame to 50ms/frame and back should be counted as far more detrimental than a smooth transition from constant 10ms to constant 25ms.

  • The overwhelming bulk of GPGPU users are still on two and three year old cards. When they move up they'll either buy the discounted 7000 series once the 8000 series arrives or not. Either way, the GCN issues will be worked out and they won't have ever experienced this issue.
  • Ordered Nvidia (Score:2, Informative)

    by watermark (913726)

    I had a newish card, AMD 4xxx HD, but they force me to use the fgrlx-legacy driver. The fgrlx-legacy drivers don't work with the newest xorg, and the ones that work with the older xorg are missing features essential for steam. AMD scaled down the Linux team recently. I just bought an Nvidia.

  • I've had a 7970 since early 2012 and have never had this issue on any of the resolutions I commonly use (1920x1080 and 2560x768 eyefinity). Obviously other people have had these issues, but I've personally never experienced them, and I'm sure I'm not the only Radeon owner who was spared from this bug.
    • by tlhIngan (30335)

      I've had a 7970 since early 2012 and have never had this issue on any of the resolutions I commonly use (1920x1080 and 2560x768 eyefinity). Obviously other people have had these issues, but I've personally never experienced them, and I'm sure I'm not the only Radeon owner who was spared from this bug.

      Well, to detect it means you need to really run a bunch of video cards together on the games you play. It was discovered only because a video card was putting out great numbers, but gameplay seemed "less smooth

      • by Cinder6 (894572)

        I have two 7950s and notice it. The game will just randomly slow down for a few frames, then go back to running smooth as silk. It's annoying and distracting, but it's also one of those things that isn't nearly so common or noticeable when you're just running one card. Had I known of the issue when I bought them, I would have gone with nVidia.

        • by jakobX (132504)

          This is a different problem and is common to every crossfire/SLI setup. If you want to fix it you have to play with tools like radeon pro.

  • It's too late. I've been a hardcore AMD/ATI guy for over 10 years and I'm done. They could have fixed this crap BEFORE they were nearly bankrupt, now it's a bit late. It's one thing when AMD was getting beat up by intel and ATI was getting beat up by NVIDIA. But drivers from both companies were just shit. I was buying $600 video cards and getting crappy performance for years... Their duel monitor implementation is ridiculous. AMDs CPU setup on older systems was borderline ridiculous... they finally fixed th

    • by Pubstar (2525396)
      Care to elaborate on the dual monitor implementation? I have had no problems running a dual or tipple monitor setup since my HD4460. Hell, I got a 8150FX running in quad core mode and it cleans up in benchmarks against my friends i5 2600k (i do have a slightly higher OC than him).
    • by epine (68316)

      Unfortunately, IT'S NOT PRESCOTT!!!(TM) was the high point in the AMD story arc. They bought themselves some additional time and good will being less pederastic toward binary blobs than the Green Goblin.

      Despite this, I've also grown weary of the difference between the vision AMD proclaims and the reality in the channel six months later.

  • I always had to terminate that process to gain some input responsiveness back, otherwise the only workaround is to have a multi-core system.

If builders built buildings the way programmers wrote programs, then the first woodpecker to come along would destroy civilization.

Working...