Frame-pacing driver aims to revive the Radeon HD 7990

When the Radeon HD 7990 first hit the market back in April, it didn’t get the sort of reception one might expect for a graphics card that could easily claim to be the world’s fastest. The trouble it encountered had been brewing for quite a while, as PC gamers became increasingly aware of a problem known as microstuttering that plagues multi-GPU configurations like AMD’s CrossFire and Nvidia’s SLI.

The problem has to do with the fact that frames are rendered in interleaved fashion between the two GPUs. That division of labor ought to work well, at least in theory, but the GPUs can go out of sync, dispatching and delivering frames at uneven intervals. In the worst cases, the two GPUs might dispatch or deliver frames at practically the same time. The second frame doesn’t really offer any visual benefit when that happens, although its presence will inflate FPS averages.

Since the 7990 is a CrossFire-on-a-stick product with two GPUs lurking under its expansive cooler, it’s prone to microstuttering problems just like a pair of video cards would be.

Although we’ve been stalking this problem for a couple of years, we were ready to pinpoint it when the 7990 arrived. We conducted a deep investigation, capturing each frame of animation produced by the card and converting them into detailed, frame-by-frame benchmark results and slow-motion videos. We concluded that two Radeon GPUs were often no better than one, visually speaking—and that AMD’s thousand-dollar Radeon HD 7990 wasn’t really any better for gaming than a much cheaper video card with a single GPU.

There was a ray of hope, though. Our results showed that Nvidia had largely mitigated the worst effects of microstuttering in SLI using a feature called frame metering. Frame metering is what it sounds like: the timing of frame delivery to the display is modified so that new frames are displayed at more even intervals. Generally speaking, this trick works, producing perceptibly smoother animation. AMD supplied us with a prototype driver to use with the 7990 with a similar feature called frame pacing, and we found that it seemed to work pretty well, too. Trouble was, the prototype was based on an older branch of AMD’s driver code, and it was in no state to be distributed to 7990 owners.

Now, at the start of August, has AMD finally released a beta driver with frame pacing to the public. You can grab the Catalyst 13.8 beta here from AMD’s website. This driver comes from the current code branch and adds a few other new features, including application profiles for OpenGL-based programs, full support for OpenGL 4.3, and a 3-8% improvement in Metro: Last Light performance. We’ve been playing with an early build of this driver for several days, and we’ve captured some detailed, frame-by-frame results to illustrate what it does.

This first beta driver does have some limitations. Frame pacing only works with games that use DirectX 10 and 11. AMD tells us DirectX 9 support is coming later, but the company isn’t yet willing to say when that will be.

Also, frame pacing support is limited to resolutions of 2560×1600 and below. Those with 4K monitors—who might be, you know, pretty likely Radeon HD 7990 buyers—will have to wait. Frame pacing at 4K resolutions is “in the works,” but we don’t know when to expect that, either.

We haven’t yet had time to try it, but frame pacing should work with cards other than the Radeon HD 7990—and not just the Radeon HD 7000 series, either. The Radeon HD 6000 series should benefit from frame pacing, and even older cards might, too. AMD couldn’t give us a generational cutoff point for frame pacing support. Maybe that pair of Radeon X850 XT cards I’ve been saving will be rejuvenated!

Speaking of things we don’t know, AMD didn’t have any interest in telling us exactly how its frame pacing mechanism works. We expect that it adds a small amount of latency at some point in the frame delivery process in order keep frame-to-frame display intervals fairly consistent. Beyond that, we don’t have specifics about how this goal is achieved. We do know that frame pacing doesn’t require per-game profiles. Generally, just turn it on, and it works.

In another bit of happy news, frame pacing is enabled by default for CrossFire setups in the Catalyst 13.8 beta. Users may disable it if they choose via the Catalyst Control Center setting shown above. Earlier this year, after we’d pinpointed the microstuttering problem using FCAT, AMD told us frame pacing would be an option, but not the default, because frame pacing can add a very small amount of lag to the user input-visual response loop. Once we saw how poorly the 7990 performed without metering and how effective the prototype driver was, we lobbied AMD to change its policy. Surely most users would benefit from the smoother animation with frame pacing enabled. Fortunately, it looks like AMD eventually agreed with that assessment.

I’m also pleased to report that I didn’t perceive any increases in input lag with frame pacing enabled via the Catalyst 13.8 beta. As for the rest of the experience, well, let’s have a look.

The results on the following pages were collected using the same setup as in our original Radeon HD 7990 review. The only exception is that we tested the Radeon HD 7990 with the Catalyst 13.8 beta, both with and without frame pacing enabled. We’re comparing against older drivers for the GeForce cards (and for the Radeon HD 7970), so be aware that they may have a bit of a handicap. However, our primary focus will be on the 7990 and frame pacing.

Tomb Raider

Here’s a video of our test run in Tomb Raider, along with screenshots showing the quality settings used. At these settings, especially with 4X supersampling, this game is intensive enough to stress any of these high-end graphics solutions.

Ok, here’s where things get a little complicated. The plots below show frame-by-frame rendering times for each graphics card, as captured with two different tools: Fraps and FCAT. Fraps measures when a frame is finalized by the game engine and handed to the DirectX API, relatively early in the rendering process. Fraps results get us as close as possible to the game engine itself, whose internal simulation timing determines the content of each frame—i.e., where objects are positioned in the game world as they move through time. FCAT analyzes the contents of the screen to tell you exactly when a frame reached the display. Both tools are useful for determining animation smoothness, because an interruption in either frame dispatch or frame delivery can disrupt the illusion of motion.

Fraps and FCAT results tend to correspond pretty closely, generally speaking. They don’t correspond exactly here because we weren’t able to use both tools at once. The results you’re seeing in the plots come from a single test run, conducted manually, with each tool. (We did test three times per tool per card, though, for FPS averages and such.)

Frame time
in milliseconds
FPS
rate
8.3 120
16.7 60
20 50
25 40
33.3 30
50 20

The button labeled “HD 7990” will get you results from the Catalyst 13.8 beta with frame pacing disabled, and the button labeled “HD 7990 paced” will show you how frame pacing changes things. Flip back and forth between them, and you can see a dramatic difference. Without pacing, the 7990’s results from Fraps and FCAT both are all over the map. Look at the zoomed-in plots below the main one, and you can see the telltale see-saw pattern that comes from microstuttering. The pattern is very pronounced, too, with frame-to-frame intervals ranging from three to 40 milliseconds and back again.

With pacing enabled, the 7990’s Fraps results still have an intermittent high-low latency pattern, but it’s not a constant problem. The FCAT plot becomes much, much smoother. Tomb Raider appears to advance its simulation time using a moving average or some sort of set interval, rather than simply sampling the current frame time. We’ve found that those relatively small (typically under 40 ms) intermittent variations in the Fraps plots don’t translate into perceptible hiccups or slowdowns. Since the game works that way, I’d give more weight to the FCAT results in the charts below.

If you only care about FPS averages, then the good news here is that frame pacing doesn’t appear to reduce the number of frames rendered over time. If you care about smooth gaming, then the latency-oriented 99th percentile frame time may be the more helpful metric. Frame pacing vaults the Radeon HD 7990 from the middle of the pack to the front ranks, just ahead of its direct rival, the GeForce GTX 690.


Here’s a look at the entire frame latency curve. As you can see, without pacing, the 7990’s curves from both Fraps and FCAT start out low and escalate steadily, matching or crossing the curves from a single 7970. With pacing enabled, the 7990’s latency curves are flatter, and frame times are dramatically lower in the right half of the plot. That avoidance of high-latency frames should translate into smoother animation, so the frame-paced results are much more desirable.


We can look at the amount of time spent working on high-latency frames in order to get a sense of the “badness” in each card’s performance. The more time spent on high-latency frames, the more likely you are to notice a problem. For Tomb Raider, we can ratchet our threshold of badness down to 33 milliseconds, the equivalent of 30 FPS or two display refresh intervals on a 60Hz monitor, and the 7990 with frame pacing still aces the test. FCAT tells us the 7990 with pacing only spends two milliseconds beyond the 33-ms threshold. That’s pretty fluid.

Those are the numbers, and hopefully you have a sense of what they mean. The impact of frame pacing is palpable. When testing this driver, I first played each game with frame pacing enabled, and then I disabled pacing and played the same section again. The difference in the smoothness of the animation was comical. I laughed. With frame pacing, the 7990 delivered a nicely playable experience. Going without pacing after having used it was painful—easily slower and chunkier.

