Driver software to be tweaked to reduce Radeon frame latencies in series of updates

To our surprise, we recently found that the GeForce GTX 660 Ti generally outperforms the Radeon HD 7950 in our latency-focused tests in many of the latest games, despite the fact that the Radeon is based on decidedly beefier hardware. Although the Radeon cranked out conventionally respectable FPS averages, it often produced a number of long-latency frames interspersed throughout our testing sessions. Follow-up testing confirmed the problem isn’t confined to Windows 8, and we even posted a slow-motion video illustrating the issue. We concluded that AMD has work to do in optimizing its drivers for the latest games.

Earlier today, in my blog post, I noted that AMD’s David Baumann had posted in a thread at Beyond3D, stating that a host of different software-related issues are potentially responsible for the Radeon’s latency issues. He claimed the slowdowns in Borderlands 2 are a buffer-sizing issue that could be addressed via a Catalyst Application Profile (CAP) update.

After seeing my blog post, Baumann contacted us to provide some additional insight into the situation, including word of a series of driver updates in the works intended to smooth out frame latencies. He writes:

The comment that you quote was just one update that highlights that some things can be tweaked fairly easily (although since coming back today I learn that it is not quite as easy as the BL2 fix does actually need to be implemented in the driver so we will have to QA a new build rather than releasing a CAP). Over the early part of the year you’ll see a few driver updates help this across a variety of games.

We’re pleased to see that AMD will be addressing these issues soon, even if Borderlands 2 can’t be patched via a CAP update.

The most intriguing revelation in Baumann’s correspondence, though, concerns one specific technical contributor to the frame latency problems on HD 7000-series Radeons based on the GCN architecture: less-than-optimal memory management in software.

Additionally, when we switched from the old VLIW architecture to the GCN core there was a significant updates to all parts of the driver was needed – although not really spoken about the entire memory management on GCN is different to prior GPU’s and the initial software management for that was primarily driven by schedule and in the meantime we’ve been rewriting it again and we have discovered that the new version has also improved frame latency in a number of cases so we are accelerating the QA and implementation of that.

So a specific portion of AMD’s driver code needs some additional attention in order to perform optimally on the year-old GCN architecture—and AMD has accelerated an overhaul of it after discovering that the new revision can alleviate frame latency issues. Wow.

Although we’re not happy about the situation facing current Radeon owners, we’re gratified to see that AMD has taken notice of the problems and is working to resolve them. We’re also thrilled by the possibility that our latency-focused game testing may have helped nudge one of the major GPU makers into making changes that could result in improved gameplay fluidity for PC gamers going forward. Stay tuned to TR for additional updates on this situation as they become available.

Comments closed
    • Fighterpilot
    • 7 years ago

    Recommending the GTX660ti over the HD7950 because of some minor benchmark latency spikes is just wrong.Only a tiny percentage of owners have even noticed it and it hasn’t affected gameplay. or image quality.
    TR is appropriate to mention it but to change its card recommendation over it is astonishing.

    Here’s what [H] had to say…

    “Middle pricing band – This pricing band was far less competitive as the Radeon HD 7950 with Boost simply demolished the GTX 660 Ti across the board with regards to raw frame rates and overall game play experience across our suite of testing”

    “The most interesting stand out from this evaluation has been the NVIDIA GTX 660 Ti, which is priced significantly higher than the Radeon HD 7870 while turning in similar performance numbers. The GTX 660 Ti has lost its value at this point in time, lending way to the Radeon HD 7950 which has benefited from falling prices and new drivers.”

      • oldDummy
      • 7 years ago

      Hello FP.

      While you have a point regarding the severity of poor drivers; the main problem is less obvious.
      If AMD had knowledge of these limitations without addressing them, it shows disregard for their customers and lack of integrity on a corporate level. Allocating resources to fix this problem should have been job one. Lack of resources may be a factor…..
      This brings to mind the tail end of the 3dfx implosion.

      • Willzzz
      • 7 years ago

      Thing is over the games tested the 660 Ti won on FPS as well, mainly due to the 7950 choking quite badly on AC3. Clearly readers are free to draw their own conclusions based on the games they want play.

      There weren’t any convincing losses for the 660 Ti over at [H], like with most purchases it depends upon price at the end of the day.

    • michael_d
    • 7 years ago

    My 7970 is a good performer as it is. I have set all 3D options to max in Catalyst and it runs quite well at 2560×1600 even in Metro 2033 and STALKER CoP with all in-game settings set to max.

    I can see why there is delay with Catalyst update now. The last official driver was released in October 2012.

    The Catalyst 12.11 Beta has slightly better performance. I assume it has the aforementioned optimizations but for me the problem is that STALKER crashes during loading due to Access Violation Exception. There are probably other issues as well that is why it is delayed.

    • a_non_moose
    • 7 years ago

    I have to admit that this did not seem like a big deal at all until I started seeing “issues” over time.

    I admit, I update grfx drivers sparingly (3 to 6 revs) barring any major issues with whatever I’m playing.

    Long stint with Arkham CIty, particularly the combat maps which on my XFX 6870 did very well in
    the 12.1 Cat era. Smooth because of the simple maps even with lots of thugs/action.

    12.6 I noticed “hits” before any cues were given…thought it was my imagination until I had more than
    one “could not miss it if I tried”.

    12.12…gawd, it’s even worse now and extends to gadget use. Used the GEP and the “arc” lasted
    a lot longer and floated in mid air when it was done.

    Granted a Q9550 @ 3.2Ghz ain’t top of the line now-a-days, but it is no slouch either and the only
    thing that has really changed (aside from a few more games like ACIII…and specific GAME issues)
    is the drivers.

    But the article did make me wonder, and it seems that the latency does come into play in small
    and subtle way and the problem got worse much faster for previous gen and older cards and
    the current crop.

    I’ve owned a 9800, x800, 3850, 4850 and the 6870 and next system is going to have a 660 simple
    because of a future hackintosh build (and 10.7 + 4850 =’ed much pain, but works despite being
    the only version NOT supported…WTH was that about? Neither here nor there, but still)

    • jonjonjon
    • 7 years ago

    isn’t this the story of amd drivers. sorry we released garbage for the past year. it should be fixed in the future maybe. how lame. the 7950 should smoke the 660ti but amd is such a mess they cant do anything right. i just told my friend to get a 660ti i really wanted to tell him to get a 7950 but i don’t believe/trust amd. of course they are going to do PR damage control. your articles made them look pretty bad. im curious how many people still work in the amd driver dept compared to nvidia. they probably laid half of them off.

      • clone
      • 7 years ago

      garbage would be drivers that don’t work… AMD’s drivers work very well but for the moment Nvidia’s drivers comparatively speaking run smoother frames… that isn’t garbage.

      we don’t live in a black and white world, why would anyone presume to make an issue so basic, so definitive….. absurd.

      I’ve found this whole situation to be overblown by a wide margin, TR is to be thanked for finding a way to show performance differences between AMD and Nvidia gfx cards but just because their are differences does not mean that AMD’s cards are “garbage”, more to the point AMD’s HD 7950 will most likely prove to be notably faster and longer lived than Nvidia’s 6xx series.

      the fact that Tech Report had to accelerate time lapse recordings in order to find a means to show the flaws that supposedly quantify AMD’s 7950 as “garbage” is a testament to how minor the problem is.

        • jonjonjon
        • 7 years ago

        you are really going to make excuses for amd’s drivers? i never said the card is garbage just the drivers and support. maybe try reading the first post without your amd biased glasses on. its pretty clear to me that you get better driver support from nvidia. i’m also not an nvidia or amd fan. i buy whatever card has a better price/performance at the time. why would anyone buy a lesser card just because its made by a certain company? but that’s another topic.

        its pretty black and white to me. a year later and their drivers still suck. they basically admitted their memory management is a mess. they release a new architecture and instead of rewriting the memory management system to take advantage of it they slapped some half assed patched version together using the previous gen code. this is a year later stop with the no excuses. if you don’t have enough people to support your products then maybe you hire some more. ohh i forgot this is amd we are talking about. isn’t kepler a new architecture that’s been out for even less time then gcn? excuses excuses. they are also “accelerating the QA” sound like a good idea what could possibly go wrong? it would be funny if it wasn’t true.

        your going to try to tell me no one at amd looks at frame latency? what else would you look at while debugging a driver? fps? doubt it. you really expect me to believe that? they knew there was a problem but if no one else noticed then why spend time and money to fix it? had this not been caught by TR amd would have went on their merry way posting fps benchmarks and it wouldn’t have been fixed. with amd’s track record i don’t even believe them when they say it will be fixed. when the 8000 series is released? if this was the first time i would give them a pass but they have a history of this kind of stuff. maybe you should look at the amd enduro drivers and all the poor people who bought laptops with amd gpu’s.

        [url<]http://www.anandtech.com/show/6480/amd-enduro-beta-take-three[/url<]

          • clone
          • 7 years ago

          [quote<]you are really going to make excuses for amd's drivers? i never said the card is garbage just the drivers and support.[quote<]its pretty black and white to me. a year later and their drivers still suck. they basically admitted their memory management is a mess.[quote<]your going to try to tell me no one at amd looks at frame latency?[/quote<][/quote<][/quote<]1: excuse only applies if you are arguing that Nvidia has not one single driver issue which is not the case, both Nvida and AMD have driver issues which doesn't make either companies drivers garbage, they are actually very good, but they do have some issues...... issues that are being overblown. 2: that's brutally obvious given your inability to place the issue in relative terms, the drivers have issues just like Nvidia's drivers have issues, that AMD's current issues may be more focused in memory management.....so?..... ummm, great, they found the source and will fix it? their is no such thing as a perfect driver. 3: said no such thing, quote me on that. P.S. to be clear if you want the best performance today for the identical price because the differences are less than minimal then sure go ahead and buy Nvidia's 660..... but knowing where the potential lies for the same price I wouldn't hesitate to buy an HD 7950 and I'm saying this as the current owner of an Nvidia GTX 460 and someone who's owned/used 25+ Nvidia cards over the past decade and a half.

    • ronch
    • 7 years ago

    This is one of the reasons why TR is my foremost go-to tech site. I hope you guys don’t drop the ball like some other tech sites out there that used to fly high but lost focus later on. It’s all about credibility/honesty, humility, and witty and interesting writing styles.

    • ronch
    • 7 years ago

    So does this mean AMD should reduce Radeon prices such as the HD7950 to the level of their corresponding same-latency GeForce (say, 660 Ti) until they’re able to completely resolve these frame latency issues? I mean, knowing this, why would someone pay more for a high end Radeon that doesn’t offer a tangibly better gaming experience than a cheaper GeForce?

      • Voldenuit
      • 7 years ago

      [quote<]So does this mean AMD should reduce Radeon prices such as the HD7950 to the level of their corresponding same-latency GeForce (say, 660 Ti) until they're able to completely resolve these frame latency issues?[/quote<] That would be a short term solution at best. nvidia could easily reduce prices on Kepler (since it's already a smaller die, and the 660Ti is already a twice-harvested chip) and launch AMD into a bloody price war. While that might be attractive to consumers in the short term, it's not an attractive prospect for AMD. Their best bet is to fix the latency issue in drivers as quickly as possible. In the meantime, they might face rougher sales than they'd like, but the onus is on AMD to fix the extant problem with their cards for the benefit of their current and future users alike.

    • indeego
    • 7 years ago

    Color me Very Very surprised! ATI Drivers aren’t working right?

      • Krogoth
      • 7 years ago

      Nvidia drivers aren’t exactly saints either. They have their fair share of stupid, stupid issues over the years that the fanboys make into “OMFG the WORLD IS ENDING!” type problems.

      The problem is that modern GPUs are very complex pieces of hardware that depend heavily on the software to work effectively. The entire 7950/660Ti latency “fanboy shitstorm” simply highlight this. On paper, the 7950 should be pulling ahead of the 660Ti, especially when memory bandwidth becomes a factor. However, simple code issues can easily negate the memory bandwidth advantage.

      We are long past the days where official driver releases used to cause CTDs and BSOD like clockwork (even 3Dfx had this happening).

      Fanboys will always be fanboys.

        • indeego
        • 7 years ago

        Anecdotal: ATI drivers, standard on the OEM desktops I buy, have had constant issues, in basic 2D, for many, many years. Their driver release notes confirm.

        Nvidia, home builds, I’ve rarely had issues with.

        My anecdotal is enough to tell me to avoid ATI from here on out.

          • Bauxite
          • 7 years ago

          The plural of anecdote is not data.

          Not an anecdote: nvidia managed to kill quite a few laptops well before their time, and the true owners of these were either poorly compensated or not at all

          • Krogoth
          • 7 years ago

          What kind of issues?

          Nvidia drivers also have a history of a number of stupid, stupid issues in 2D mode over the years as well.

