AMD Comments on GPU Stuttering and will offer latency or FPS orientated driver option

[CAT-THE-FIFTH]

More details here:
http://www.anandtech.com/show/6857/a...-roadmap-fraps

In a typical AMD move, AMD will ultimately be leaving this up to the user. In their July driver AMD will be introducing a multi-GPU stuttering control that will let the user pick between an emphasis on latency, or an emphasis on frame pacing. The former of course being their current method, while the latter would be their new method to reduce micro-stuttering at the cost of latency.

[vinnyT]

Thanks Cat, interesting reading.

[Willzzz]

Why can't we have both?

The fact that AMD says that we can only have one or the other suggests that despite their protestations the Present call is indeed in the wrong pace.

AMD are saying that in order to have regular frames displayed that they would need to introduce a wait, thing is that they wouldn't if the START of the pipeline was more regular.

[kalniel]

Like I commented on this article in the other thread, it's very disappointing to read AMD only just noticing the problem and addressing it. Most of the article is basically on the defensive.

Tech report's analysis with the FCAT tools is also worth a read in this context:
http://techreport.com/review/24553/i...-capture-tools

[HalloweenJack]

well at least AMD are saying `yes there is an issue`

as allways NVIDIA are silent on SLI micro stuttering

[Willzzz]

Well the thing is if you don't notice, then you don't notice. Nvidia was hardly going to tap them on the shoulder and say "you've missed something", at least not until they were comfortably ahead in terms of providing a solution to the problem.

I wonder whether the problem is in terms of predicting timings inside the pipeline and then accounting for them or getting the pipeline really regular in the first place.

Or is it actually the applications themselves? Are the problems actually outside the drivers? Do the games not react properly to the video cards? If so how does Nvidia trick the game engines into behaving? Why do some games developed with AMD still behave poorly?

Actually Crysis and Hitman do seem to behave, but Tomb Raider is just awful according to Tom's.

[CAT-THE-FIFTH]

well at least AMD are saying `yes there is an issue`

as allways NVIDIA are silent on SLI micro stuttering

The same goes with the options they are giving,it gives the user control over what they feel will suit them.

In our discussion with AMD, AMD brought up a very simple but very important point: while we can objectively measure instances of stuttering with the right tools, we cannot objectively measure the impact of stuttering on the user. We can make suggestions for what’s acceptable and set common-sense guidelines for how much of a variance is too much – similar to how 60fps is the commonly accepted threshold for smooth gameplay – but nothing short of a double-blind trial will tell us whether any given instance of stuttering is noticeable to any given individual.

AMD didn’t have all of the answers to this one, and frankly neither do we. Variance will always exist and so some degree of stuttering will always be present. The only point we can really make is the same point AMD made to us, which is that stuttering is only going to matter when it impacts the user. If the user cannot see stuttering then stuttering should no longer be an issue, even if we can measure some small degree of stuttering still occurring. Like input lag, framerates, and other aspects of rendering, there is going to be a point where stuttering can become “good enough” for most users.

Hardly being defensive in anyway,they are just being honest.

[kalniel]

This is the bit that strikes me as particularly arrogant and ingorant. Good that's it's changed now, but it's too late really.

The shortest answer also the bluntest answer: AMD had a stuttering problem because AMD wasn’t looking for a stuttering problem. AMD does a great deal of competitive analysis (read: seeing what NVIDIA is doing) on overall performance, but AMD was never doing competitive analysis for stuttering.

Because stuttering is such a complex issue and AMD had such great knowledge into their drivers, AMD assumed that stuttering was occurring due to the applications and the OS, things that were out of their control. Furthermore because those things were out of their control, AMD assumed that they were happening to NVIDIA and Intel GPUs too. After all, there wasn’t any kind of competitive analysis to scientifically confirm this. AMD never saw that NVIDIA cards weren’t experiencing as much stuttering, and consequently never saw that they did in fact have more control over stuttering than they first thought.

That's just a failing really. Someone else proved that nVidia suffered from less stuttering, so now AMD act. They should have spotted it before.

The defensive bit is where they go on about fraps stuttering not being represented in the real world. Tech Report prove that's not the case for big spikes.

[CAT-THE-FIFTH]

This is the bit that strikes me as particularly arrogant and ingorant. Good that's it's changed now, but it's too late really.

That's just a failing really. Someone else proved that nVidia suffered from less stuttering, so now AMD act. They should have spotted it before.

No different to when it took third parties to show the bumps issues Nvidia had with some of their earlier GPUs,even though it caused a high rate of failures among certain chips.

It took them time to even acknowledge that problem existed(I don't think they even apologised) and before then people were being palmed off with updates to their laptops,which increased the fan speed. Its only when the problem could not be buried,they ended up acknowledging the issue. Intel OTH proactively solved the issue with the B2 chipset problems,before any real problems started being displayed.

The same goes with the Nvidia's own stuttering problems which people were complaining about,and it took a decent amount of time before Nvidia said it would solve the problem.

So,Nvidia has done exactly the same too like AMD.

On the AMD side,the GSOD issue too. There are loads of issues both companies seem to miss out,until someone "discovers" them.

They both have had a history of doing this,ie,sitting on issues until they are too big to be ignored.

Its not really acceptable.

The defensive bit is where they go on about fraps stuttering not being represented in the real world. Tech Report prove that's not the case for big spikes.

http://blogs.nvidia.com/2013/03/with...-what-you-get/

The problem: FRAPs accurately measures frames when they are transferred from a game engine. But a lot happens between that point and when a game gets to the screen. In fact, we found that gamers weren’t imagining things: there can be a big difference between what users see on FRAPs and what they experience.

Two of the problems: what we call ‘drops’ and ‘runts.’ Drops occur when frames that are counted by FRAPs are never displayed. Runt frames, by contrast, are displayed, but only for a few lines of the full 1080p that should be displayed.

Pfft,it seems Nvidia also does not seem to like FRAPs too.

So,it seems both AMD and Nvidia are apparently wrong then,in their judgement about FRAPs.

Edit!!

Both of the major GPU makers, AMD and Nvidia, have told us that the results from Fraps don't tell the whole story—especially when it comes to multi-GPU solutions.

That is from the TR article,you linked to.

So,I am failing to see why AMD is considered defensive,if Nvidia basically thinks the same??

Hasn't TR sort of proved both of them wrong then??