Since we capture each and every frame during our FCAT testing, we can convey a sense of the experience with some videos. I’ve slowed things down to half speed, from 60Hz to 30Hz, so the differences between the solutions should be even more evident. Please do realize we’re using YouTube here, so it’s kind of like filtering espresso through a sock, but I’m hoping you at least get some of the flavor. Ignore the trippy overlay colors, which are used by FCAT to mark each frame. Watch for things like the fluidity of the ground passing under Laura or the motion of her arms and hair, and I think you’ll notice the gap in fluidity.

First, here’s a side-by-side video showing the impact of frame pacing:

Here’s just the 7990 without frame pacing:

Here it is with frame pacing:

And now the GeForce GTX 690:

My hope is that these videos illustrate what we’re talking about reasonably well. See what you think.

Crysis 3
Crysis 3 easily stresses these video cards at its highest image quality settings with 4X MSAA. You can see a video of our test run and the exact settings we used below.

The 7990 is a wreck without frame pacing, especially when it when it comes to delivering frames to the display. The jitter pattern is pronounced, with punishing frame times of 60-70 milliseconds on the high side. With pacing turned on, the jitter dissipates both in Fraps and FCAT, leaving the 7990’s latency plot looking similar to a single-GPU solution’s—only with lower frame times.

Meanwhile, the GTX 690’s Fraps results show some stuttering that’s smoothed out by the time the frames reach the display. I understand Crysis 3 uses a moving average of the last 250 ms worth of frame times in determining how much to advance its simulation from frame to frame. As a result, that variance in the GTX 680’s Fraps results shouldn’t have too much of a negative impact.



Frame pacing makes everything better. The 7990 trades blows with the GeForce GTX 690 for the top spot in each of our latency-focused performance metrics.

Once again, here are some videos illustrating the different cards’ performance.

The 7990 without frame pacing:

With frame pacing enabled:

Now the GTX 690:

Finally, the Titan:

Yep.

Sleeping Dogs

The 7990’s stutter pattern should be obvious by this point, and you can see how frame pacing mitigates the problem. Sleeping Dogs is yet another game that appears to use a smart timer to advance its simulation. Those occasional “heartbeat” sequences in the Fraps results, where you’ll see some frame times north of 80 ms, don’t tend to translate into perceptible hiccups as long as the FCAT line is relatively smooth.



The trends in the numbers are familiar by now, but it’s worth pointing out that a single Radeon HD 7970 GHz Edition card is superior to a Radeon HD 7990 in many of our latency-oriented performance metrics when frame pacing is disabled.

Let’s look at some videos. First, the Radeon HD 7970:

The 7990 without pacing:

The 7990 with pacing:

And the GTX 690:

Once again, I think the difference is apparent. Here’s hoping the videos convey that impression.

Conclusions

If you have any sort of graphics setup that relies of AMD’s CrossFire technology, you’ll want to download the Catalyst 13.8 beta and install it right away. Since AMD has wisely decided to enable frame pacing by default, gamers should see the benefits of that feature in any DirectX 10/11 games immediately. Based on our benchmark results, the slow-mo videos we’ve captured, and our own seat-of-the-pants impressions, I think we can say with confidence that AMD’s frame pacing solution appears to work just as well as Nvidia’s SLI frame metering. The Radeon HD 7990’s frame production and delivery results look much more like what you’d get from a single fast GPU—and that’s exactly the behavior you’d want.

AMD still has work to do: extending frame pacing support to 4K displays, DirectX 9 games, and hopefully OpenGL, too, eventually. Also, we refrained from testing Far Cry 3 for this article because of a lingering problem with that game. AMD tells us a fix is coming. This is just the first beta driver release with this feature, so these limitations should come as no shock.

I get the sense that AMD hopes this driver release will restore some of the luster to the Radeon HD 7990. Given how well it seems to work, I suppose that could happen. We don’t often recommend graphics cards that cost nearly a grand—although, hey, this Sapphire model is down to a paltry $920. Dual Radeon HD 7970s will cost less and perform the same. However, I still like a lot of things about the 7990. We’ve only tested three games so far with the new driver, but at least when frame pacing is an option, the 7990 can plausibly claim to be the world’s fastest graphics card. AMD will throw eight (quite decent) games in your face if you buy it, which should be a disconcertingly pleasant experience. And I still think the 7990’s super-quiet cooler is a revelation, although less of one after THG uncovered some overheating issues in smaller enclosures. You’ll need a beefy case with good cooling to house one of these beasts. Still, for the right setup, the 7990 could be a bragging-rights goldmine and a significant source of your daily requirement of buttery smoothness.

Before you plunk down nearly $1K on this glossy black-and-red monster, though, consider that multi-GPU cards like the 7990 are very much hardware-software combo solutions. The results on the last few pages have given us a nice object lesson on that front. Making two GPUs work well together is a constant effort that requires frequent software releases to support new games. With the Cat 13.8 beta driver release, AMD appears to be well on the road to fixing this particular problem.

But realize that literally years have passed since the first forums and tech websites, particularly in Europe, started talking about microstuttering. Nvidia claims to have implemented its frame pacing solution back in the G80 generation of cards, a claim we still need to confirm. We know AMD did nothing back then. We first talked to AMD about microstuttering nearly two years ago, for our first Inside the second article. We illustrated the problem with frame-by-frame plots at that time. Evidently, AMD did nothing. Only when we irrefutably pinpointed the problem at the display level with the aid of the FCAT tools supplied by AMD’s competition did Team Radeon begin taking microstuttering seriously. Work on a driver fix finally began. With this issue exposed for all to see, the firm went ahead and released the Radeon HD 7990 to consumers when a fix was months away—and marketed the product on the basis of FPS scores, a metric multi-GPU solutions with microstuttering essentially inflate artificially.

There is a lingering question of trust here. If another issue crops up with multi-GPU configs that impacts gameplay but isn’t easily quantified, will AMD do the right thing for its customers? I’m afraid I can’t make any plots or high-speed videos to answer that one. You’ll have to make that call yourself.


I make no attempt to meter the timing of my output on Twitter.

Comments closed
    • RachelHarg00
    • 6 years ago
    • MaryGreen06
    • 6 years ago
    • SelenaChao6
    • 6 years ago
    • JillHoff9
    • 6 years ago
    • GTVic
    • 6 years ago

    What is with the GTX 680 poor performance in the Tomb Raider – Time spent beyond 33.3 ms?

      • Damage
      • 6 years ago

      The GTX 680 is averaging 22 FPS. Every single frame it renders takes longer than 33.3 ms. You need two GK104s in order to run this game fluidly with 4X supersampling at 2560×1440, which is very intensive.

      Now, a GTX 680 will run this game well with, say, FXAA. But not 4X SSAA.

        • USAFTW
        • 6 years ago

        4xSSAA , 2560×1440, That’s 10240×5760. Effectively two 680’s only get a total of 2 gigs, would that be enough to drive that?

    • kamikaziechameleon
    • 6 years ago

    Does this work for regular Crossfire cards?

      • ronch
      • 6 years ago

      What do you mean by ‘regular’?

        • LukeCWM
        • 6 years ago

        I believe he means not-Crossfire-on-a-stick. And I believe the answer is yes.

    • Laykun
    • 6 years ago

    The question is, why didn’t AMD do this YEARS ago. A year ago (probably more now) I tried 3 7970 cards in CFX with poor results. Broken Tri-fire support and frame delivery problems. I ditched the cards and replaced them with 2 GTX 670 4GB cards (for nvidia surround 5670×1200). If AMD had their shit sorted back when I got the 7970s I’m sure I’d be a much much happier customer.

    That’s not to say the GTX 670s are bad, but I bet I could have gotten better performance out of 3 7970 cards and had a much less harrowing experience (returning 3 7970s in NZ is not an easy feat, it took months).

    Oh well.

    • CBHvi7t
    • 6 years ago

    I did not understand the issue.
    afaik the game on the CPU sends objects and textures to the GPU. Most of the time it will update their positions and then trigger the processing of the next frame.
    In the case of two GPUs the CPU will send the same objects to the cards but trigger the picture generation with different positions at different times. The rendering might take twice as long as the frame period.
    How will this lead to stutter that would not be there in the case of one GPU? Why would one frame finish much earlier than the other?

      • sparkman
      • 6 years ago

      The game would run at 1,000,000 frames per second if the card could handle it, but no card can.