          Personally, the only problems that I ever had in recent years with video card drivers involved getting my 4850 CF working correctly with certain games, but that’s standard fare for SLI/CF setups. Single-card cards never game me any sort of issue on a Nvidia, Intel and AMD GPU.

            • Meadows
            • 7 years ago

            Will “stupid, stupid issues” become your new “trading blows” copypasta?

            • Krogoth
            • 7 years ago

            Will you try to come up with better rebuttals?

      • rrr
      • 7 years ago

      Sure ATI drivers don’t work, since company doesn’t exist, and it was the case for a few last years.

    • Voldenuit
    • 7 years ago

    TR reports problem.

    AMD confirms problem and commits to a fix.

    Somehow, AMD fanboys are upset.

    Welcome to another day on the internet.

      • Chrispy_
      • 7 years ago

      This needs bumping to the top.

      It would seem some people are never happy – perhaps they should get outside more often….

      • maxxcool
      • 7 years ago

      bumpitty L;)

      • Firestarter
      • 7 years ago

      Well, to confirm a problem in the drivers is to confirm that AMD isn’t the most perfectest company after all, which of course is a notion that just doesn’t fly with a real fanboy.

        • lilbuddhaman
        • 7 years ago

        Woah woah woah. AMD is certainly not perfect, but Ati on the other hand…

        “they’re like angels” ::flutter hands like wings ala Boondock Saints::

      • alienstorexxx
      • 7 years ago

      thank god they asume his errors and fix them.
      i know a lot about company’s that don’t learn how to treat their users, but they don’t mind, as they keep selling, they’re fine.

      • RenatoPassos
      • 7 years ago

      QFT.

    • kristi_johnny
    • 7 years ago

    It’s good to see AMD took notice of TR’s reviews.
    Pity that AMD start working hard on the drivers after a review revealed some problems with their drivers, and that after 1 year since the 7000 series was released.
    maybe GCN is that complex and writing drivers for it is hard, i don’t know. I hope they release these drivers fast and really improve performance, otherwise people will start looking for Nvidia only, and that’s not that good.

    • sweatshopking
    • 7 years ago

    this comment section is HILARIOUS. honestly, idk why these TR guys don’t go crazy with the people on the site sometimes (myself included).

      • nanoflower
      • 7 years ago

      It’s not just this site. I’ve seen the same sort of comments over on Slashdot after this article came up on that site. There’s only so many times you can stand up to the onslaught before you just realize it isn’t worth the effort.

    • ermo
    • 7 years ago

    Does this mean that the 7970 GHz Edition cards w/3GB RAM and 384bit memory bus (which are generally a little bit cheaper than GTX680s w/2GB RAM and 256bit memory bus) will finally be worth buying, assuming that AMD actually deliver on their promises with the driver memory management rewrite?

    Interesting development, to say the least…

      • ermo
      • 7 years ago

      To those of you who didn’t ‘get it’: The obvious point is that up until now, the AMD hardware may have theoretically been more powerful, yet the driver situation has meant that the competing offering in the form of the GTX 680 would often deliver smoother gameplay, which TR’s testing shows can be more important than raw pixel pushing power aka FPS.

      If AMD manages to fix this, the competitive situation with NVidia will be such that the 7970 GHz edition will be the better deal, hands down (less expensive, faster, just as smooth). You could even get a 7970 + a GF 6xx just for PhysX for around the same amount it would cost you to get your hands on a GTX 680 4GB. And the 7970 would still be better at 3x1080p and 4k resolutions, not to mention that the 7950 will become an absolute killer performer for the price compared to the 660 Ti.

      Bottom line: If you’re pining for a GTX 680, you might want to hold your horses until the AMD driver improvements forces NVidia to lower their prices in order to stay competitive.

        • Great_Big_Abyss
        • 7 years ago

        I didn’t know you could run Physx (even on a separate card) when your primary card is an AMD card. Wow. The World is saved.

        (actually, you can’t….and the only time I’ve mixed brands in the same rig was for folding, and it’s a huge pain in the ass. The drivers don’t play nice with each other.

    • Sargent Duck
    • 7 years ago

    @DaveBaumann

    Thanks for replying to some of the comments and thanks for the updates.

    I must say though, since buying my 7870 in late November and putting in over 70 hours in BL2, I don’t think I ever really noticed any off these problems mentioned. Maybe I did and just wasn’t bothered with it. Whatever the case, good to see AMD working on addressing the issue.

    Been a fan of ATI since the 8500 days and continue to be a fan. Regardless of what happens with AMD (the processor division), I hope the graphics section lives on.

    • tjoynt
    • 7 years ago

    In case anyone missed it, “… the initial software management for that [GCN memory management] was primarily driven by schedule…” means that they rushed out their drivers before they could fully optimize their memory management code for the new architecture, which could quite credibly lead to the stuttering issue. This is not to say that they necessarily did anything wrong or made a bad choice: no one has infinite time or resources and compromises must be made somewhere. The software engineers or QA team might have noticed the issue, but decided to go ahead because the deadline needed to be hit. They might have fixed the issue anyway when they rewrote the memory management software, but TR’s focus on the issue probably light a fire underneath them to make it a priority.

    My further discussion has already been eloquently elucidated by UberGerbil, saving me a good 10-15 min. Cheers! <raises a glass of beer>

    • zzz
    • 7 years ago

    People complain about AMD having crap drivers and a whole host of people say ‘they’re good enough now, they’re totally not as terrible as they used to be’. Just pointing it out that this topic itself specifically says ‘no, AMD still hasn’t gotten their sh*t together’ concerning drivers.

      • Majiir Paktu
      • 7 years ago

      I had this experience buying a GPU recently. I mentioned that I was only looking at Nvidia cards because I didn’t want to subject myself to AMD drivers—and I got flamed to hell over it. The 7970 is better, they said. The drivers are great now, they said. I almost believed them, too, because I don’t think I gave myself enough credit for understanding AMD’s position in the market. Either way, my GTX 670 says hello.

        • aspect
        • 7 years ago

        But if you look at the techreport reviews the 7970 does not have the same problems, it is only the 7950 that has these latency/stuttering issues.

        • Voldenuit
        • 7 years ago

        My last 3 GPUs have been ATI – 4850, 4870 and 5770 (3 separate builds, which is why they’re all clustered around the same level of performance).

        I just upgraded to a Zotac Geforce GTX 660, which is a massive jump from the 5770 (about 2x+ the performance), and already I’ve found the Geforce drivers to be less buggy. In the past, whenever ATI released a new Catalyst driver, it would make some of my media files unplayable under DXVA (they were always different files affected with different releases, so I never felt like the driver team was moving forwards – only sideways). With the 310.70 WHQL GeForce drivers, the 660 has played back every media file I’ve thrown at it flawlessly so far.

        Gaming experience is a lot smoother, but that’s probably due as much to the added horsepower as to any latency problems on the 5770 (5770 is VLIW, not GCN).

        I’m pretty CPU and GPU agnostic; I usually get whichever camp gives me the best value, silence/power and features at a given time. The decision to go lime green this time was a combination of a good deal on the 660, TR’s frame latency article, and a desire to try out hardware acceleration in Adobe’s Mercury. So far, so good.

    • Lee_144
    • 7 years ago

    That is one long-@ssed sentence:

    “Additionally, when we switched from the old VLIW architecture to the GCN core there was a significant updates to all parts of the driver was needed – although not really spoken about the entire memory management on GCN is different to prior GPU’s and the initial software management for that was primarily driven by schedule and in the meantime we’ve been rewriting it again and we have discovered that the new version has also improved frame latency in a number of cases so we are accelerating the QA and implementation of that.”

    Ok, maybe the “-” substitutes one period, but gosh!

      • DaveBaumann
      • 7 years ago

      [quote<]That is one long-@ssed sentence:[/quote<] Meh. Give me that one - first day back at work, lots to catch up on and just a quickie comment via mail :p Besides, it's nowhere near as convoluted as some of the word structures I published at Beyond3D!

    • HisDivineOrder
    • 7 years ago

    Kinda sad it takes over a year for AMD to DECIDE to fix something they clearly knew was broken, but hey… that’s bureaucracy for you. It takes someone shining a light on a problem in order for it to get fixed.

    The saddest part is how many fanboys came out of the woodwork to attack TR for shining said light. I think each and every one of those people who “defended” AMD against these “awful” attacks by TR should stop and realize how far gone they really are. TR is trying to help you and you’re attacking them in favor of the company that de-prioritized fixing an inherently broken system they already knew was broken?

    Perhaps you should be TR fanboys instead of AMD or nV or Intel fanboys, eh?

      • Essence
      • 7 years ago

      Thats a load of rubbish, the reason people attacked TR was bcoz of the games tested and over half being nvidia twimtbp games and the Rare fastest card TR used for nvidia (That card was only available for the Asia market & even then you cant find it). TR could have easily shut everyone up by doing a proper review with the most played games and not the “New Games” as that can be twisted anyway you like via choice of games.

      Most people would prefer to know/see settings, Location of bench in game or even raw data (TR Video), so they could replicate those benchmarks which i don`t remember TR showing to anyone apart from the game settings like AA/Res etc

      I think all review sites should start listing which section of the game bench is being used as each section will produce vastly different results.

      Having said the above, why is it that we suddenly started caring for the slight latency discrepancies? If people go back to the amd 6 series vs the nvidia 5 series, nvidia has massive latency spikes that could give one an heart attack. Never saw anybody lift an eye brow.. so back to my point about double standards… oh nvm

      [url<]http://www.hardocp.com/article/2012/11/12/fall_2012_gpu_driver_comparison_roundup/3[/url<]

        • Fighterpilot
        • 7 years ago

        Don’t waste your breath.
        Since the 7 series cards sailed past the nvidia competition in frames per second and price there’s been a desperate attempt to install slower but smoother as the new gospel.
        Retro testing as you mention will be dismissed as “irrelevant’ and will disappear like the dodo the second nvidia cards can compete for speed and price again.

          • Essence
          • 7 years ago

          I agree with you and have known that to be the case since AMD dropped their prices. The bonus and problems came when amd managed to squeeze extra juice from their new found GCN hardware. nvidia could have either compete, drop prices or start a propaganda war, unfortunately for us customers nvidia went with the later (notice this came just before holiday sales). so far we only have 1 other site (called hardware something) and TR saying this thing about latency, but both tested BL, skyrim and the usual culprits with max games not exceeding 3/4 that favor nvidia more. when i see a proper games review (7/8 games) I will be the first to agree, until then its all marketing and smoke for nvidia`s lackluster performance and price.

          How do you have a slower card, smaller die chip, locked down PCB and Components BUT a higher price? MARKETING + fanboys = win win

          VS

          A faster card, bigger die chip, unlocked PCB and Components BUT a lower price? NO MARKETING fud = lose lose

        • Damage
        • 7 years ago

        Well, you got one fact right, and that’s that we were attacked, from reddit to YouTube to a host of tech forums across the web, a strangely cohesive group of posters said many of the same things about us. They weren’t true, but you repeated them enough they created an impression, which I suppose was the point.

        What’s amazing is how every single one of the stated “reasons” you attacked us is false.

        Here is the article in question:

        [url<]https://techreport.com/review/23981/radeon-hd-7950-vs-geforce-gtx-660-ti-revisited[/url<] *You claim we tested "over half" TWIMTBP games. The games we tested are: -Borderlands 2 (TWIMTBP) -Guild Wars 2 -Sleeping Dogs (Gaming Evolved) -Assassin's Creed III (TWIMTBP) -Hitman: Absolution (Gaming Evolved) -Medal of Honor: Warfighter (Gaming Evolved) -Skyrim Gaming Evolved games: 3 TWIMTBP games: 2 Non-affiliated games: 2 Setting aside the fact that we prefer to test the most notable games regardless of whose label is on them, your claim is false. *You claim the GTX 660 card we tested is rare and only for the Asian market. Yet it is a US-market card that remains to this very moment in stock at Newegg right here: [url<]http://www.newegg.com/Product/Product.aspx?Item=N82E16814500269[/url<] This card cost $20 *less* than the Sapphire 7950 we tested, and *both* cards tested were the highest-clocked variants of their type available at Newegg at the time we published. *You claim we don't show the settings at which we tested, although we listed them exhaustively with screenshots of the in-game menus: [url<]https://techreport.com/review/23981/radeon-hd-7950-vs-geforce-gtx-660-ti-revisited/3[/url<] *You claim we didn't show the in-game locations where we tested and asked for raw data, including video. Yet we *embedded video in the very review* showing our testing session for each game. Again, see here: [url<]https://techreport.com/review/23981/radeon-hd-7950-vs-geforce-gtx-660-ti-revisited/3[/url<] We then posted video recorded at 120 and 240 FPS showing our Skyrim test session side-by-side on the GeForce and Radeon to illustrate the differences. See here: [url<]https://techreport.com/review/24051/geforce-versus-radeon-captured-on-high-speed-video[/url<] *You repeated the trope that we saw "vastly different" results from other sites, yet none of them test like we do. Whatever you were expecting, time to re-think. *You ask when PC gamers started caring about "slight" latency discrepancies. The answer, of course, is at the beginning of things, back when 3D gaming began. We've fought the battle for fluidity for years. Only in the past year have we used the language of frame latencies to describe it, but we've always wanted fluid gameplay. You are welcome to pretend PC gamers don't care about in-the-moment performance, just FPS averages, but you would once again be wrong on the facts. Now that we've established you were factually wrong on every "reason" why TR was "attacked," one question remains: why *was* TR attacked? Kinda makes you go: hmm.