      So yes while the game at a high level controls the generation of each frame, it normally does so only when the card signals that a new frame can be displayed, and much of the low level scene generation is handled by the card itself.

      Ex: the game says “draw a bunch of triangles at this list of locations” then the card goes off and does that and after the triangles finish drawing then the card will display that new frame.

      With microstuttering, one possible failure mode is that the game says “can we draw another frame yet?” and GPU1 answers “yes”. Then the game says “draw these triangles” and GPU1 starts working on them. Then the game says “can we draw another frame yet?” and GPU2 says “yes” and the game sends another almost-identical batch of triangles for GPU2 to draw. So for a whole 2 frames we have a very high theoretical frame rate.

      But now we have run out of GPU’s. The game says “can we draw another frame yet?” but both GPU’s on the 7990 are busy finishing the earlier frames. So the game blocks, waiting for a GPU to free up. Eventually GPU1 finishes, followed shortly by GPU2, meanwhile the game wakes up (and finds that it’s simulation timer has advanced far while the game was sleeping), and another two nearly-identical frames get fired off.

      These pairs of nearly-identical frames are causing the reported fps number to double but really almost no more additional information is being displayed to the human player’s eye. It’s as if there is only 1 GPU, but you paid for 2.

        • Waco
        • 6 years ago

        Very good concise explanation.

    • 0g1
    • 6 years ago

    Revive the 7990? If they want to do that, they should make it less than two 7970s which cost $620. Right now the 7990 is $820.

    • ronch
    • 6 years ago

    I’ve always preferred AMD graphics over Nvidia for the last 10 years or so. Now I admit I have a bias towards AMD but in this case it’s not just fanboyism — they just generally offer better ‘raw’ performance for the money (where I live, anyway) so I always tend to pick AMD. When the frame latency issues cropped up here on TR that kinda soured my perception of AMD and affirmed the age-old belief that their graphics drivers are inferior to Nvidia’s to many people. Now it’s good to see that they have at least closed the gap. I don’t think their latest beta is quite as polished as Nvidia’s frame pacing mechanism yet, but at least they’re on the same ballpark. I hope they take this as a learning experience that it’s not just about winning FPS tests, but ultimately, it’s about providing a tangibly better gaming experience than the other guys.

      • Bensam123
      • 6 years ago

      You know Ronch… You come off as a really confused fanboi, perhaps even a convert that still somehow thinks he’s a AMD fanboi.

        • ronch
        • 6 years ago

        Not confused, Ben. I like AMD but I’ll call it as I see it. And I’ll be the first to admit what they’re failing to do instead of covering up for them. How is that confused?

        Also, 10 years ago AMD’s graphics division was ATI. Never was an ATI fan and yet I almost always chose ATI because at the time of purchase they always seemed to offer better products for the money.

          • Bensam123
          • 6 years ago

          Nah, dude Chuckula totally mind fucked, then converted you. I’ve seen you make these really awkward posts for the last few months since he did it. Stop acting like a fanboi, when you aren’t one… It comes off as ungenuine and sarcastic. Not just that, but it seems like you don’t even support AMD most of the time and just choose to rail on them in a pleasant way while calling yourself a AMD fanboi.

          [quote<]And I'll be the first to admit what they're failing to do instead of covering up for them. How is that confused?[/quote<] Because it's not what fanbois do. You don't need to 'admit' to AMD failing either. It sounds like you somehow think you're responsible for the actions of the company.

            • flip-mode
            • 6 years ago

            Dude, you’re going after ronch in a very personal way which is clearly against this website’s comment policy.

            • ronch
            • 6 years ago

            No harm done. I don’t take things too personally and instead try to remain rational. Well, unless a particular gerbil is really getting very offensive.

            • flip-mode
            • 6 years ago

            A little bit of encouragement to abide by the rules could go a long way toward improving behavior.

            • Bensam123
            • 6 years ago

            Once again, tell that to your butt bud Chuckula… Oh wait, you chose to look the other way when someone you ‘like’ is doing it. Moral high ground my foot. Get off your high horse before I drag you down.

            • flip-mode
            • 6 years ago

            I don’t read every post in the thread. Any one of us can lead by example. If you think someone is being out of line then you should politely tell them or else report them. It would be great if Scott would appoint a front page mod. We don’t need to be angelic, but this stuff does get out of hand and the quality of the discussions most certainly suffer for it.

            Edit: I searched the comments and there are five occurrences of the word “chuckula” – 1 occurrence from the only post of his and he didn’t do anything inappropriate, 4 occurrences were you referring to him, usually to say something nasty about him.

            • Bensam123
            • 6 years ago

            When you participate in said threads I can only assume you know what’s going on in them.

            You searched the comments in this thread? So what I’m citing which happened from Nov to about April applies to your logic? I eventually told him to fuck off. I don’t reply to his threads or anything he writes anymore with some really rare exceptions and he knows it. He plays things like a politician, he stabs you in his back, you’re only his friend when you agree with him or hate on AMD (hence the Ronch thing I’m pointing out here), he fights dirty, he doesn’t look to make logical arguments, but simply make fun of or berate whoever he’s arguing with. Heck it got someone banned because he nagged them so far.

            [url<]https://techreport.com/discussion/24879/intel-core-i7-4770k-and-4950hq-haswell-processors-reviewed?post=736962[/url<]

            • ronch
            • 6 years ago

            Hey, chill out, dude. Why are you so riled up? Flip-mode is right. I don’t know why people say all sorts of nasty things just because they’re hiding behind a user name. How would you like to talk the way you’re doing now [b<]in person[/b<] at the next TR BBQ? No, of course you wouldn't do that. A little respect and camaraderie goes a long way even on the Internet.

            • Bensam123
            • 6 years ago

            Saying you AREN’T a fanboi isn’t nasty or a personal attack. Heck how can it be a personal attack when calling someone a fanboi in the first place isn’t one? I’m being very strict handed here because Flip is accusing me of personally attacking you and citing forum rules, which is something that is quite punishable. It should be dealt with in a serious fashion.

            I’m going out of my way to point out that he not only hasn’t cited this before when I’m under the gun, he has also participated in such personal attacks with Chuckula. Hypocrisy at it’s finest.

            And that hurts. I actually had considered attending the TR BBQ. I’m not saying stuff like this to be mean, I’m doing it because you appear to be his star lapdog which he toys with whenever he feels like it and I find it quite disgusting because you don’t seem to have any idea it’s going on.

            • ronch
            • 6 years ago

            Well, I’m used to heated discussions here on TR, and in other forums as well. Yeah, I can sense what sort of guy Chucky is, but it doesn’t bother me. Come on, this is just the TR forums. I don’t know any of you guys personally. Why should i get all riled up? If anything, I use these discussions just to make my mind work sometimes.

            So, in the end, I’m not talking to anybody here. I’m just talking to my FX. It’s like this..

            ronch <—> AMD FX <—> TR gerbils.

            It’s like I’m the CPU, the FX is the chipset, and you guys are the hard drives, or whatever. The CPU only talks to the chipset. Haha. Whatever.

            • Bensam123
            • 6 years ago

            Fanbois aren’t overly rational, that’s the thing.

            • ronch
            • 6 years ago

            You must be referring to Apple fanbois. No, AMD fanbois are quite a bit different.

            • Bensam123
            • 6 years ago

            Example A, Spigzone…

            • Bensam123
            • 6 years ago

            Tell that to Chuckula dude and his six months of BS I put up with. I’m not going after him personally, I’m calling him out on not being a AMD fanboi, it in fact seems he’s almost the opposite.

            • ronch
            • 6 years ago

            No, I don’t think he’s an AMD fanboi. Not sure he’s an Intel fanboi either. But it doesn’t bother me at all. Why would it?

            • Bensam123
            • 6 years ago

            Reply to Flip about you, not Chuckula.

            • ronch
            • 6 years ago

            Jeez, Ben. What are you trying to do, make me scream at the top of my lungs that I’m an AMD fanboi? Or prove that I’m not a fanboi? Well, then. Let me explain. I like AMD and you may say that I’m sort of an AMD fanboi. But I’m a different sort of AMD fanboi. Unlike most AMD fanbois I try to be rational. I like AMD because they keep PC computing from being monotonous. You could say I’m a Cyrix fanboi too, although that obviously matters little now. And AMD has obviously been less than leading edge these days. And I don’t like that because things start to become monotonous. And when I don’t like something I talk about it.

            Get it?