          • Essence
          • 7 years ago

          AM not claiming, I just repeated what people have been saying (Am sure you know that yourself) to the person i replied to . If you go on any site, you will see the same questions. i assure you its not me, i only got carried away bcoz of what he said… when i said marketing fud i meant nvidia in general and not TR or any site btw

          Edit. btw that link i posted was not from my comment on that site which is being critical of you, otherwise i would be an idiot bcoz thats pretty blatant.

            • Damage
            • 7 years ago

            So if you can go to “any site” and see the same set of falsehoods posted about us… why do you think that is? Where do they come from? Spontaneous mass reading comprehension failure? And why are you repeating them without regard for the true facts?

            • Essence
            • 7 years ago

            I think thats up to you to disprove (More games maybe a good start). Good luck

            Personally i dont think its that big a problem as some people are making it out to be (Critics of TR), some people will always be unhappy of certain reviews, its how the reviewer reacts that makes it an issue i think, but i think you knew that anyway.

            • NeelyCam
            • 7 years ago

            TR’s results from a uniquely detailed test showed an issue. An AMD engineer admitted the issue and said a fix is being worked on.

            Why are you still whining about this? Why are you taking this so personally?

            • Turd-Monkey
            • 7 years ago

            The internet echo chamber has “decided” on those “truths” and will self perpetuate them.

            • MadManOriginal
            • 7 years ago

            Yeah, that’s about the gist of it and it happens aalllll the time. I don’t think there’s any kind of shady conspiracy going on if that’s what Scott is hinting at.

          • l33t-g4m3r
          • 7 years ago

          So that’s 7 games you tested out of how many total games available? Great job. I’m not saying you haven’t found something, but you certainly put the rose colored glasses on when it comes to discovering if this is prevalent across the board, or whether it’s something that should actually warrant recommending a less capable card of the opposite brand over. Especially when AMD says it’s something they have a solution to, and will be releasing shortly.

          The question you CLEARLY haven’t answered in your article is, does this exist in games amd does well in? That’s why you’re getting flak. You didn’t answer that, and it might be that it doesn’t exist. That, and a lot of past nvidia owners know for a fact FERMI has horrible stuttering issues, but that’s completely passed over because it was only one generation back. Even though Nvidia is STILL SELLING THEM.

          Questions people might want answers to:
          How does a VLIW 69XX fare in comparison, being it has the old memory architecture?
          Do well performing games also experience stuttering?
          Why not throw in a Fermi card in for good measure?
          Powertune?

          Now these are just questions you might want to answer if you were impartial, but if you have an agenda and are only interested in pointing out a singular issue to make amd look bad, feel free to just ignore everyone calling for additional information.

          Regardless of how obviously slanted the article was, it still brought a problem to light, and I don’t disagree with it. I only disagree with [i<]how it was done[/i<]. There wouldn't have been be ANY of this controversy if it had been done in a less polarizing fashion.

            • sweatshopking
            • 7 years ago

            why do you bother posting? you just ignore the facts when they’re presented to you.

            • superjawes
            • 7 years ago

            He’s not ignoring facts, he’s just saying that TR needs to TEST ALL THE GAMES in order to genuinely prove their point. Honestly, though, I’m still waiting to see the l33t-g4m3r benchmarks for the 7950 on the original DOOM.

            /tapsfoot

            • l33t-g4m3r
            • 7 years ago

            7 Games are not representative of the thousands that remain. I’m not saying TR needs to test all the games, just a few games that we know work on AMD cards. Otherwise, this could just be a problem with only unoptimized games. Sure it’s a problem, but is it bad? Probably not. TR hasn’t tested for it, but somehow 7 games magically mean all the other games have the same issues. I doubt that very much, and that’s why people are arguing about it.

            If you wanna be an ASS about it, then be an ASS. The problem with that is it makes you look like you’re guilty, and you lose even more credibility.

            • superjawes
            • 7 years ago

            There’s a couple issues with that statement. First, as Scott mentioned, he only ever said that the results were representative of the 7 games tested. Said games are relevant because they are current games, and it is very likely someone purchasing a [b<]new[/b<] video card will make a decision based on games that actually put the GPU's feet to the fire (and are candidates for purchase in the same timeframe), not something lightweight from five years ago. Secondly, even if you were to choose from the "thousands" of available games, which ones do you even choose? How do you choose them? Who's to say that you can't still choose games that the Nvidia card does well in, or even games that only serve to pad AMD's results? Is that fair? You've now introduced a very real problem of testing creep, multiplying the amount of time and work required to provide what you consider "fair" testing. There are still only 24 hours in a day, last time I checked. I'm being "an ASS" about it because your attitude on this matter has always been critical of TR because they haven't met your unrealistic expectations of what is "fair." So "if l33t-g4m3r ain't happy, ain't nobody happy." I've tried the reasonable approach (and I may have been tricked into taking it again here), and it only gave me a headache. So unless you want to settle down and have a nice, calm chat, I'll continue to "be an ASS about it" and I will wear it proudly.

            • MrJP
            • 7 years ago

            You’re wasting your energy. All of this was pointed out to him in the comments under the previous articles and he didn’t pay any attention then either. Sometimes it makes me wish the comments had a blacklist system.

            • superjawes
            • 7 years ago

            I know…I was one of them pointing these things out. At least I can just be an ass now 😆

            • sweatshopking
            • 7 years ago

            you haven’t seen if scott has a legit response. if he does, l33t will ignore it and keep posting his insane ramblings despite the reality staring him in the face. ask l33t if volcanoes make more co2 than humans, and see what i mean.

            edit: scott has posted.

            • l33t-g4m3r
            • 7 years ago

            Where did I say I ignored anything? WHERE? I swear every post you make is a lie! I fully acknowledge TR’s findings, they just aren’t very comprehensive.

            It’s good that they are doing such tests too, but you don’t make blanket statement claims about a card when you’ve only tested 7 games.

            • Damage
            • 7 years ago

            Point of fact. We never claimed our findings were comprehensive. We limited our comments to the scope of the games we tested. From the conclusion of the first 7950 vs. GTX 660 Ti rematch article:

            “Instead, we have a crystal clear recommendation of the GeForce GTX 660 Ti over the Radeon HD 7950 for this winter’s crop of blockbuster games. Perhaps AMD will smooth out some of the rough patches in later driver releases, but the games we’ve tested are already on the market—and Nvidia undeniably delivers the better experience in them, overall.”

            You’re making a classic straw-man argument, ignoring the facts of the situation. You guys keep doing this, arguing against things we never said, blowing qualified statements into gross generalizations, and getting faux-enraged over obviously false assertions we never made.

            Take some valium, maybe, and sit by the fire for a bit. Sip some herbal tea. Listen to classical music. I dunno. Just chill, reflect, and make your future disagreements count for something, if you must continue having them so frequently. Life is too short to be pretend angry all of the time.

            • sweatshopking
            • 7 years ago

            scott comes in with the facts.

            • l33t-g4m3r
            • 7 years ago

            You’re agreeing with what I’m arguing about. The article wasn’t comprehensive, and was based on 7 games. That’s all I’m saying, and apparently so are you. The controversy stems from the fact that people are misconstruing your findings to mean that every game has this problem, when it quite possibly does not. Add Nvidia trolls to the mix, and Voila! Instant flamebait.

            Then there’s SSK, and he’s so consistently full of it, that you can’t ever take anything he says serious.

            • sweatshopking
            • 7 years ago

            [quote<] Add Nvidia trolls to the mix, insane people, and AMD fanboys and Voila! Instant flamebait. [/quote<] ftfy

            • tfp
            • 7 years ago

            But then you ignore AMD saying there is a general issue with the driver and they are working on fixing the problem.

            If AMD doesn’t need more games tested why are you so worried about the results?

            • l33t-g4m3r
            • 7 years ago

            Nope. I’m not ignoring that, but I don’t think it effects all games equally across the board either. What I’m concerned about is that the 7 games tested were not a accurate representation of other games, and that this paints amd in a more negative light than is necessary. That’s what I’m concerned about. Perspective. 7 games give you none, only a glimpse into a problem that may not effect other games as much. After all, it took this long for anyone to notice, and AMD didn’t have it on high priority until now.

            It’s not like amd wasn’t working on it either, they were and it was going to be released eventually. TR just happened to push it into sooner rather than later.

            • Grated
            • 7 years ago

            Not really sure what you expect here… A review with 20 games every time? I’m pretty sure TR doesn’t have the people to do an extensive test run with 20 games for every review…
            Everything can always be better, but based on this review, the conclusion was correct (and too bad it was against AMD this time… ). Those were the facts of this particular review.
            If you’re in the market for a graphics card, read more than one review ;-).

            • l33t-g4m3r
            • 7 years ago

            You don’t? I clearly stated only the addition of one or two games that we know WORK on amd hardware. That’s not 20, and I’m getting tired of this stupid meme.

            The only point of including an additional game would be CONTROL. You know, the scientific method? Perhaps I’d like to know how the card performs NORMALLY, and not with some game that has had little or no optimizations.

            Yes, the conclusion was correct, that’s not the point. I’m saying without a control game the review was skewed to make AMD look worse than it might actually be. None of the 7 games picked seemed to be in any way optimized on the GCN architecture, and AC3 was so bad it ended up with it’s own graph. If every game picked performed similarly to AC3, don’t you think it’s a bit disingenuous to base a general conclusion on? Damage even said the findings were not comprehensive, so what’s so hard for you to get here? The findings are not comprehensive, they only point out that an issue exists. I’m merely pointing out that this bug isn’t the end of the world scenario like the damn shills are claiming. It’s not.

            The ONLY game I’ve even seen complained about by 7950 owners in this forum is Rage, and even at that they incorrectly blame the game over their drivers. The memory management apparently isn’t a big deal, outside of unoptimized new games like AC3, or games that use new rendering techniques and OpenGL like Rage. It’s a mountain out of a molehill situation.

            • sweatshopking
            • 7 years ago

            wtf. you have 7 “control” games. quit being silly.

            • l33t-g4m3r
            • 7 years ago

            Bullspit. Those were all games that performed poorly on GCN hardware, with skyrim being the exception because it is dx9.

            I clearly stated the control game should be an older dx11 title that we know performs well.

            ps. I have no clue why I’m even replying to YOU, outside of basic manners, when it’s quite obvious that all your posts are pure trolling bs. You don’t even bother to think about somebody else’s perspective, and your replies are all bait to get a reaction. Good day.

            • Grated
            • 7 years ago

            The games were released between August & November. With AMD having early access to games as well, there was enough time for AMD to get the games to work (& with 3 of these games under the Gaming Evolved program. If those don’t work on GCN, what does…)

            I want my new games to work when I buy them, not in 6 months or a year.

            • l33t-g4m3r
            • 7 years ago

            I don’t disagree. AMD fubar’d their drivers, and it shouldn’t take 6 months to fix things, but that’s about what it’s been taking them. I’ve noticed a steady decline ever since their 5X series, and it’s been downhill since. Hell, that’s why I went nvidia after my 4870, because things were getting out of hand. Not just with drivers either, since amd was skimping on the hardware too. The 5X series could have been shipped with 6X tessellation hardware, but they skimped to sell the 6 series to the 5 owners. Screw that, I went with fermi, and haven’t regretted it, even if it does have issues that were ignored here. No card is ever perfect, so you just have to deal with the shortcomings.

            • sweatshopking
            • 7 years ago

            Ps – because I am the light to your darkness?

            Pot and kettle again! Thanks for at least trying to be genial!

            • l33t-g4m3r
            • 7 years ago

            You know how satanists invert everything because that’s their twisted mentality? That’s you. You’re like a sadistic satanic freak.

            If we ever want to know the actual truth about something, just invert what you say and there it is.

            • sweatshopking
            • 7 years ago

            I don’t actually know any Satanists, so ill have to take your word for it. Wait though, if I say everything backwards and I say volcanoes produce less co2 DOES IT MEAN THEY PRODUCE MORE?!?!?

            • l33t-g4m3r
            • 7 years ago

            You need to STFU about volcanoes because you’re so wrong it’s not even funny.
            [url<]https://www.scientificamerican.com/article.cfm?id=how-do-volcanoes-affect-w[/url<] [url<]https://www.scientificamerican.com/article.cfm?id=iceland-volcano-airspace[/url<] [url<]http://www.geology.sdsu.edu/how_volcanoes_work/climate_effects.html[/url<] [url<]http://vulcan.wr.usgs.gov/Glossary/VolcWeather/description_volcanoes_and_weather.html[/url<] That icelandic eruption that grounded air travel was so bad that it caused immediate climate change in the region by blocking out the sun. We even felt the effects here in America. You need to just shut the hell up about this stuff, because you don't know what you're talking about. PS. That logic doesn't work because you admitted you were saying it backwards, so that's a satanic double negative, and you don't have to know you're a satanist to be one. All evil and mentally deranged people are basically satanists, whether or not they admit it. [url<]http://www.nydailynews.com/news/national/gabrielle-giffords-shooting-frightening-twisted-shrine-arizona-killer-jared-lee-loughner-yard-article-1.149211[/url<] Was Loughner an admitted satanist? I dunno, but when you have a skull shrine in your backyard, you pretty much are one in my book. Same thing with your twisted backwards mentality. You're basically a satanist, whether or not you admit or realize it. It's just a generalized term for deranged madmen who have certain common characteristics, which you happen to have. Inverted thinking happens to be a key trait among devil worshipers, although it seems to be a natural phenomenon for you. Technically, that just means you're a sick freak, but you might as well be a satanist in my book.

            • superjawes
            • 7 years ago

            [quote=”Link 3″<]Volcanic eruptions can enhance global warming by adding CO2 to the atmosphere. However, a far greater amount of CO2 is contributed to the atmosphere by human activities each year than by volcanic eruptions. T.M.Gerlach (1991, American Geophysical Union) notes that human-made CO2 exceeds the estimated global release of CO2 from volcanoes by at least 150 times.[/quote<]What was the point of this volcano argument?

            • l33t-g4m3r
            • 7 years ago

            SSK claims it doesn’t affect climate, while the 2010 eruption of Eyjafjallajokull caused an immediate drop in temperatures worldwide. Perhaps it doesn’t contribute greatly to warming, but volcanoes certainly can cause immediate effects on climate, not to mention pollution. Either way, I don’t believe Co2 causes warming or is a pollutant. It’s air, and for the most part, certain people in government are hoaxing the citizens about warming in order to pass economy destroying taxes that only serve to line the pockets of the politicians. Take the taxes out of the argument, and then maybe you could get people to agree on something, but for the most part it’s pointless, since you could shut off the grid worldwide, and it would only contribute 1-2% towards doing anything. It’s all a hoax to raise taxes, IMO. The carbon trade system is a scam where mega corps can buy credits from 3rd world countries and shut down smaller competition, it’s Eco-fascism. It would be nice if people decided to look into other causes of warming, like perhaps the sun and it’s cycles, and government destruction of the atmosphere from nuke testing, but when it’s all a singular drumbeat for carbon taxes, you know somethings off. Especially when this is all an admitted Marxist takeover tactic. Marx invented fake environmentalism to pass laws to take over production.

            Either way, randomly spouting VOLCANOES111!!!!!ELEVENTY!!!!, every freaking post is retarded. Get over it, SSK. Volcanoes do change the climate and pollute the air. That’s a fact.

            • sweatshopking
            • 7 years ago

            LOL.
            nice way to change the argument. you CLAIMED that volcanoes have put more co2 into the atmosphere than mankind has ever. obviously volcanoes CAN impact weather. I’m not denying science, you are. Nice try, though smarty pants.
            i wish i knew how to search the front pages, so i could pull up your comment.

            The facts is, you have a serious problem with facts. nobody is saying there isn’t more to the story in terms of warming, or that co2 is the ENTIRE story. I never made any comment suggesting such a thing. YOU claimed volcanic eruptions put more co2 than man EVER did. when i posted the FACTS, YOU refused them. Making me out to be crazy is nuts. i’m ALWAYS down with reality and science. You’re the one who has issues with them.

            • l33t-g4m3r
            • 7 years ago

            LOL. You’ve been constantly harping Volcanoes don’t impact climate, or pollute the air, but now are claiming that you’re down with that, and you are Mr. Reasonable with an open mind. No, buddy, you have been anything BUT reasonable, and you care more about your stupid crusade than anything else, facts or openness be damned. Global warming is a theory, but what isn’t a theory is these so called “cures”. I’m not concerned about climate change in the slightest, what concerns me is what the government supposes to do about it. The planet has gone through plenty of climate change before we were here, and will continue to change after we’re gone. It’s not the end of the world for temperature to go up or down a few degrees every hundred years or so. There are bigger problems in life, and this one doesn’t concern me.

            This isn’t a problem that will concern anyone down the road either, especially if the economy doesn’t improve. The more people like you call for taxes and regulations in a recession, the more it looks like you are contributing to the problem, or don’t care about it, and the public perception will become so negative that you will have destroyed any chance to do anything for a long time.

            Best not to be so damn divisive about it, and upon discovering the full facts about the situation, only implement measures that are completely bipartisan, are economically neutral, and show respect to your fellow man. Otherwise, forget it.

            A good first step would be to stop inappropriately bringing up volcanoes in discussions about video cards, or other non-related discussions. Then again, that might make you seem less schizophrenic than you actually are, but that’s a trade off we’d all be willing to accept.

            • sweatshopking
            • 7 years ago

            idk what the f you’re talking about. I NEVER said volcanoes couldn’t impact climate. that makes no sense, and it’s stupid. i’m not stupid, so i wouldn’t say that. I NEVER SAID THAT. QUIT MAKING STUFF UP. LINK OR IT DIDN’T HAPPEN. DAMN I WISH I KNEW THE PAGE WE WERE ARGUING ON SO I COULD SHOW YOU THE CRAZY YOU ARE!!!!!!

            you’re agreeing to STAY ON THE CONTENT OF THE SITE WITH OUT RIDICULOUS RAMBLINGS ABOUT GOVERNMENT/RELIGION/GUNS/ETC?!?!?!? I’M DOWN. IF WE CAN AGREE TO THAT, I’LL STICK TO IT. BUT, IF YOU KEEP BRINGING UP CRAZY, I’LL KEEP CALLING YOU OUT ON IT.

            • bjm
            • 7 years ago

            You’re only mad because of the conclusion that was reached, not the methods that was used to get there. TR has always reviewed about the same number of games for each of their reviews. Some reviews were shorter, some a bit longer, but seven is about the normal range. Why are you only crying for more games now?

            I’ll put it simply for you: because the recent events have portrayed AMD in a bad light — that’s why. Had these recent reviews ended up portraying nVidia in a bad light, you would not be arguing these same positions. What would your posts look like if it was reversed? Easy, [url=https://techreport.com/discussion/22384/amd-radeon-hd-7950-graphics-processor?post=611707<]they'd look like this.[/url<] And that review only had six games! When TR's methods end up portraying AMD in a good light, their methods are "appreciated" and "loved". When AMD gets the bad end of the deal, you cry, "What?! Just seven games? That is not indicative of anything at all! Test more games until AMD is proven to be better!" Your combative attitude and you-have-an-agenda! posts only show you to be the very thing you claim to hate -- a biased, nerd raging, hypocrite.

            • sweatshopking
            • 7 years ago

            i wasn’t talking about this post. i mean you ignore facts in everything. why bother yelling at scott, when what he says will mean nothing to you? you’ll just keep on yelling the same crap regardless of what the truth is. if you can’t handle it, you just stop on this discussion only to bring it up again next time and act like the facts didn’t exist.

            • nanoflower
            • 7 years ago

            So the fact that Scott tested 3 (I think that’s the number) games that were built with AMD’s help is something that you are either choosing to ignore or have blissfully remained ignorant of even though he’s brought up the fact multiple times. TR did a good job of covering games that are popular as well as showing games that were built independent of AMD/Nvidia as well as including games that AMD and Nvidia helped optimize.

            As for your suggesting that they test numerous other cards the issue has already been somewhat addressed by AMD stating that the issue is due to a the driver for the new GCN cards. Testing a VLIW card would only confirm what AMD says about the problem not being associated with those older cards. In addition there’s the issue of time. The more cards/games/configurations to be tested the greater the time it takes. I don’t blame Scott for limiting himself to a few cards and a mixed set of games as even that small set must have taken some time to work through. It’s enough to get everyone’s attention and I’m sure other sites will follow up on the work that Scott has done. That will include testing a wider variety of cards and games. (If for no other reason than they will use what they have available and/or prefer.)

            • MadManOriginal
            • 7 years ago

            The controversy all came from people on the interwebs going crazy over the results for all the wrong reasons, not because of ‘how it was done.’

          • theleif
          • 7 years ago

          As a long time AMD fan I thank you for these articles.

          • erwendigo
          • 7 years ago

          Wow, fucking shit! The first part is very similar, in essence, to my previous response to this fallacious comment.

          It´ld be very frustating see laments about the inclusion of SOME Twimtbp games, when the review includes gaming evolved games too. And more games.

          It´s… a absurd accusation!!

        • Silus
        • 7 years ago

        LOL, more than half the games in TR’s benchmark suite is part of AMD’d Gaming Evolved program. Your post just shows how clueless you are.

        Latency discrepancies are a new way to assess smoothness. What in the past was a reviewers opinion of how smooth the game was on a certain config (you posted a hardocp link, so you should know what I’m talking about this [H] always included that perspective in their GPU reviews) can now be seen with numbers. This is how things work in the real world. Methodologies evolve and a better one should always be used to better show the strengths and weaknesses of hardware. Just like that hardware evolved to do more than it did in the past. GPUs can do much more than play games today and NVIDIA started it quite a few years ago, but for some reason (fanboyism) some people disregarded that and actually criticized NVIDIA for their compute push with previous GeForce generations, because what mattered were the fremerates!!! Years later and with AMD at last doing something with compute, compute is suddenly very important because AMD’s high-end desktop GPU is actually better in compute than NVIDIA’s high-end desktop GPU. Double Standards right ?

          • Spunjji
          • 7 years ago

          I liked your post up until the point where you started the unrelated rant about compute. Firstly you managed to ignore the fact that AMD were the ones to initiate a shift towards general-purpose computation on GPUs (see R580 and Folding@Home). For their part, nVidia deserve credit for pushing much harder and much more successfully with GPU compute once they got down to it.

          However, this brings me to your next point, where you claim that nVidia’s compute focus was criticised because of fanboyism. You conveniently ignore two factors – firstly, that there is *still* not a concrete consumer use for GPU compute even today. Secondly, that Fermi (which is presumably what you’re referring to) was inherently and irrefutably compromised by this focus in its first iterations. It was delayed and used more power than necessary to achieve its performance goals.

          So, yeah, not really fanboyism there.

            • Silus
            • 7 years ago

            You give too much credit to fanboys. They criticized NVIDIA’s GPU compute push based on the fact that AMD had nothing to compete with it. Now that it does, compute is a major feature that should not be ignored and that’s where the double standard comes from.
            I would disagree that there’s no concrete consumer use for GPU compute. There’s plenty to use a GPU on, other than games. Plus, NVIDIA’s GPU compute focus was mostly for servers, not consumers. Consumers just got it as a bonus initially, but there’s not much money to be had with compute in the consumer space.
            Fermi was not compromised in any way by this focus and many of its problems/delays were related with the problematic 40nm transition, which were easily fixed a few months later, once the process was stable. The only architecture that was compromised by this focus was the one in GT200, that got its 30 DP Stream Processors added in at the last minute, which made the chip an unnecessary monstrosity. Fermi was designed to be what it ended up being and it was quite successful in both consumer and server spaces.

            And I didn’t ignore GPGPU from AMD with Folding@Home, but that was it. That’s not a push when not a single investment was made after that. So either we have very differing opinions on what investment is, or AMD did almost nothing in regards with GPGPU until now. Even now I don’t see much of anything from AMD in this regard…

        • maxxcool
        • 7 years ago

        Paid troll go F$%ck your self…

        Username:Essence
        Rank:Gerbil In
        TrainingGroups:
        Add friend
        Add foe
        CONTACT ESSENCE
        PM:Send private message
        USER STATISTICS
        Joined:Mon Jun 04, 2012 1:54 pm
        Last visited:Wed Jan 02, 2013 9:47 pm
        Total posts:0 | Search user’s posts
        (0.00% of all posts / 0.00 posts per day)

          • Lazier_Said
          • 7 years ago

          He isn’t necessarily a paid troll.

          He could be an unpaid moron.

            • maxxcool
            • 7 years ago

            :/ which is worse ?