            • Bensam123
            • 6 years ago

            No, I’m simply pointing out that you aren’t one so you should stop pretending you are one. I actually find it quite disheartening, because Chuckula totally gave you the go around till you can’t tell up from down. I’ve watched it happen over the past few months and it’s finally reached a point where I feel I should say something because I’m tired of watching you walk around and bump into things till Chuckula gives you the OK and points you in a different direction. You’re like a broken Roomba that can’t vacuum the floor properly.

            You are in no way a AMD fanboi anymore. You used to be, then Chuckula ‘set you straight’ to the point where I wouldn’t want to buy any of AMD products based off any of your posts. The only thing that makes me think that you would be a AMD fanboi is the fact that you keep calling yourself that… It’s not the words that count, it’s your actions that define you I think is how it goes.

            I don’t like fanbois, I don’t think being a fanboi is a good choice, but if you’re going to swear by it, which is what it seems like you’re doing, then perhaps you should actually abide by it. Flip is giving me shit because he knows I’m right and there is definitely something ‘programmed’ about the way you’ve been acting.

            In this case, don’t take my word for it, introspection is your friend.

            • ronch
            • 6 years ago

            [quote<]You are in no way a AMD fanboi anymore.[/quote<] LOL! I don't know whether to take this as a 'congrats' for snapping out of it or as a sort of damning statement. Ok, Benny, I admit, I [b<][u<]LOVE[/b<][/u<] AMD but I'm no AMD fanboi. Er... is that right? I'm confused! 😀

            • Bensam123
            • 6 years ago

            Indeed… Would you buy a AMD product based off your post? What about your posts where you make jabs at AMD in order to have a back patting time with Chuck?

            Take a step back, then look at your post(s) and tell me that’s something someone who favors AMD talks about, let alone a fanboy.

            I mean you can be a AMD fanboy, you can be a Intel fanboy, you can be a Cyrix fanboy for all I care. But when you aren’t acting the way you talk it comes off as sarcasm or even as a shill. I guess you aren’t being paid in money, so ‘friendship’?

            • ronch
            • 6 years ago

            Oh no, Benny. I don’t stab AMD just to get a pat on the back from Chuckula. Never even crossed my mind. I like AMD but it doesn’t mean I won’t criticize them the same way you love your kid but criticize him/her every time he/she does something stupid.

            FWIW, I’ve owned far more AMD chips than any other CPU brand. I originally planned to join AMD back in college (that was many years ago), but I got busy with other things. Lovin’ my FX-8350. It’s the one material thing I am most in love with. Now that’s fanboyism, is it not? But don’t ever tell me to get an FX-9590 even if they give it to me for free. I won’t take it.

            • ronch
            • 6 years ago

            No, Chuckula didn’t convert me, I converted [u<]HIM![/u<] I once made some [url=https://techreport.com/forums/viewtopic.php?f=1&t=88345<]AMD FX wallpapers[/url<] and Chucky did use them! LOL! Lighten up, bud.

            • Bensam123
            • 6 years ago

            Dude, when you first started arguing with him he wasn’t all friendly or buddy buddy with you till he finally messed you up enough where you don’t even really support AMD anymore. Of course he said he used them for a half a second to take a publicity shot then remove them. This is called saving face (You know what kissing babies is right?). How many people do you know that keep wallpaper tiled, especially geeks? It didn’t take too much to buy your friendship apparently. Talk is cheap, look at actions instead of words.

            So he had to put up a wallpaper long enough to take a screenshot, meanwhile you are spending most of your time saying how awesome AMD is followed by all the things that are wrong with them (which makes it seem like your appreciation of them is simply sarcasm) and then the occasional ‘friendly jabs at AMD’. Notice how he never makes jabs at Intel without them dripping in sarcasm (usually intended to put down something I or someone else has said to highlight one of the pitfalls of them)? Yeah it only goes one way.

            This is what operant conditioning looks like in real life.

            [url<]http://en.wikipedia.org/wiki/Operant_conditioning[/url<] He berates you and puts you down when you like AMD too much or point out good things about them. He rewards you (with friendship) when you make fun of AMD and he acts neutral when you spend the majority of your post pointing out the bad things about AMD. This is putting aside that you're essentially like their pet AMD fanboi, which they use to save face and avoid the banhammer. "See how neutral they are? They have a AMD fanboy for a friend!". They definitely weren't friendly with you when you actually supported AMD. Friends aren't your friends if you have to say or do things in order to keep them as friends, they accept you for who and what you are without you needing to change. I'm using 'friends' in a really loose context here because I definitely wouldn't consider them that. Chuckula is one of those people that you don't take for face value and you definitely should look a bit deeper then platitudes.

            • Fighterpilot
            • 6 years ago

            Why do you even care what some mediocre forum troll has to say?
            You are making him out to be some kind of Machiavellian character…dude…”Chuckula”?
            He’s low rent dude…don’t waste your breath.

            • Melvar
            • 6 years ago

            Ronch has Chuckholm syndrome.

    • ronch
    • 6 years ago

    Don’t these frame latency issues crop up in lesser GCN cards such as the 7770 and 7790? I have a 7770, and while framerates are ok, there are times when the fluidity would just have a ‘hiccup’ even in older games such as Amnesia. I paired it with an FX-8350 so I don’t know if it’s the CPU or the GPU, though.

      • jokinin
      • 6 years ago

      I’ve been having these same issues you’re describing with a radeon 7870 and a Core i5 3550, so i don’t think this has to do with either your cpu or your gpu. Having said that, it has improved noticeably with the latest set of drivers (13.xx series). Frame latency was quite visible for me in all 12.xx series drivers.

        • ronch
        • 6 years ago

        So the Core i5 also experiences hiccups while gaming? I also experienced hiccups (very rare) with System Shock 2 (old but awesome title). I attributed it to the FX but I’m inclined to think AMD’s drivers are partly to blame. As expected, apart from those rare hiccups SS2 runs like butter on the FX.

          • jokinin
          • 6 years ago

          Yes it does ocasionally. And I don’t think it’s the cpu. Ok, I don’t have the fastet gaming cpu around, but it’s an ivy bridge, quad core, 3.3GHz, and 6MB cache. Besides i also have a 2012 H77 chipset motherboard and 16GB DDR3, so I wouldn’t call my rig underpowered, and yet, i’ve been having this issues since i built this computer about a year ago.
          I’ve noticed this with various games : Diablo 3, XCOM Enemy Unknown, Civilization 5, …
          I don’t think it has to do with your FX cpu which I think is more than capable of gaming, it has to do with AMD drivers.

            • ronch
            • 6 years ago

            Perhaps TR should compare AMD and Nvidia graphics on both Intel and AMD platforms. That should tell us a few things.

      • flip-mode
      • 6 years ago

      It’s a crossfire issue, not a GCN issure. Gaming on a single GCN card will not produce the multi GPU stutter.

        • Waco
        • 6 years ago

        This. Many game engines hiccup from time to time and it doesn’t always have something to do with your GPU driver.

    • bwcbiz
    • 6 years ago

    Back when the FCAT tool was first used to analyze frame latency issues, I said I was taking those results with a grain of salt, simply because the tool was released by nVidia.

    This article demonstrates that either 1) the tool is honest and I need to take back that remark, or 2) that AMD has learned how to game the tool. Possibly both. In either case, no shame to nVidia and I take back my remarks. Even if the tool was rigged, that would mean that AMD found a way to rig their driver to exploit the tool, and both companies are even in the scamming. And I think it’s much more likely that both AMD and nVidia have found a legitimate tool and a legitimate driver tweak to investigate the issues of frame stuttering.

    • anotherengineer
    • 6 years ago

    Hmmm, nice, however the latest non-beta driver is back at 13.4?!?!

    [url<]http://support.amd.com/us/gpudownload/windows/Pages/radeonaiw_win8-64.aspx[/url<] When will we get an updated non-beta driver?

    • Chrispy_
    • 6 years ago

    Nice progress, but it still proves that you want the best single-GPU available before SLI/Crossfire should even become an option though.

    • USAFTW
    • 6 years ago

    Finally some news of comfort for people stupid enough to invest in dual GPU single card configs.
    I love the Idea, But I wouldn’t forget about making large chips and rely on simply doubling the cores.
    That’s why nVs high ends are great hits.
    Still, only took them 2 months to get a driver ready for a product Instead of rolling it out with the hardware.
    No matter what happens, I’m going AMD for my next GPU upgrade, I like their costumer focused approach to business. They may be lying, but I don’t know about that.