          • Essence
          • 7 years ago

          I don’t expect anything less from nvidia focus groups, most of them are 18 to 21 years old. It also shows what type of a company nvidia is by hiring people like you. nvidia have been known for shills, even the UK BBC news published an article over it.

          Am glad AMD r*ped nvidia in its own game i.e. FPS, then quickly came Compute & next will be so called “smoothness” over a couple/few driver updates for the 7950 since it’s the only card shown to have some problems in couple of nvidia flavored games. Fanboys have always said nvidia is smoother, am glad that can be PROVEN wrong eventually with data to back it up one day… on that sunny day the drums of smoothness will beat hard lol. So, until then Nvidia needs to drop prices hard (now), it r*ping the customers with higher prices and inferior cards.

          100% of sites that reviewed or tested the 7970 GE vs 680 admit and show the 7970 is faster & SMOOTHER Inc. TR but most never mention it in the conclusion, why? Neither do we have this shouting as in the case with 7950 aftermarket fan with stock speeds vs highest overclocked 660 ti… well It’s the only thing they have which ironically can’t be proven yet lol… 99% of people in every forum are recommending the overclocked 7950 over the 660ti and with good reason… The fundamentals haven’t changed one bit.

          including NVIDIA NOT DROPPING PRICES.

          Waiting for the 20 nvidia fanboys for down thumbing as usual on this site. They all seem to be hanging here since TR gave them some hope again LOL

            • sweatshopking
            • 7 years ago

            why are you here?

            • maxxcool
            • 7 years ago

            Hahahah… idiot. If you think TR is Nvidia-centric in its reviews I invite you to self fornicate and go elsewhere. Tr is the most unbiased site I can think of that looks decent, reads great and is supported by the “industry”.

            Even AMD agreed that they F#$%ed up. You should be happy, not a but hurt 15yr old Trollololol.

        • shank15217
        • 7 years ago

        Anything that helps improve the game play experience for AMD card holders is a good thing. These tests may actually make a real difference in our gameplay and also ended up doing some serious QA for AMD in general for the cost of free. AMD certainly appreciated this. I don’t see why you need to bash them for not having the best possible drivers or bash TR for doing a great service for AMD and the people who own these cards.

        • erwendigo
        • 7 years ago

        “Thats a load of rubbish, the reason people attacked TR was bcoz of the games tested and over half being nvidia….”

        Borderlands 2: TWIMTBP

        Guild Wars 2: NEUTRAL

        Sleeping Dogs: GAMING EVOLVED

        Assasin´s Creed 3: TWIMTBP

        Hitman Absolution: GAMING EVOLVED

        MoH: Warfighter: GAMING EVOLVED

        SKYRIM: NEUTRAL

        Ouhh yeah!!, all of the games are pronvidia. All of them. Shut up fanboy, and don´t lie more!! Is a very fair selection for AMD cards. 3 Gaming evolved (proAMD) games, 2 Twimtbp and 2 “non affiliates” games.

        And no, Skyrim isn´t a TWIMTBP, if the game runs better in geforce is for the better drivers, no more, no less.

        The same about Guild Wars, that supposedly runs better in radeon but it isn´t a gaming evolved title.

        About this other:

        “and the Rare fastest card TR used for nvidia (That card was only available for the Asia market & even then you cant find it). ”

        More fanboy´s lies. This card is available, widely, in Europe (you know, the FIRST market in units for this class of hardware). Not only a “rare and asiatic card”. And the problem detected here in TR isn´t related with the pure performance, moron. Don´t launch crap like a monkey. It´s about the drivers. Gains of pure performance of, maybe 10%, doesn´t hide this problem.

      • DaveBaumann
      • 7 years ago

      [quote<]Kinda sad it takes over a year for AMD to DECIDE to fix something they clearly knew was broken, but hey... that's bureaucracy for you. It takes someone shining a light on a problem in order for it to get fixed.[/quote<] Not sure how you derived that from the comments. The memory manager was already undergoing a rewrite for separate reasons, it is a development phase such that tried to see if it made any improvements to the types of frame latencies shown here and it has done in some cases so the reasons for prioritising this and getting it through full testing have increased.

        • Silus
        • 7 years ago

        Radeon owners certainly hope that AMD is actually doing their job on drivers. Unfortunately, AMD is not known for their stellar driver support. Bugs are always part of software, but AMD is consistent in having too many and taking way too long to fix them. That’s why people criticize you for taking so long to solve something that’s been in the “field” for quite some time now.

          • BlackStar
          • 7 years ago

          Nvidia drivers have always crashed *much* more often than AMD drivers on my various gaming rigs. They still do.

          Anecdotal data just refuted your anecdotal data. “Not known” my ass.

            • Silus
            • 7 years ago

            It always amazes me that the argument in these subjects always go to: “I never had problems so there” or “company Y also has problems, so there”

            Neither argument makes AMD’s drivers less horrible than the competition. Yes the competition also has problems. Software has bugs, but AMD always has more and they take longer to fix them, which hurts anyone that is their customer. This is not anecdotal evidence, it’s fact. Fact that has been mentioned time and time again in various articles/reviews for years and years in countless sites, yet you and others like you, continue to dispute what is the reality for years and years. Have fun living in AMD’s RDF.

        • willmore
        • 7 years ago

        Plus, if memory serves, these latency spikes weren’t seen in the initial review of the AMD card, so the issue hasn’t been known for a year, it’s come about much more recently.

      • jensend
      • 7 years ago

      They didn’t know it was broken, they knew it was non-optimal and that they could do better.

      In any non-trivial real world software or hardware system there will be tons and tons of areas where it could have been improved. To ship something in finite time you have to make tradeoffs with incomplete information about the impact of the tradeoffs you’re making.

      They decided that they needed to ship what they could come up with quickly, that it was worth starting work on a rewrite, and (apparently) that this rewrite didn’t merit the highest priority imaginable.

      In making those decisions they had imperfect information about how much better they’d do with a rewrite of these routines, about how much effort/development and QA time that would take, and about how the improvements from that rewrite would compare to many other areas of possible improvement. With the additional information from trying to debug the problems that have surfaced and from testing of the now-apparently-nearly-completed rewrite, they’ve decided to adjust their priorities.

      You can claim that they should have been better informed about the latency impacts, that they should have had different priorities in deciding where to put developer effort a year ago, etc. (Be careful of hindsight bias in doing so.) But there’s zero evidence to support your claim that the decision was due to the intrigues or laziness of some shady bureaucracy.

        • DaveBaumann
        • 7 years ago

        [quote<]You can claim that they should have been better informed about the latency impacts, that they should have had different priorities in deciding where to put developer effort a year ago, etc. (Be careful of hindsight bias in doing so.) But there's zero evidence to support your claim that the decision was due to the intrigues or laziness of some shady bureaucracy.[/quote<] To be clear, we have not had a single customer complaint or issue arise directly related to what is portrayed here. Indeed, these type of frame render time graphing graphs have been used for many years at HardOCP and no comment have arisen about this for single GPU, most reviews other reviews also show min FPS, are an indicator to similar things (though you need to consider how min FPS is actually calculated); indeed, as far as I had seen, TR's prior reviews using this method didn't highlight anything particularly untoward. Likewise, we all use our GPU's and, for instance, I've put may many hours in to Borderlands 2 (207 in fact :gasp:!) on other an HD 7970 Ghz Ed or a HD 7870 and I genuinely can't notice any affects on the gameplay related to this to have even flagged it up as something I can perceive. However, we take it seriously and this has occupied quite a lot of architectures and software engineers time over the past few weeks to see where this is coming from and whether it can be improved.

          • MrJP
          • 7 years ago

          To be clear, the HardOCP plots and minimum FPS reported elsewhere have never shown this sort of issue because they use FPS averaged on a per-second basis rather than the individual frametimes used here.Hence they could only ever hope to capture low frequency drops in performance rather than the higher frequency stuttering in theory being flagged in this analysis.

          Having said that, I’ve got to agree with you that I haven’t “felt” this as a problem with my 7950. Perhaps even more shockingly, I didn’t really perceive a massive problem with stuttering with my previous 4850 Crossfire setup, despite the stuttering being clearly evident in the frametime numbers of some benchmarks I ran pre- and post-upgrade. Perhaps some people just don’t have much sensitivity to this.

          This does lead me to think that the one missing strand to TR’s investigation is some controlled blind testing with a small range of people to try to identify the threshold at which the stuttering becomes a significant problem. In fairness Scott did include comments about the perception of the smoothness in several of the benchmarks in the 7950/660Ti review, but I’d still like to see a more scientific approach to establishing where the limits really are.

            • flip-mode
            • 7 years ago

            Blind testing is a good idea.

            • DaveBaumann
            • 7 years ago

            [quote<]Having said that, I've got to agree with you that I haven't "felt" this as a problem with my 7950. Perhaps even more shockingly, I didn't really perceive a massive problem with stuttering with my previous 4850 Crossfire setup, despite the stuttering being clearly evident in the frametime numbers of some benchmarks I ran pre- and post-upgrade. Perhaps some people just don't have much sensitivity to this.[/quote<] Yes, I think this is the area that needs further investigation. It is interesting that TR has does this type of analysis, and flagged an "issue" but to all intents and purposes I have no frame of reference that indicates that this is an "issue" that affected people in most of the cases; as mentioned nobody has really reported this an issue that affects their gameplay.

            • superjawes
            • 7 years ago

            I think this “issue” is something that could get raised in coming years, though. “Smoothness” is mostly a subjective analysis. If you see a movie in the theater, you might say that everything is smooth, but [i<]The Hobbit[/i<] has jarred a lot of people who saw it in 48 FPS (whereas 24 FPS is "normal). Some people are going to immediately prefer the higher rate because they like how much smoother it seems to them, but others will continue to see the traditional rate because they are used to it, and the higher rate seems strange to them. Both groups have different "smoothenss" standards. However--and this is why I think we're going to hear a lot more about smoothness in coming years--now that Pandora's box has been opened and higher rates are available, I think more people are going to prefer 120 Hz televisions and 48 FPS movies, and visual inconsistencies will become less tolerable.

            • Ryu Connor
            • 7 years ago

            I’d argue it’s because no one figured there was anything that could be done about it.

            Other assumptions would most likely take precedence. Such as believing that I the user had set the settings too high. That this is the game engines fault or that this specific area is just more taxing in some way (back to having pushed my settings too high).

            I’m also not sure I’d say none of your users noticed. The previous articles detailed many people tossing out a variety of fixes for this issue. Claiming that PowerTune adjustments or RadeonPro could make it all better, with people implying that it had fixed it for them.

            I personally find it a little disingenuous for a representative of AMD to slide into this thread now and start a subtle spin about if they are perhaps doing all this engineering work for nothing.

            • willmore
            • 7 years ago

            Or that anyone who though that the game didn’t feel ‘fluid’ enough just backed down the settings until even the long frame times were <16ms and they effectively went away.

        • Silus
        • 7 years ago

        They didn’t know it was broken ? Give me a break. The tests that are done in their labs for products and drivers should be quite thorough. Of course bugs will always exist, but this is something that AMD always neglects and is clear with their usual driver state.

        “They decided that they needed to ship what they could come up with quickly, that it was worth starting work on a rewrite, and (apparently) that this rewrite didn’t merit the highest priority imaginable.”

        This “argument” should not used to defend AMD, as it just gives an even worse impression of them if that’s what they do. They should release what was thoroughly tested and working, not something that they come up quickly…

          • Spunjji
          • 7 years ago

          I would love to live in the perfect little world in your head, but in reality this is what all companies in a capitalist economy have to do in order to survive.

            • Silus
            • 7 years ago

            In order to survive, they have to release half baked drivers with promises of fixes that only see the light of day months and months later ? Sorry to burst your “reality” but that’s not what happens. Half baked products (hardware or software) are met with a huge backlash. Only die hard fanboys can consider a constant problem to be a “necessity in order to survive”. Radeon owners that spent their hard earned money on those graphics cards, some of them quite expensive, probably don’t see it quite like you do.

            • Firestarter
            • 7 years ago

            The drivers work in a vast variety of different games and setups, both old and new, on several OSes. I wouldn’t say they are ‘half baked’. A performance (not functional!) problem with a few of the games is not something that would prevent AMD (nor Nvidia) from releasing their drivers, as at least they know the game works, even if performance is suboptimal. And the performance issue that has been identified by Techreport is not something that would easily show up in testing, evidenced by the fact that Techreport has so far been the only major publication to find it.

          • RenatoPassos
          • 7 years ago

          Have you ever developed complex software? And what about complex software that is in the cutting edge? What about real-time or soft real-time apps? Heavy multi-threading?

          Most important of all: how many software QA guys/gals do you know? Did they tell you that they will never find every single bug? What about smoothness being kind of subjective – even after you do stuff like “Inside the Second”?