    • beck2448
    • 6 years ago

    According to Hardware Heaven, and HotHardware this process still has a way to go.

    • Bensam123
    • 6 years ago

    Interesting article, it’s good to note AMD is coming through with their latency enhancements. Speaking of which, any word on when AMD will release a driver with their rewritten memory subsystem they talked about when this first came about?

      • ermo
      • 6 years ago

      It’s already in the driver (since the 13.3 beta I think), but AMD realized that they’d still need to do per-game optimizations, so apparently it wasn’t the panacea they thought it would be…

        • HisDivineOrder
        • 6 years ago

        People always want a quick fix, even if there isn’t one. So companies often state they have one and then quietly acknowledge later it’s more complicated than that.

        Case in point.

        • Bensam123
        • 6 years ago

        13.3 included optimizations, not the rewritten memory management which they said they’d release a few months later.

        This may just mean they’re waiting till 8xxx to release them then.

    • tbone8ty
    • 6 years ago

    Does this work for mobile apu crossfire setups?

    Specifically the a10-4600m + 7670m (7660g + 7670m)

    Ask Amd about APU support for frame pacing.

    • beck2448
    • 6 years ago

    Kudos to TR and others for helping to get AMDs head out of its butt and stop denying what gamers have been noticing for years. Releasing products that don’t actually work as advertized is no way to keep a good reputation. The 7990 should never have been released when it was. I hope this wake up call is enough.

      • HisDivineOrder
      • 6 years ago

      Of course, it’s not. How do I know? Because they went ahead and released the 7990 while emphasizing FPS’s instead of holding it and acknowledging their Crossfire drivers needed a lot of work. If they’d “learned their lesson” from their “wake-up call,” then they wouldn’t have done that.

      They milked consumers foolish enough to trust that a 7990 consisting of 7970’s in Crossfire would actually work like it should (ie., like the marketing says it does). They took advantage of the people who trust AMD enough that they don’t need to read reviews or about frame latency because AMD’s got their back, right?

      Except no, they don’t. Didn’t they also just say the 7990’s going to be taken out of production… in about a month?

    • BIF
    • 6 years ago

    Scott, I am thinking of doing precisely that: plunking down money for one (or more) of these. But I’m a dork; I don’t plan to game with it; just fold. That’s a lot of money so I would like to know (assuming it is knowable at this point), do the new drivers either improve, break, or make no change for folding?

      • derFunkenstein
      • 6 years ago

      Worst case scenario is that you step back a driver rev for your folding rig. If you’re never going to game, there’s no reason to install this driver.

    • laggygamer
    • 6 years ago

    Is there anyone else who is bothered by the input lag sli adds? I’m just wondering if amd’s frame pacing has the same problem. Anybody know?

      • Voldenuit
      • 6 years ago

      Anandtech’s Ryan Smith seemed to think that AMD’s frame pacing delays some frames to even out frame interval – if he is correct, that would add some input (as well as output) lag into the system.

      That said, if you’re really concerned about input lag, the biggest bottleneck to input lag is typically the LCD. Moving to a low-input-lag monitor or 120 Hz display would probably reduce input lag more than turning off frame pacing.

      EDIT: corrected author of Anandtech article

        • Jason181
        • 6 years ago

        I’d actually say it feels like less input latency. I think that’s because the uneven frame delivery meant that you’d get lightning-fast input on one frame, but only a very small portion of that frame might be rendered, meaning you wouldn’t see the input changes for another swap interval. Then the next frame takes inordinately long to render, so it compounds the problem, possibly missing several swap intervals (120hz monitor here).

        • BestJinjo
        • 6 years ago

        “Not only does this driver validate everything we have worked on for the last two years but the fact that AMD has decided to enable the frame pacing fix by default emphasizes that fact even more. Evenly paced frames results in a smoother animation and does not mean that your input latency increases in any way.”
        [url<]http://www.pcper.com/reviews/Graphics-Cards/Frame-Rating-Catalyst-138-Brings-Frame-Pacing-AMD-Radeon/Closing-Thoughts[/url<]

      • Prestige Worldwide
      • 6 years ago

      I haven’t been bothered by any noticeable input lag in BF3 / 670 SLI / 120hz. I suppose a best practice would be to limit pre-rendered frames to 1 or 0 if you’re really worried, assuming you have the cpu and gpu power to still have a smooth experience with minimal prerender overhead.

      But I do disable SLI in almost every multiplayer game I play because that much power simply isn’t needed in most games I play (FEAR Combat, COD4 Promod, CS:GO, etc.)

      BF3 and Metro 2033 / Last Night are really the only games I have that necessitate SLI on highest in-game settings, and as of this week, my 2nd GTX 670 is now in my living room gaming PC hooked up to my HDTV.

        • l33t-g4m3r
        • 6 years ago

        +1 for podcast reference, but I don’t understand why people bother with mid-range SLI, opposed to just buying a single high end card.

        • laggygamer
        • 6 years ago

        Well actually in BF3 for me according to cpu/gpu graphs in game, my gpu frame delivery goes down considerably compared to single card. I thought SLI was broken in that game now. It also feels like 30ms of input lag is being added independent of fps, I thought it was just being caused by SLI. I’m at 120hz with sub 5ms input lag monitor

    • odizzido
    • 6 years ago

    AMD seems to be very dedicated to gaming of late. I am pleased.

      • HisDivineOrder
      • 6 years ago

      AMD is dedicated to the only markets they have any ability to make profits in. That’s discrete GPU’s, console contracts, and …that’s about it.

      So when you look at those two lines, gaming sorta comes out as something they really need to emphasize. Neither seem to have a particularly bright future as areas of huge growth, though. Consoles are about to be replaced by tablets (where AMD has no solid plans to be a big player) and discrete GPU sales are going to contract as desktops contract unless someone devises a way to manufacture, market, and sell an external GPU that connects to a NUC-like device for an affordable price that functions the way we wish it might.

        • ronch
        • 6 years ago

        I think AMD’s CPU division makes more money than their GPU division.

          • Firestarter
          • 6 years ago

          Or losing less?

    • DancinJack
    • 6 years ago

    So, this driver totally borked my graphics. I dont’ have CF, but I installed y’know, because I can. Driver kept crashing, system would become non-responsive, and resolution was all screwed up.

    🙁

    HD7850 2GB for those interested.

      • Stickmansam
      • 6 years ago

      its beta for a reason

      7850 2Gb as well but I rarely get any beta drivers/OS stuff since I don;t want something to screw with me

        • ronch
        • 6 years ago

        Yep. Beta is beta. Not too long ago I criticized AMD for taking so long to release a non-beta driver intended to fix their frame latency issues. Can’t remember for sure but some kids here didn’t seem to take that comment well. Well, how could anybody take AMD’s taking too long to release a non-beta driver well?

        Edit – stupid tablet toy computer’s fake keyboard sucks

      • Jason181
      • 6 years ago

      I have two 6970s in crossfire, and it does feel a bit smoother. No crashes yet.

      • Kaleid
      • 6 years ago

      Fine with my hd7850..

    • Musafir_86
    • 6 years ago

    [quote<]Maybe that pair of Radeon X850 XT cards I've been saving will be rejuvenated![/quote<] -FYI, all pre-DX11 cards have been moved to Legacy support phase and different driver releases; and pre-DX10 cards are no longer being updated with new driver since Catalyst Legacy 10.2. Regards. P/S: I know it's a joke, but still...

      • flip-mode
      • 6 years ago

      If AMD were wasting resources on tuning up DX 9 and 10 cards that would be stupid.

      Bah, ninja edit.

        • DrDillyBar
        • 6 years ago

        Depends on the update I suppose.

    • DarkUltra
    • 6 years ago

    And that, as they say, is that. Now we must look for other inconsistensies, maybe input lag or game engine update rate. I think the physx effects in Borderlands 2 are kinda laggy.

      • Jason181
      • 6 years ago

      It goes from somewhat acceptable in a single player game in most areas to a slideshow in areas like caustic caverns in multiplayer.

      • PixelArmy
      • 6 years ago

      How about frame pacing in Borderlands 2 (DX9)…? This fix only applies to DX10 and 11 games,

    • wingless
    • 6 years ago

    The ONLY reason I got a Nvidia GTX 760 and not an HD Radeon 7950 LAST WEEK was because I thought AMD would take months to get this driver out. F**K ME! I had planned on going 7950 Boost Crossfire or 760 SLI a couple of weeks after I made the first GPU purchase.