          We’ve got so used to software we keep forgetting: computer and software engineering is incredibily new compared to building engineering, architecture, medicine, etc. and it has become too big, too complex and pervasive already. And that means the thing is [i<][b<]not[/b<][/i<] as predictable and precise as some thousands of years old sciences. And yes, some will do better at it. And you [i<]always[/i<] compromise when you are trying to deliver real world software ("goldplating" and "over-engineering" are anti-patterns, you know). [b<]The point:[/b<] maybe AMD is lying and they did know of the issue and chose not to fix it. Or maybe they are too incompetent, or dishonest, or both. And maybe everything DaveBaumann said is true. I will never know so I will, at least, give them the benefit of doubt. Sorry for the long post. [b<]P.S.:[/b<] I currently own a GTX 670 card and have previoulsy owned a GTX 280 and a Radeon 9600 XT, just in case that is somehow important.

      • Silus
      • 7 years ago

      I was about to post just that. AMD disregarded all their problems because they were getting some high framerates and winning in some benchmarks. So it’s not that they decided to fix them…they were just caught with their pants down and now they have to own up to it.

      As for fanboys, that’s what they do. it’s not about reality, it’s about illusion and how their “favorite brand” must always be put in a good light, otherwise whoever did the review is biased and/or should re-run their tests.
      I mentioned it in the follow-up article that Scott did that included Windows 7, but TR should NOT give in to fanboy pressure. The first tests with Windows 8 were more than enough to prove the point. Still Scott made two more reviews, which obviously made his point even stronger but it shouldn’t be necessary.

      • Kaleid
      • 7 years ago

      Earlier measurements done by TR say that the latencies were not as common, or didn’t even happen at all. It certainly seems to be that with newer drivers the problems have increased.

      • Krogoth
      • 7 years ago

      Welcome to software development 101, my friend.

      You will be surprise how many “stupid, stupid” issues management is willing to slip under the proverbial rug if they can away it, especially if resolving the problems will delay the release by weeks if not months.

      • rrr
      • 7 years ago

      nVidia still is pretending problem of significant CPU overhead leading to major performance problems on dual core CPUs (and possibly quads, if LGA775 based) if paired w/nVidia cards doesn’t exist…

      Applicable to Windows versions newer than XP.

    • quasi_accurate
    • 7 years ago

    Good for TR to investigate the issues, although I’ve personally not noticed anything on my 7970. It’s possible I’m just not as sensitive to latency as some folks.

    • Ashbringer
    • 7 years ago

    What about users with 5000 and 6000 series cards? SOL?

      • MrJP
      • 7 years ago

      Taking the information at face value, you might not have the same problem in the first place.

        • HisDivineOrder
        • 7 years ago

        He did say it was GCN-related.

      • Philldoe
      • 7 years ago

      I can’t really say I’m seeing these issues with my 6870. I have played a lot of skyrim and tried to replicate the stuttering, but everything is smooth as warm butter for me. Maby it’s the setting I use. Everything all the way up with the exception that I do not use AA.

    • tviceman
    • 7 years ago

    In the meantime, it would be awesome of TR retested / benchmarked with older AMD’s drivers (before the 12.3 update) to see how the games they scored big frame rate increases in were affected in the latency department.

    • Bensam123
    • 7 years ago

    Huh, makes me wonder what changing powertune to +20% is doing then…

    It’s good that AMD is fixing this as well. I wonder how a bigger buffer will effect actual latency though (the amount of time it takes the system overall to shoot out a frame).

      • Ryu Connor
      • 7 years ago

      Placebo.

        • Bensam123
        • 7 years ago

        Derp, confirmation bias?

        You know this is like saying ‘NOU’ with psychology terms right?

          • Bensam123
          • 7 years ago

          For people that aren’t clever enough to get this. In this case a confirmation bias would be Ryu confirming that it was a ‘placebo’ based on the fact that what I’m describing currently doesn’t have a quantitative metric associated with it.

          But at the same time doesn’t prove anything because it’s entirely possible for a problem to be comorbid or there being more then one problem happening at the same time. Just the same as it’s possible this isn’t a confirmation bias (although likely is in this case).

          I appreciate popculture psychology becoming mainstream, but watching Bones or House doesn’t necessarily mean psychological terms apply wherever you see a possible example of them.

            • MadManOriginal
            • 7 years ago

            I’ve watched House. I think you’re obsessive about Core Parking and that conclusion is hard to deny. Point disproven!!

            • Bensam123
            • 7 years ago

            I believe fixated would be a better diagnosis doctor.

            • yogibbear
            • 7 years ago

            WHISKEY!

      • Arag0n
      • 7 years ago

      Remember we discussed about it? I told you whatever the reason was, AMD was likely to be able to solve it by driver update and consequently be a driver issue. Seems time will prove me right. Still, you may be right to point power tune, it may just be the handling of power isn’t properly done and increasing power budget hides the problem.

        • Bensam123
        • 7 years ago

        I’m unsure of the point of your post. It’s like you’re saying ‘I told you so’, therefore there is no problem with powertune, then go on to say that I may also be correct. You know it’s entirely possible for there to be more then one issue happening at the same time and consequently more then one right answer? Hence why you say ‘I told you so’ then contradict yourself immediately.

        There is a thread on the forums discussing powertune as well.

          • Arag0n
          • 7 years ago

          Note that I’m saying power tune may be related and increasing power budget hides the problem, but still, it can be solved via driver update and consequently is a driver issue.

          I agree more than one issue may be happening at the same time, but as you read AMD expects to solve them all via driver update.

          Don’t be mad because I do not deny that the problem may be power tune related!

          As long as the solution comes via driver update I will be the right one

            • Bensam123
            • 7 years ago

            Maybe powertune was the problem all along, but increasing the buffer size helps elevate this issue? Correlation doesn’t imply causation.

            It’s entirely possible for them to update powertune technology via a driver as well, if you can influence it by a slider bar in the CCC. I suppose I could lay down a blanket statement and be right too. Even a broken clock is right twice a day.

            • Arag0n
            • 7 years ago

            Lol, so you would prefer me to say I’m right, you were wrong? AMD will solve the problem via driver update… it makes the problem driver related, no matter which is the underlying reason for it.

            • Bensam123
            • 7 years ago

            You already said ‘you’re right, I’m wrong’. I’m simply telling you that making broad sweeping claims doesn’t make you a oracle. I could say the sun will rise tomorrow and also be right. It deserves absolutely no credit though and if there was a more pressing issue or even confounding issues, NEITHER is wrong… which in the case is true.

            Read the new article.

            • Arag0n
            • 7 years ago

            I was right and you were wrong regarding drivers being able to fix the issue. Still, I acknowledge you may be into at least one of the underlying issues with powertune.

            Seems you prefer people just to say, “i told you so”. Just because I was right about drivers being able to solve the issue does not make your research around the issue invalid.

            Is it so strange that despite I “win” the discussion we had before I concede you still may have a point? Don’t be a bad loser!

            • MadManOriginal
            • 7 years ago

            [url=http://xkcd.com/386/<]Duty calls![/url<]

            • Arag0n
            • 7 years ago

            Hahaha! So true!

            • Bensam123
            • 7 years ago

            Wow dude… The hypocrisy is beyond me.

            I say I think you’re a conceited douche for being petty enough to say ‘I told you so’ especially when there is more then one right answer, you say _I_ think that’s what I want you to say. I say there can be more then one right answer, you agree, then also say I’m wrong because you’re right… Which doesn’t make sense as the aforementioned scenario involves both of us being right. Then you try to brag, while attempting to be humble, how does that even work?

      • tfp
      • 7 years ago

      What does the +20% powertune impact? Really with a 20% increase in performance on a GPU it’s easy to expect that some issues with the drivers could be hidden by decreased latency in cache/RAM/Memory interconnects on the graphics card. HW can make up for performance issues in drivers or games up to a point.

        • DaveBaumann
        • 7 years ago

        20% on PowerTune does not, in anyway, translate to a 20% performance improvement.

          • derFunkenstein
          • 7 years ago

          Thanks for clarifying what it does.

            • tfp
            • 7 years ago

            Yeah thanks a lot Dave…

            • Bensam123
            • 7 years ago

            Thanks Dave!

          • Bensam123
          • 7 years ago

          .

        • Bensam123
        • 7 years ago

        Hitching and responsiveness (without having a better way to describe it). We’re trying to figure it out on a thread in the forums here:

        [url<]https://techreport.com/forums/viewtopic.php?f=3&t=83535[/url<] It's not a direct +20% increase in performance though as Dave said.

          • tfp
          • 7 years ago

          Power tune allows for dynamic clocking of the GPU and from the sounds of it you have increased the boundary of the TDP by +20%. This should allow the GPU to run faster under a heavy work load. So no it does not increase performance by a percentage set.

          This should be helpful for “smoothness”. Though looking at your graphics one can’t see by how much easily.

          First link on Google when searching AMD Powertune:

          [url<]http://www.amd.com/us/Documents/PowerTune_Technology_Whitepaper.pdf[/url<] [quote<]w. With this technology the GPU can minimize power during light workloads such as idle mode by enabling reductions in voltages, engine and memory clock speeds. In such cases, the GPU is in the lowest power state of voltage and frequency. .... Minimized Performance Impact on Applications which would Otherwise Exceed TDP AMD PowerTune technology allows TDP constrained GPUs to ship with greater nominal clock speeds in the highest power state due to the mechanism by which it handles applications which exceed TDP. Without AMD PowerTune technology, applications which exceed the GPU TDP are forced to lower power states (such as intermediate or lowest states) and pay a very steep performance penalty as a result of drastically reduced clock speeds and voltages. In the AMD PowerTune enabled GPU, the clock speeds in the highest state can be dynamically managed to hold the TDP budget in a way that was not otherwise achievable. The goal for applications that exceed the TDP budget is to maintain operation in the highest power state, but dial back on runtime power by modulating the high power state clock to keep the TDP range slightly below the absolute limit. This keeps the GPU away from the undesirable performance penalty of a forced state reduction arising from a thermal throttling event.[/quote<]

            • Bensam123
            • 7 years ago

            Yeah… I was talking about the impact on the user, not specifically what it does.

            I’m still working on trying to figure out how to quantify it. Hitches only happen so often and responsiveness is something that’s really hard to gauge. I’m still working on figuring out the same thing for core parking.

      • cynan
      • 7 years ago

      I guessed I missed the whole Powertune debate. I thought it simply permitted increased maximum current draw. I just assumed the slider was there to give the GCN products a competitive TDP for reviews at stock speeds. I thought everyone who bought a 7950 or 7970 automatically turned it to +20% (especially if overclocking).

        • Bensam123
        • 7 years ago

        Aye… From what I’ve seen monitoring GPUZ, it increases power draw in general, not just operating at maximum thresholds. You can actually see it impact FPS when running a program like Furmark or OCCP. However, trying to quantify this in real world examples is proving challenging, just like with core parking.

        Qualitatively there is enough evidence to suggest that it actually does impact overall, but you can’t apply a benchmark to that or offer exact scenarios where it will be beneficial to go +20 or disable core parking.

    • Waco
    • 7 years ago

    I’d rather have a driver that doesn’t crash when watching HD video after long periods of uptime. My 7750 has no home now thanks to it crashing my HTPC twice in as many days regardless of what driver I install.

      • brute
      • 7 years ago

      [quote<] driver that doesn't crash when watching HD video[/quote<] distracted driving is dangerous, crash or not, the potential to get hurt is pretty big

      • l33t-g4m3r
      • 7 years ago

      It’s probably because it’s a cheap card with a cheap chip, using cheap components. RMA it. I bet asic quality is on the low end, and the cooler isn’t powerful enough to keep the card stable. That’s what it sounds like to me, because those are the standard symptoms of any low quality semi-defective card, regardless of brand. Nvidia TDR’s and bumpgate anyone?

        • LoneWolf15
        • 7 years ago

        Cool story, bro.

      • vargis14
      • 7 years ago

      Waco my HIS Icooler 7750 does not crash watching any HD videos unless it is overclocked. One day after playing BF3 with a big overclock on my card, I started watching a video and wham driver crashed 3 times in a row until i reset the card to the stock profile.

      Running stock my 7750 does not crash watching HD videos…so if you have even a mild overclock set it back to stock and see what happens.

    • Hattig
    • 7 years ago

    Good work by AMD in actually looking at the problem. I can understand how the new architecture could lead to suboptimal drivers, and in this case I guess the memory management is stalling when freeing up internal structures within the GPU, leading to a slow frame, and then a fast frame or two because there’s now loads of free memory – and a better system could spread things out better. Memory management is not an easy thing, and I imagine on a GPU with 1000s of cores each with local and shared memories it is probably something close to a nightmare unless you are very familiar with the hardware.