    Oh the feeling I have right now…

    FYI – My experience with Nvidia drivers has not lived up to expectations: My Windows 8.1 Preview installation got bricked by the display drivers for some odd reason. PhysX crashes Planetside 2 in fairly regular intervals of 15 minutes. There is no OpenCL 1.2 support for my GPGPU habits.

    I’ve had been looking forward to going to Nvidia for a long time to experience greener pastures, but I may not hold on to this GTX 760.

    PS: I’m no fanboy; just a disgruntled consumer of both AMD and Nvidia products as of late.

      • MadManOriginal
      • 6 years ago

      ‘Preview’ – you keep using this word, I am not sure it means what you think it means.

        • wingless
        • 6 years ago

        It means buggy Microsoft BETA! I know this. I gambled, but I didn’t expect Nvidia to drop the ball too.

          • Airmantharp
          • 6 years ago

          I don’t even know what bricked mine. I’ll be waiting for the retail release to install 8.1 again.

          • Klimax
          • 6 years ago

          Beta is beta is beta. If you are not prepared to deal with bugs then you are not definitely supposed to be on beta. (Although previews generally are between alpha/beta stages and there are some fixes coming through WU)

      • l33t-g4m3r
      • 6 years ago

      AMD’s drivers have been complete garbage ever since the beginning of dx11, and when they dropped monthly updates and fired all those employees. Vista is barely supported anymore, and 7/8 is unstable beta trash.

      I bought a 7970 for 3d, but tridef is ridiculously slow, buggy, and a hassle to use. It annoyingly restarts steam to play games, and game support is spotty. Aside from that, Magrunner (unreal engine) ran @ 30 fps. THIRTY FREAKING FPS on a $400 card! Then Hard Reset ran slower than my 470. The only game that I saw any performance out of was Metro: LL, and that’s probably because AAA titles were getting more optimizations vs indie games.

      Needless to say, I returned it and bought a 780, because I don’t think Kepler is a great product aside from the 760+, which at that point I might as well spend a bit extra and get a superior gpu. I have experienced a few problems with nvidia’s drivers, but the latest version has fixed a lot, and I’m not having problems playing games or using 3d.

      Another thing that I don’t like is these cheap aftermarket coolers that provide inadequate cooling and don’t exhaust hot air. The 780 uses an awesome blower, and there literally is no competition to it. AMD has nothing, aside from a few monstrosities that are too large to fit in my case.

      Overall, my advice to people would be hold off for price cuts and driver updates, and just tweak the graphics settings if you need more performance. Disabling SSAO and using FXAA usually does the trick. Fermi cards could completely sit out this cycle doing that. AMD guys might want to switch, considering the driver issues, although it may not be an problem depending on what games you play. I dunno, but I’ve had enough of AMD’s drivers, and IMO they’re worse now than they’ve ever been before, with maybe the exception of this current driver. Things might be slowly turning around, but it still doesn’t seem good enough for what these cards cost.

        • Stickmansam
        • 6 years ago

        I might be a bit biased as all my cards are AMD but throughout XP/Vista/7 I haven not yet had any driver or really bad game performance (vs expectations for the card). Then it might be I don’t play indie games as much.

          • l33t-g4m3r
          • 6 years ago

          I used AMD exclusively from 98-Vista. DX8-DX10. I never had problems playing games like I did with the 7970. It shouldn’t have performed worse than a 470, but it did. It could be due to poor optimization of indie titles, but IMO it shouldn’t be that dependent on game specific code. If you don’t play those games, you should be fine, but I had difficulty tolerating inconsistent framerate between games on a card that shouldn’t be having those problems. AMD needs to rewrite their driver for better general purpose performance, instead of tweaking for individual titles.

            • BestJinjo
            • 6 years ago

            Not sure what games you are playing. The opposite is true – AMD’s cards tend to perform faster in less popular games while NV cards perform better in the most popular titles. I upgraded from GTX470 SLI to 7970 overclocked. I’ve never played a single game where 470 was faster than 7970.

            In Crysis 3, 7970 Ghz is nearly doubling the 470:
            [url<]http://gamegpu.ru/images/stories/Test_GPU/Videocards/2013_1/test/c3%201920.jpg[/url<] In TR, 7970 Ghz is more than 2.3x faster: [url<]http://gamegpu.ru/images/stories/Test_GPU/Videocards/2013_1/test/tr%201920.jpg[/url<] In Bioshock, 7970Ghz is again 2.3x faster: [url<]http://gamegpu.ru/images/stories/Test_GPU/Videocards/2013_1/test/bi%201920.jpg[/url<] In M:LL it's nearly double the speed: [url<]http://gamegpu.ru/images/stories/Test_GPU/Videocards/2013_1/test/metro%201920.jpg[/url<] In GRID 2, it's double: [url<]http://gamegpu.ru/images/stories/Test_GPU/Videocards/2013_1/test/gride%201920.jpg[/url<] In Remember Me, it's more than double: [url<]http://gamegpu.ru/images/stories/Test_GPU/Videocards/2013_1/test/remember%201920.jpg[/url<] In Castlevania: LoS, it's 2.5x faster: [url<]http://gamegpu.ru/images/stories/Test_GPU/Action/Castlevania%20Lords%20of%20Shadow%20demo/test/clos%201920.jpg[/url<] 2.7x faster at 2560x1600: [url<]http://gamegpu.ru/images/stories/Test_GPU/Action/Castlevania%20Lords%20of%20Shadow%20demo/test/clos%202560.jpg[/url<] In COH2, 7970 Ghz is 2x faster than GTX480: [url<]http://gamegpu.ru/images/stories/Test_GPU/strategy/Company%20of%20Heroes%202-2/test/coh%201920%20low.png[/url<] Voodoo GPU power rating for 7970 Ghz is more than double of a single 470: [url<]http://forums.anandtech.com/showthread.php?t=2298406[/url<] I also played on a friend rig with HD6970. We did a cool comparison. A single overclocked 470 to 750mhz was slower than a stock 6970 in Crysis / Crysis Warhead. You must have had some magical 470 + a broken 7970 stuck at 500mhz GPU clock. What are demanding indie titles anyway? For indie titles a GTX660 is fast enough, not need for $400-600 GPUs.

            • l33t-g4m3r
            • 6 years ago

            LOL. Those are less popular games? Those are completely mainstream Triple A titles. I’m speaking about the performance delta between mainstream and indie. Nvidia doesn’t have this massive deficiency, whereas AMD does. Plus AMD’s 3d support is abysmal.

            Where did I say I was comparing Crysis with my 470 to a 7970? I specifically mentioned MAGRUNNER, and HARD RESET, using moderate amounts of AA/AF. Less popular games are performing below acceptable limits, whereas more popular games like Borderlands2/Metro are fine. Not that the 470 can’t run newer games, because it certainly did pull off smooth and playable framerates.

            This isn’t the first time AMD’s had performance issues with indie titles. Back when I had a 4870, it performed terribly with The Ball. AMD is profiling popular games for performance instead of general purpose optimizing, and it’s been downhill ever since Catalyst AI was first introduced on the 9×00 series.

            Edit: I also like how you say you are “not sure what games” I am playing, when I specifically mentioned Magrunner and Hard Reset. I haven’t been a fan of recent mainstream games like Bioshock, so I’ve been playing a lot of “indies”. ROTT, most recently.

            Another game I know performs like sludge is Mortal Kombat. The game literally ran in slow mo. Almost everyone with an AMD card cannot play this game, and there are numerous amounts of complaints in the forums. Runs fine on Nvidia cards.

            • flip-mode
            • 6 years ago

            It’s possible that you were doing something wrong. :shrug: I never had any show stopping bugs with any of my AMD cards. It could also be your, uh, hyperbolic view of things.

            • l33t-g4m3r
            • 6 years ago

            30 FPS in Magrunner had nothing to do with me “doing something wrong”, or my “hyperbolic view of things.” The game just performed bad, and I wasn’t about to keep using a $400 card that performed worse than a 470 because of bad drivers.

            Edit: Side mention of Mortal Kombat which is unplayable because it runs in slow mo. Nvidia cards run it fine. Check the forum complaints.

            Don’t get me wrong, the 7970 SHOULD be fully capable of 60 fps in that game, but the drivers are holding it back. That, and Tri-Def sucks the big one for performance, stability, compatibility, and ease of use. I’d rather AMD just support 3d natively like Nvidia does.

            I’ve used AMD/ATI long enough to know when they have problems, and until they fix the drivers there is no point in me using their hardware. I wouldn’t call this never ending stream of Beta / Alpha updates a proper fix either.

            Sure, the card performs fine in Crysis using standard display methods, and that might be good enough for 90% of the public, but if the indie titles I play and 3d doesn’t work, then it’s not good enough for me. AMD still has a lot of work to do.

            • Waco
            • 6 years ago

            You’re doing something wrong. There’s nothing else to it.

            • l33t-g4m3r
            • 6 years ago

            And you’re making biased assumptions without any evidence. There are real problems with AMD’s drivers, you just won’t notice it playing triple A games under “Gaming Evolved”.

            Tridef and Radeon Pro are also very poor pieces of software, and it would be a million times better if AMD would support more features out of the box, instead of allowing buggy hack job software to act as a substitute.

            • Waco
            • 6 years ago

            Am I? Hmm…I’m not sure I ever said AMD drivers were great, but the problems you’re experiencing aren’t symptoms of a healthy/properly configured system/software.

            I play a lot of indie games. Up until very recently there were three (yes, 3) 7970s in my house. Even, *gasp*, Magrunner ran fine (hell, it ran fine on a 7750).

            So yes, you’re doing something wrong.

            • l33t-g4m3r
            • 6 years ago

            3 7970’s? Now I know you’re full of it, and this was all conveniently “up until very recently”, and you “play a lot of indie games”. Right. I don’t believe you. While you’re at it, how’s ROTT and Interstellar Marines run?

            The only thing I did differently from most people was using RadeonPro, Tridef, and forced AA/AF. 4x / 16x + MLAA.

            Both tridef and rad-pro were a bit buggy, but I’ve tried it both ways. RadeonPro apparently does have some rep for causing slow downs, according to the Mortal Kombat wiki. Perhaps that was the cause of that game, but that doesn’t explain the rest.

            Regardless, I didn’t like the experience. I was dual booting with Vista, but AMD apparently doesn’t care about releasing standardized drivers like Nvidia does, so gaming on Vista was n/a, not to mention 3d is completely out of the equation because AMD doesn’t support it, and tridef sucks.

            I will say this, I forced on AA and AF in AMD’s control panel. AMD’s had performance problems in the past with forced AA, especially with the Unreal Engine, but I figured that wouldn’t be a problem with GCN. Nope. Still garbage.

            The only way you’re not having problems, is if you aren’t using any AA, or using ingame / MLAA. I could believe that, but otherwise no, you’re lying.

            I know you’re full of it, because I had no problems playing popular games like Borderlands or Metro. If there was “something wrong” with my setup, those games technically should have ran like crap and they didn’t. YOU’RE LYING. Also, all the games I had problems with on the 7970 run perfectly fine on a 780. Same AA/AF settings. The real problem is AMD’s subpar drivers.

            • Waco
            • 6 years ago

            You have some issues. Yes, I had 3 7970s between multiple rigs. Would you like the current list backed up with a picture or something else equally moronic?

            You’re the only one complaining about said problems and I have personal experience that leads me to believe you simply have something configured very wrong or very stupidly.

            I’m not sure your ranting responses deserve any more of a reply than that.

            • l33t-g4m3r
            • 6 years ago

            Pics or you don’t have 3 7970’s. Also, I clearly listed my configuration @ 4x AA 16X AF + MLAA. Indie games didn’t run very well, whereas popular games were fine.

            I don’t see how I CONFIGURED SOMETHING WRONG, WHEN POPULAR GAMES RAN FINE. Indie games usually have less graphic features than say Metro: Last Light, which was getting 60 fps. There is no other plausible explanation, other than AMD has broken drivers, and you’re a lying AMD fanboy. You also haven’t mentioned Jack about using Tridef, which I clearly stated is a complete bug-fest, and runs slow as molasses when it does work.

            Another thing that I noticed is that Nvidia has much smoother motion @ lower framerates. 30 fps was playable on a 470, but 30 fps on the 7970 was a jittery mess.

            Whether or not I had a few games “MISCONFIGURED”, (which how the F do you even do that?), means absolutely nothing, because the horrendous overall experience was why I got rid of the card, and it all works fine on Nvidia using the same settings. AMD’s DRIVER SUCKS.

            Edit: You remind me of those AMD fanboys who spewed crap everywhere about the Rage problems. AMD F-ed up the drivers. There is no other explanation. Also, amd is still slower than nvidia using tessellation. The 7970 got better frame-rates than a 470 in the Heaven benchmark, but the 470 was smoother, and had less dips. I don’t want higher framerates if they jitter all over the place, not to mention it was inconsistent and frequently dipped to 30 fps.

            • Waco
            • 6 years ago

            So when you force stupid settings you expect everything to work fine?

            Also, I review computer hardware…so yes, I have a lot laying around. I’m not taking pictures just to appease you.

            You can configure your computer poorly all you want but don’t blame AMD. They have issues with their current (and past) drivers but the drivers aren’t your problem here.

            I wish I could understand why I feel compelled to respond to your rants…

            • l33t-g4m3r
            • 6 years ago

            Right, so 4x AA and 16x AF is “MISCONFIGURED”, according to you. In fact, I shouldn’t be using ANY AA and AF, because the $400 7970 is clearly OVERWHELMED by those settings in dx10 level indie games. Right. You’re full of it. AMD’S DRIVER SUCKS. PERIOD.

            • Waco
            • 6 years ago

            You’re forcing FSAA and MLAA on games that don’t support it…and you expect things to work perfectly.

            I never said AMD’s drivers don’t suck (I would know, I ran dual 4870X2s in the past). I said you’re using stupid settings.

            • l33t-g4m3r
            • 6 years ago

            Because it DOES work perfectly, with NVIDIA. AMD is the only one that has problems with forced AA, and I figured they would have fixed the performance issues with GCN and driver updates. Nope. Still crap. AMD had an excuse with the 4800 series because 10.1 support was needed to get optimal performance, but there is no legitimate excuse now.

            The whole point of forcing AA is for games that don’t support it, as well as having a level visual experience across the board. DUH. There’s plenty of option-less games that need it, and AMD shouldn’t be taking a massive hit in old UE3 engine based games. If forced settings are a problem for AMD, but not Nvidia, then AMD has a driver issue, not a “settings” issue, when most in-game settings are either non-existent or low quality.

            Also, AMD didn’t have this issue post-dx10. DX9 performed perfectly with forced AA. This performance hit with forced settings is a recent issue, and it’s only a problem with AMD.

            AMD’s not getting any more purchases from me until this is fixed.

            • Fighterpilot
            • 6 years ago

            Thank God….now you can just whine and moan about Nvidia for a change.
            Thank heaven for small mercies.

            • ermo
            • 6 years ago

            [quote<]"AMD needs to rewrite their driver for better general purpose performance, instead of tweaking for individual titles."[/quote<] They did rewrite the memory-management code specifically to try to extract more performance out of the GCN architecture. And even with that rewrite, they realized they'd still need to focus on per game optimizations. My hunch is that AMD needs to be more proactive in reaching out to devs and help them optimize their render-path for AMD GPUs, just like NVidia is doing with TWIMTBP. Apparently, it's possible to extract a large amount of performance from GCN by using known clever tricks. So yeah, it would seem to me that AMD are (finally!) beginning to make the right moves with their two-pronged Gaming Evolved strategy (better drivers + developer reach out), and the net result is there for all to see: Gaming Evolved titles running on a 7970 perform on par with or beats GTX 680 and GTX 770 more often than not. And this definitely wasn't the case when the GTX 680 was initially released.

            • Firestarter
            • 6 years ago

            [quote<]They did rewrite the memory-management code specifically to try to extract more performance out of the GCN architecture.[/quote<] They did? Do you have a source for that? I remember them talking about it as something they could do but I haven't seen any confirmation yet that they've actually done it.

            • Damage
            • 6 years ago

            Yeah, they did roll out the new memory manager, I believe in Catalyst 13.3 beta, IIRC. I could swear I’ve written about this, but can’t find my own source to cite.

            The improvements weren’t that dramatic, and when I asked about it, the AMD rep told me they’d realized that the memory manager alone wouldn’t be a universal cure for high frame latencies. He emphasized that AMD has changed its tuning process to look for trouble spots where there are rendering slowdowns, though. I believe the revamped memory manager can help with tuning once problems are found. The bigger change is that AMD’s process now includes frame latency awareness.