    Should it have happened at all? Probably not, but software running on the metal like a GPU driver is hard, and you start off with something that works reliably, and then optimise the components that need it.

    • Ratchet
    • 7 years ago

    AMD is working on a detection algorithm that will detect when a high-speed camera is recording the screen and disable their benchmark-winning routine and enable their frame-latency winning routine.

    • My Johnson
    • 7 years ago

    I’d like to ding them for a driver conflict with Logitech mice. I’ve seen [url=http://forums.logitech.com/t5/Mice-and-Pointing-Devices/Performance-MX-bad-tracking-in-very-low-speed/td-p/649746<]others with the same issue[/url<] and I've narrowed it down to a conflict with AMD video drivers.

      • Deanjo
      • 7 years ago

      I haven’t experienced any slowdown relating to AMD’s drivers but I have heard a few problems that users are experiencing with Windows 8 and older Logitech mice.

        • mnecaise
        • 7 years ago

        Count me as one example of persons with Logitech and Windows 8 issues. The Logitech wireless keyboard mouse combo I bought does not work correctly with my new Dell laptop running Windows 8. Both the website and the box clearly say it’s Windows 8 compatible. Mouse driver seems to conflict with the Dell / Elan touchpad driver and the keyboard alphanumeric keys don’t work (weird). Driver locks up the laptop when you try un-pairing the keyboard or unplug the receiver. Logitech acknowledged the problem as theirs only after I swapped the keyboard (shipped at my expense) and then contacted them three more times. They acknowledged it but haven’t gotten back to me with a solution.

          • Deanjo
          • 7 years ago

          That doesn’t surprise me. I love logitech keyboard/mice/joysticks but it seems like everytime a new version of windows comes out I have quirky things happen with them or they drop features that never are re-introduced (such as loss of force feedback with the Wingman force, loss of DPI setting ability on a MX500, bluetooth never pairing etc etc). Fortunately they continue to work in other OS’s just fine and those are what I spend most of my time in.

            • mnecaise
            • 7 years ago

            Yeah, it works fine on my Windows 7 boxes here and at work. That piece of information, combined with my re-iterating it’s a brand new laptop with a fresh OS install and verifyably good USB ports is what convinced them it wasn’t my problem. I think they have issues with the new OS that they haven’t thoroughly ironed out.

            • My Johnson
            • 7 years ago

            I haven’t see anything like this across windows versions and I use Logitech exclusively. Another funky thing when it comes to mice though:

            My spouse has a new HP machine and it would constantly wake up after a few minutes, or so, from idle. After I switched out the included HP mouse (I use it on the AMD platform machine now) and gave her the Logitech gear, it stopped doing it!

            So I ended up fixing two problems.

            • Deanjo
            • 7 years ago

            If you stick with the base models you are usually fine. It’s when you purchase their higher end models that they seem to drop support for newer windows versions of windows and the only replies you will get from logitech are ones like these.

            [url<]http://forums.logitech.com/t5/PC-Gaming/Wingman-Force-joystick/m-p/939698[/url<]

            • MadManOriginal
            • 7 years ago

            Yeah, it’s just when you try to use their junky software that there’s problems. The basic functions of a straightforward mouse are fine, sucks for things like joysticks that might actually need drivers or software though, but I bet those would still work OK as standard HID inputs.

    • bthylafh
    • 7 years ago

    As a recent purchaser of a 7850, thanks for your work keeping DAAMIT honest.

      • sschaem
      • 7 years ago

      honest ?

        • bthylafh
        • 7 years ago

        ya srsly. Thank/blame TR for recommending their cards in the most recent system guide.

    • Deanjo
    • 7 years ago

    Hmmm, having a bit of deja vu here from AMD as this sounds a lot like the spin they gave with Bulldozer and scheduler issues. Guess we will see when TR gets to re-review.

      • DPete27
      • 7 years ago

      HA, that’s the exact same thing I was thinking when I read “software-related issues are potentially responsible for the Radeon’s latency issues.” It’s always somebody else’s fault isn’t it.

        • UberGerbil
        • 7 years ago

        Except in this case it’s their own driver software they acknowledge to be the problem and they’re pledging to fix it. So yeah it is similar to the Windows scheduler issue except for the significant way that is different. Now, if it turns out the updated software doesn’t have much effect, then we might talk about the similarity (and spin and deflecting blame). But so far AMD’s worst offense appears to be tardiness in reacting to the issue; the spin and protestations seem to have been mostly from their fans, not from the company itself (not surprising, since there doesn’t seem to be much in the way of marketing people left at AMD to do any spinning).

    • MadManOriginal
    • 7 years ago

    I wonder if smoothing the frame times will reduce the average FPS. That would be an interesting outcome, and it would mean that the vast majority of sites will see it as a negative driver change, and by sheer weight there will probably be a lot of complaining and posts on forums saying ‘don’t update to the new drivers!’ That’s all just speculation based on assuming that frame times were sacrificed in favor of average FPS, even if unintentionally, but it would put AMD in tough spot.

      • DaveBaumann
      • 7 years ago

      [quote<]That's all just speculation based on assuming that frame times were sacrificed in favor of average FPS, even if unintentionally, but it would put AMD in tough spot.[/quote<] This is not due to solely chasing average FPS. One of the software engineers looking into this contacted me to point out that for at least two of the titles where we've made improvements the average FPS should be improved, maybe by a reasonably large degree in some cases.

        • derFunkenstein
        • 7 years ago

        All the more reason to chase these down faster than a year after release, eh?

      • cynan
      • 7 years ago

      I would imagine that smoothing the frame times would not reduce average FPS as rule once the memory handling issue is sorted. Until that happens, you could be right.

      The cynical side of me suspects that these latency issues that have reared their ugly head recently could very well be due to AMD “cutting corners” in pursuit of higher average FPS numbers, though. The reason for this suspicion is that these latency problems did not exist to the extent they did a few driver releases ago (again, looking at the August 660 Ti review).

      I imagine it could have something to do with making too many top level optimizations to suit specific games or game engines as they are released in order to compensate for the underlying memory handling issue, but what do I know.. Until the situation becomes more clear (if that ever happens), I think AMD deserves some benefit of doubt. At least for now.

    • tbone8ty
    • 7 years ago

    this is awesome news for GCN card owners

    looking forward to seeing the 7950 regain its position

    congrats TR

    pop some of that left over bubbly!

      • tviceman
      • 7 years ago

      “Looking forward to seeing the 7950 regain it’s position”

      Why does this matter to you? Losing sleep over the whole ordeal?

        • MadManOriginal
        • 7 years ago

        epeen is on the line!

      • juampa_valve_rde
      • 7 years ago

      I knew that architecture had some extra horsepower left 🙂

    • maxxcool
    • 7 years ago

    @#$%ing awesome 🙂 How about that… one lonely tech site changed AMD’s perspective and brought about a positive change. GJ TR…

    • hansmuff
    • 7 years ago

    While I commend AMD for owning up to this and working on various efforts to improve the situation, I have to wonder what if anything they would have done without TRs excellent work.

    The architecture is about a year old, so I think they had enough time to address this post product launch but chose not to; surely it can’t take a year to implement better memory management.

    Does their QA not notice these things or were they ignored?

      • superjawes
      • 7 years ago

      [quote<]Does their QA not notice these things or were they ignored?[/quote<] I think the issue might be that [b<]only[/b<] TR seems to do the testing that fleshes out these issues. FPS measurements can easily be fudged to hide bad frames, so the metric might say that the 7950 is faster than the 660 Ti by a significant degree and the book is closed.

        • UberGerbil
        • 7 years ago

        Having worked in software development, I can tell you that it often takes a nudge from the outside to get internal red flags taken seriously. Management tends to prioritize things that get mentioned in reviews; if nobody is complaining about something, it can linger for a long time. Moreover when it’s a subjective impression it’s hard to even get traction on what you should be trying to do. Without metrics you can waste a lot of time fixing something that isn’t actually broken, or isn’t the thing that actually needs to be fixed. Without good tests that isolate the problem, it’s hard to prove you’re even making any progress.

        And something like a rewrite of an architectural foundation like the memory manager are the kind of thing that gets put off indefinitely. Even if it’s so obviously grotty that everyone has that cognitive itch whenever they look at it, the task is so disruptive to the schedule that it takes real courage to take the time to step back and do something like that. Usually you don’t have the resources to put a separate team on it (if they’re so good that you trust them with that, you can always make the case for putting them to work on something else) and the time devoted to it is time you could be optimizing a more obvious feature for the next game or benchmark or product or whatever. Good organizations have the discipline to just plan for periodic major maintenance like that, but not many organizations are that good — even when they aren’t wracked by layoffs and fleeing talent.

          • jpostel
          • 7 years ago

          Too true. You saved me the time to write just about everything you just wrote. Thanks for that.

          The other issue is that have been seeing software getting sloppy (sloppier?) due to the overwehelming power of the hardware. Why optimize something when you have cycles to spare? It’s not like they are working with 32k of memory, so [i<]the entire concept[/i<] of optimization is lost unless it is for a benchmark or game. It seems like something that would have been caught 10-15 years ago. I just re-read that last paragraph and I am officially "that old computer guy".

          • ermo
          • 7 years ago

          Well aren’t you the demotivational poster of the day to aspiring C.S. students-cum-software engineers…

          (see what I did there?)

          • superjawes
          • 7 years ago

          I take it by the upvotes that you aren’t the only person here who has QA experience 😆

          Yeah, I’ve dealt with some of that, too. In fact in my current job, some of our [i<]new[/i<] product development is driven by quality concerns. Even if you have a process that minimizes quality issues, you're still going to have others come up, and getting those fixed requires a lot of noise, and usually a lot of money (meaning losses if the problem goes unchecked).

        • nanoflower
        • 7 years ago

        Agreed. There’s always something that’s more important to get to and never enough people around to get to everything (even before AMD started letting people go.) So if their tests show that the FPS numbers are good enough no one is going to sign off on looking at some “minor” glitch that isn’t hurting the FPS numbers. I’ve been down that road before and it takes someone making the glitch a big deal to get management to sign off on taking the time to look into and fix the problems. In this case the glitch is real and TR is the someone that has helped to get AMD’s attention.

        Let’s just hope that what Baumann said is accurate and AMD can fix this glitch in their driver.

      • Firestarter
      • 7 years ago

      > Does their QA not notice these things or were they ignored?

      I guess most internal performance testing is done using (semi-) automated average FPS benchmarks for the current series of ‘hot’ games. I doubt that the QA people bother with performance testing though, I bet they concentrate more on the ‘does the game crash’ and ‘does it look right’ type of testing using a whole host of different hardware/software setups.

      • HisDivineOrder
      • 7 years ago

      They had bigger fish to fry. Like laying off employees, paying for awesome three-game bundle coupons, complaining when their CPU’s aren’t reviewed exactly the way they want you to by keeping you from doing a review unless exactly how they prescribe it to be, making TR editors sick and/or trying to kill them with ceiling attacks, and cancelling updated products that they’ve spent months advertising would overcome all the shortcomings of their existing products.

      They didn’t have time to go in and do something as simple as fix the memory management part of the driver when they were too busy getting more frames per second. You can’t expect them to make the more frames matter, can you? Duh?

      Feel bad for the CF guys. They’re the ones suffering the most from the stutter and they’re the ones that constantly complain about it without AMD doing a damn thing to fix it. Perhaps now CF will actually be worth using?

        • tjoynt
        • 7 years ago

        “…making TR editors sick and/or trying to kill them with ceiling attacks.. ”

        Please do elaborate: how did AMD “make TR editors sick” and WTF is a “ceiling attack”?

        Did AMD’s marketing people sneeze in their soup? Collapse a building on them?

        Normally I’d write it off as trolling, but I’m honestly interested/perplexed.

      • Sargent Duck
      • 7 years ago

      I work for a very large software company and have some friends in QA. All they do is run scripts using some programs (load runner?) which spits out the results and then they run some more scripts. They don’t actually monitor the scripts as they’re being run. It’s ENTIRELY possible that when QA’ing the 7×00 series, scripts were run, Average FPS’s were spit out and everything was good. QA people are not going to sit in front of each monitor to look for anomalies. Not when they’re testing hundreds of different scenario’s…

      I’m sure if this was Nvidia the exact same thing would have happened. Actually, since they have a marketing department they probably would have tried to sweep this under the rug. Remember the 5800Ultra? Rather than address the problem they just told gamers to go out and buy head phones (well, not exactly. But close enough)

        • UberGerbil
        • 7 years ago

        There’s many different kinds of QA. A major aspect is regression testing: making sure the new fixes or features don’t break or otherwise alter existing functionality. That is typically tested by running scripts and comparing the output with previous known good runs, looking for any discrepancies. That’s what it sounds like your friends are doing. It’s important, but it’s pretty rote. And it isn’t (or shouldn’t be) all there is to QA. There should be QA engineers working with the developers directly, to plan and implement new tests for new features or new ways to test existing features. And there should be ad hoc testing, to look for things that got missed (either because of holes in the tests or because the testing instrumentation itself hid them). None of which necessarily catches everything, because of all the permutations of hardware and software, and sometimes because something just doesn’t look like a problem (though it may turn out to be down the road to somebody else.)