            You can see they are optimizing for smoothness now in our more recent results.

            For what it’s worth, it’s also apparent that AMD doesn’t care about eliminating some of those “heartbeat pattern” spikes that show up in Fraps, so long as they don’t disrupt animation smoothness. Nvidia tends to optimize away even those, which is probably not entirely necessary but may pay dividends in scenarios where performance is just marginally good enough.

            I think there are further opportunities from both sides, extending down into GPU hardware and driver architecture and especially power management schemes, now that folks have locked on to the proper optimization target, which is consistently low frame latencies, not FPS averages. You can imagine a power-management scheme that saves up TDP “credits” when latencies are acceptable and spends them (by clocking up) when frame latencies start to rise. Excited to see how the next few years play out.

            • Firestarter
            • 6 years ago

            I can’t find anything conclusive on the memory manager, and nothing of the sort in the 13.3 beta release notes. Oh well, how they do it is secondary to the results they achieve anyway.

            • ermo
            • 6 years ago

            Scott,

            You did write about this in an e-mail reply to me dated 2013-04-08:

            [quote<] "FYI, the new memory manager is in the Catalyst 13.3 beta, but AMD has lowered expectations for frame latency improvements from it. They're now saying the key will be game-by-game optimizations for frame latencies, which is now part of their regular workflow. So.. no big fix coming, but lots of small ones as we move forward. (I really need to publish this info, eh?)" [/quote<]

            • Firestarter
            • 6 years ago

            [quote<]I really need to publish this info[/quote<] only about 4 months late!

        • CaptTomato
        • 6 years ago

        I’ve played a bunch of games on my 7950, and it’s fine, it’s really just more powerful than my older and stable 6850/4850/x800 series of ATI/AMD cards.

        • bjm
        • 6 years ago

        So, TR’s [url=https://techreport.com/discussion/24218/a-driver-update-to-reduce-radeon-frame-times?post=703064<]false controversy[/url<] with AMD's drivers wasn't false after all? Now that you've actually purchased the card despite TR's best recommendations and seen the performance issues first hand, all of the sudden you are singing: "AMD's driver has had issues "since the beginning of dx11" and that you've used AMD/ATI long enough to know their problems"? Instead, you should be saying: "Damage, I am so sorry for being a complete ass and for attacking your character. I went against your recommendations and they really DO have issues. It wasn't a conspiracy theory after all! I see the light, oh Lord Damage! You were right and I was wrong! I repent!"

          • wingless
          • 6 years ago

          What do you expect from somebody called l33t-g4m3r????!!!!!! That is the douche-baggiest and pretentious screen name I’ve seen in a while.

        • Maff
        • 6 years ago

        I also had bad 3d performance in tridef with my 7950. Then i found out that apparently you need to disconnect other monitors in order to get decent 3d performance. I know its a somewhat crappy trick you have to pull, but it ment going from 20 fps and low GPU utilisation to 60FPS and great GPU utilisation. Before i found out about the multi-monitor issue I always thought my 8150 wasn’t up to the challenge of 3d(twice the framerate and all that) but apparently it was easily solvable.

      • Bensam123
      • 6 years ago

      Update the PhysX driver or disable physics in the Nvidia control panel.

        • wingless
        • 6 years ago

        Stating the obvious on a site like this (…I sound like an ass on purpose).

          • Bensam123
          • 6 years ago

          That’s not the obvious for most people, especially the option to disable hardware accelerated physx which most people don’t know about. Although I do admire your tenacity at keep trying to bump this post, it seems to be turning viral.

      • DPete27
      • 6 years ago

      If you were so adamant about it, perhaps you should have done some googling or asked on the TR forums about this. [url=http://www.anandtech.com/show/6857/amd-stuttering-issues-driver-roadmap-fraps/6<]We've known since March 26th that AMD would release this driver fix in the July driver.[/url<]... Ok so it was August 2nd instead, but close enough.

        • wingless
        • 6 years ago

        You are CORRECT! My own, ignorant, blind assumptions lead to me feeling PISSED! I’m pissed at myself for not remembering this news from months ago. I would have marked it on my calendar as a holiday.

          • flip-mode
          • 6 years ago

          Upvoted for holding yourself accountable. Don’t know why you’d be downvoted for that.

      • HisDivineOrder
      • 6 years ago

      Windows 8.1 bricking seems pretty prevalent and not everyone having fun with the Windows 8.1 Preview is using nVidia. 😉

        • wingless
        • 6 years ago

        AMD probably plays better with Windows since they built in all that specific Win 8.1 support using the DX11 11_1 feature set. Unfortunately removing the GTX 760 and running on my i7-2600K’s iGPU still did not allow Win 8.1 to boot. Buggy Windows + buggy GPU drivers is a recipe for disaster; a lesson I learned before.

        I clearly do not learn from my past mistakes (well I no longer run RAID-0 OCZ SSDs so I can’t be too stupid).

    • chuckula
    • 6 years ago

    You see, this is how the whole hardware review process is [i<]supposed[/i<] to work. Independent review sites like TR, PC perspective, and a few others go out there and actually test the hardware to find potential problems. Then the manufacturers actually listen and take the time to understand problems that are uncovered, and the manufacturers work hard to [b<][i<]fix[/i<][/b<] the problem as best they can. Good job TR and props to AMD for paying attention to this issue and taking steps to improve the situation.

      • Voldenuit
      • 6 years ago

      What you said, except without the 2 year lag between discovery and acknowledgement+fix.

        • wingless
        • 6 years ago

        2 years? Is that an accurate statement? When was FCAT testing first introduced by TR and PCPer?

          • Voldenuit
          • 6 years ago

          TR’s first ‘[url=https://techreport.com/review/21516/inside-the-second-a-new-look-at-game-benchmarking<]Inside the second[/url<]' article was published in September 2011. Users had been complaining about microstuttering on multi-GPU configs (both amd and nvidia) for even longer. So 2 years sounds about right. PCPer was testing its own capture hardware and frame marking tools for a while before nvidia's FCAT hardware was made available, too.

            • Damage
            • 6 years ago

            Nah, they were beta testing for Nvidia. 😉

            • Voldenuit
            • 6 years ago

            Interesting. Explains why PCPer’s frame marking methodology was so similar to FCAT.

            • wingless
            • 6 years ago

            You have educated me. Thanks!

        • ronch
        • 6 years ago

        Well, yeah, except AMD has been bleeding talent profusely for the last two years.

      • spuppy
      • 6 years ago

      TR was talking about it for years, but AMD didn’t listen until Nvidia supplied FCAT hardware to TR, PCPer, Anandtech, and Tom’s.

        • beck2448
        • 6 years ago

        And Hardocp for at least two years kept saying the actual gaming experience on Crossfire did not equate with the fps numbers. FINALLY the RUNT frame issue was exposed.

          • Firestarter
          • 6 years ago

          I’ve never really liked their subjective benchmarking methods, but this is something that they were right about all along.

      • flip-mode
      • 6 years ago

      Wow, too bad I didn’t get until you had +30. See, I don’t think this is how it should work. It shouldn’t take AMD 2+ years to respond to tech press feedback. Let me quote from the article conclusion:

      [quote<]But realize that literally years have passed since the first forums and tech websites, particularly in Europe, started talking about microstuttering. Nvidia claims to have implemented its frame pacing solution back in the G80 generation of cards, a claim we still need to confirm. We know AMD did nothing back then. We first talked to AMD about microstuttering nearly two years ago, for our first Inside the second article. We illustrated the problem with frame-by-frame plots at that time. Evidently, AMD did nothing. Only when we irrefutably pinpointed the problem at the display level with the aid of the FCAT tools supplied by AMD's competition did Team Radeon begin taking microstuttering seriously. Work on a driver fix finally began. With this issue exposed for all to see, the firm went ahead and released the Radeon HD 7990 to consumers when a fix was months away—and marketed the product on the basis of FPS scores, a metric multi-GPU solutions with microstuttering essentially inflate artificially.[/quote<] So, no, the process didn't work right because AMD didn't participate in the process for far too long. The "process" is a feedback loop between AMD and the tech press. If AMD doesn't join it or takes far too long to do so then the process isn't working.

      • ronch
      • 6 years ago

      And this happens because there is still competition in the graphics industry. Do you think AMD would listen or fix this issue as well as it had if Nvidia didn’t exist? Any company would just try to get away with it.

Pin It on Pinterest

Share This