      • ronch
      • 7 years ago

      They did notice it, brought it up to management, and were given the pink slips.

    • chuckula
    • 7 years ago

    TR deserves some credit for bringing these issues to light. If it weren’t for TR being the squeaky wheel, there wouldn’t be the same effort to get these issues fixed.

      • derFunkenstein
      • 7 years ago

      I imagine Scott finds this work to be both gratifying and frustrating. Gratifying in that they pushed AMD to do something to fix it, frustrating that AMD couldn’t apparently be bothered to fix the issues on their own in a timely fashion.

        • tfp
        • 7 years ago

        This is a new method for testing I’m not sure AMD even knew there was an issue before the new benchmarks came out.

          • derFunkenstein
          • 7 years ago

          From what I can tell that would require not playing games on the hardware.

            • nanoflower
            • 7 years ago

            I think it’s more complex than that. The FPS numbers were good and the game play seemed fine. The occasional glitch could have been coming from any number of places so I could see it being overlooked. Notice that everyone else that were reviewing the various AMD GCN based cards were overlooking the issue until TR brought it up. I believe some individual users had noticed the issue but it wasn’t an acknowledged issue until TR was able to provide hard data that showed the issue.

            • Pancake
            • 7 years ago

            derFunkenstein is more right. It’s not like the AMD driver programmers don’t have access to hooks and meters into their driver running games and aren’t profiling the hell out of it to find out where delays and bottlenecks are. At least, that is what they should be doing. I used to do it as a graphics driver programmer and it’s one of the more fun parts of the job installing a bunch of the latest games and seeing how they fly.

            • Captain Ned
            • 7 years ago

            A bunch of us had a long talk with Damage at the last BBQ and it was pretty clear that Damage & Co had broken new ground in the reviewing of vid cards. From that fireside and homebrew-infused chat (thanks JBI!!) I got the distinct impression that what happens here really matters to manufacturers, and matters because of the Joe Friday (Google it) attitude.

            • tfp
            • 7 years ago

            I think you’d be surprised what is tested for in general when it comes to software and drivers. No complaints equals no issues.

            Look how long graphics bench-marking has been going on and this is the first time people have really covered “smoothness” in a measurable way. Some of that discussion started back when HT and multi core CPUs came out but it was waved off as black magic. Now with everything having multicores, fast interconnects, and integrated memory controllers it harder to blame the CPU/MB/Ram for holding games back when it comes to general smoothness of play. Back in the AthlonXP/P4 era or before I don’t think this kind of testing would have been possible as there would be to many causes.

            • derFunkenstein
            • 7 years ago

            Then QA sucks, or management ignores it. Part of my job in software QA is to bitch/lobby for change. That’s just the way my employer works – it’s what’s expected of me. Someone at AMD clearly doesn’t care enough.

            • tfp
            • 7 years ago

            Driver testing is quite a bit different and with the release cycle that they have that don’t get a ton of time. I have seen teams test for large periods of time that don’t find issues because they run the same tests they have always run. With so many after release issues that are found by the community even if they did find this interally (and we have no proof) this might not have been considered device breaking might also get a low priority. Also it could have been qualified as an enchantment, the games are playable, which would push the priority down.

            Test teams “always” push for issues to be fixed, assuming they find something. However from what I have seen 50-60% of the defects reported are not defects or are duplicates of defects already reported. The larger/moe complex the product, the more defects, the more prioritization is required. At the end of the day if the users are not caring they will prioritize other they do care about first.

            Unless someone here works on the driver, test, or as a manager on the team you really can’t say they don’t care because you really don’t know what is happening in the group.

            • derFunkenstein
            • 7 years ago

            I don’t always get what I want either, which is why I suggested that maybe they’re just being ignored. Still, if they could figure out within a couple days of TR going public with their findings, it apparently wasn’t hard to find. Everyone has priorities, and AMDs priority – a year after the first cards shipped – just didn’t care enough to prioritize it. That’s fine, I guess, but this does not inspire me to upgrade to an AMD card.

            • derFunkenstein
            • 7 years ago

            Also worth noting that for a while (as recently as last spring) we were doing 3-4 week release cycles. A lot of times things would pass QA but not quickly enough to make it into the next official release. So I get withholding stuff to make sure it’s ready and not releasing it for another cycle or whatever, what I don’t get is if Dave is claiming this will make across-the-board improvements to performance why wasn’t this started before TR made a big deal? Isn’t “good performance” the primary feature?

            • tfp
            • 7 years ago

            Before it had “good performance” and over game stability issues or bugs they always have in release maybe not.

            Also as I said somewhere else, AMD and Nvidia make releases with general improvements that help “all” games. It’s hard to say they haven’t started addressing these issues already without doing a comparison of the different driver release that have had the large general improvements.

            • derFunkenstein
            • 7 years ago

            I’m just going by what Dave wrote, that initially they weren’t going to bother, but now they are. And they’re finding it makes (in some cases) big improvements. Not to mention they’re finding it relatively quickly given the time of year this is all happening. It all adds up – to at least appear – to be low-hanging fruit that they were just passing by. And now Dave is here spinning it into “look at what we’re doing for our customers!”

            • tfp
            • 7 years ago

            Do you mean this?

            [quote<]Originally Posted by Dave Baumann Ehh, don't mind me. I've been around the internets long enough. While I like the scientific application of the analysis some work does need to be done to understand the thresholds. While I cannot deny there are spikes (the driver team have spent a lot of time analyzing where they are coming from and smoothing them out!), I can't say that in the games I've played they have been noticeable, likewise I do not see end user feedback for this type of issue (except, maybe outside of a few known problem titles).[/quote<] I don't see anywhere in that thread were Dave said they were not going to look at these issues before. He did state that the fixes are going to be harder than expected to do as well. Also the "easy" change was the buffer he lists a number of other items that need to be investigated. I also don't see any time frame for when they did all of the analysis, it could be done in the last month or two. They have fixed general issues for most if not all games in the past and they are doing this again. I think people are making way to many assumptions on how AMD has handled testing and fixing of bugs in the past.

            • DaveBaumann
            • 7 years ago

            Curious, where did you read [i<]any[/i<] of that? First off, I mentioned that we have a memory management rewrite underway for orthogonal reasons and it was at a point where we could just try it and see if it makes any improvements; it transpires that [i<]in some cases[/i<] it has helped improve things. This has raised the priority to trying to get this through earlier; that doesn't mean that we "weren't going to bother", otherwise what would have been the point of investing the development time in the first place? Again, the memory manager rewrite was being done for other reasons, and in fact there was no indications from anywhere (including TR's prior reviews) was this an area that needed to be addressed. Secondly, the memory manager change is not "magic bullet". The quote Scott originally highlighted from me was that these types of things can occur from multiple different areas of the driver - for instance the BL2 improvement primarily comes from resizing a buffer, not related to generic the memory management at all.

            • bandannaman
            • 7 years ago

            It’s hardly surprising that AMD asks their engineers to spend time optimizing the things that get the most “traction” in the market. If people believe that average FPS is what matters, and they make purchasing decisions based on that, then by gum that’s what AMD (or any company) is going to spend time on. Saying their QA sucks isn’t really accurate; no company has unlimited resources to allocate.

            Same deal here with the new frame latency awareness — if people believe that worst-case frame latency is what matters, and they make purchasing decisions based on that, then by gum THAT’s what AMD is going to spend time on. As they should.

            Kudos to TR. This is real value and real journalism.

          • Silus
          • 7 years ago

          What ? This method MUST be part of NVIDIA’s/AMD’s test labs. Memory management is surely part of any QA assurance for GPU drivers.

            • tfp
            • 7 years ago

            First they listed memory management as one issue causing some of the slow downs not all. Second the cards over all performance (FPS) is competitive. This issues found by TR is only one bigger issue in a sea of defects AMD tries to address with it’s monthly driver releases. Check the release notes, they are not sitting idle on defect fixing. No software team has unlimited resources.

            Also it’s hard to save if they haven’t already been fixing issues like this. AMD and Nvidia regularly report blanket gains in performance across driver releases. Unless someone goes back and checks the impact of those general performance enhancing releases for improvements in frame delivery consistency no one in the discussion should be claiming AMD or Nvidia isn’t already trying to address these issues even before TR started testing differently.

        • NeelyCam
        • 7 years ago

        Things move slowly; you can’t expect an overnight fix.

        Nonetheless, this is an impressive acknowledgement of the power and importance of high-quality tech media. This reaffirms TR’s place in the upper echelon of tech websites, hopefully bringing in more readers, ad revenue etc.

      • Kaleid
      • 7 years ago

      Now we need some spotlights on the DX9 flickering issue.

      [url<]http://forums.guru3d.com/showthread.php?t=368290[/url<]

      • lilbuddhaman
      • 7 years ago

      Scott’s picture is likely on a dartboard at AMD’s Radeon software division.

        • LoneWolf15
        • 7 years ago

        I doubt it. This gives AMD an opportunity to ensure the next-generation of cards with GCN don’t have this issue, while resolving the 7xxx-series issues at the same time.

        Marketing may not like it, but hardware/software engineers often appreciate feedback that helps them improve product.

          • lilbuddhaman
          • 7 years ago

          And how do you know a bunch of engineers didn’t already know they had issues with frametime?

          Are you to tell me that the guys who designed, manufactured, and programmed these chips didn’t have an intimate knowledge of the delays in time from frame to frame (along with everything else in between)?

          Are you REALLY trying to say that?

          What TR has done isn’t to enlighten anyone at AMD, it is to tell them “guess what, we’re looking at a little deeper these days, time to step up your game”

            • MadManOriginal
            • 7 years ago

            With the number of truly obvious bugs that ship in drivers and software in general, it’s very possible that the engineers weren’t aware of this particular problem.

            • anotherengineer
            • 7 years ago

            “Are you to tell me that the guys who designed, manufactured, and programmed these chips didn’t have an intimate knowledge of the delays in time from frame to frame (along with everything else in between)?

            Are you REALLY trying to say that?”

            Being an engineer, I am going to say yes.

            The people who design, manufactured and programmed are typically all different people and a lot of times under different roofs, and sometimes completely different countries.

            They also can’t perdict other vriables like virus scanners, etc. From Scott’s own data, even Nvidia has the same issues albeit to a lesser extent most of the time.

            Another issue that arrises is the amount of games that are released after a gpu is released to market, usually this requires more driver modification to get better performance.

            I cringe at the thought of how many lines of code is in todays latest ATI & Nvidia gpu drivers.

            • Ringofett
            • 7 years ago

            I don’t come from an engineering background, but still, I for one consider it a miracle of modern science, technology and engineering that these products exist, exist in vast quantities*, and asides from some bugs generally work extremely well. GPUs and CPU’s are, or at least appear to be, extremely complex devices, created on the some of the smallest scales mankind can manage in bulk.

            That these creations take vast teams, spread over vast areas, and that no/very few single persons have some omniscient knowledge of *everything* about the product doesn’t surprise me at all.

            * This point can’t be understated. Sure, one-off things can be created and vast expense, wouldn’t be shocked if some unknown government agency has impressive kit funded by some billion dollar slush fund, but these things use cutting-edge technology at a price point any penny-pinching teenager can afford.

            • lilbuddhaman
            • 7 years ago

            Well I still find it hard to believe that their team didn’t already have some internal knowledge of frame times. And I feel as they have an internal threshold that they aim for (say sub 50ms), which is a guideline, not a requirement. Now with this being public knowledge, that threshold will be much lower, and those engineers now have a harder job.

            That’s just how I see things over there, /shrug.

      • ermo
      • 7 years ago

      “Squeaky wheel gets the kick!”

      — Minsc

      • Silus
      • 7 years ago

      I would say they deserve all the credit. They pursued the issue with 3 different reviews/articles, after the uproar done by the AMD fanboys.

      • My Johnson
      • 7 years ago

      Yeah, but didn’t TR credit AMD back in the day for giving them the tools to determine frame latency?

      • MadManOriginal
      • 7 years ago

      ’69’ thumbs up…uh, wait, did I just say that?

        • sweatshopking
        • 7 years ago

        ha! I RUINED IT

    • gmskking
    • 7 years ago

    Things just are not looking good for AMD lately.

      • brute
      • 7 years ago

      i dont recall things looking good for AMD for quite some time now

Pin It on Pinterest

Share This