(05-27-2013, 11:15 AM)ExtremeDude2 Wrote: [ -> ] (05-27-2013, 11:00 AM)admin89 Wrote: [ -> ]Quote:yes, even in dolphin
Proof ? Dolphin don't give a damn about dual GPU , SLI , Crossfire
Exactly, 1 GTX 680 is faster than "1" GTX 690
In theory the GTX 690 is less that twice as fast:
http://www.hwcompare.com/12659/geforce-gtx-680-vs-geforce-gtx-690/
And since dolphin only uses "half"... :3
I suppose there is alway the problem of micro studer that the GTX 690 won't have... (doesn't effect dolphin?)
690 will have studders like any Crossfire/SLI config.
(05-27-2013, 12:45 PM)Venomx1 Wrote: [ -> ] (05-27-2013, 11:15 AM)ExtremeDude2 Wrote: [ -> ] (05-27-2013, 11:00 AM)admin89 Wrote: [ -> ]Quote:yes, even in dolphin
Proof ? Dolphin don't give a damn about dual GPU , SLI , Crossfire
Exactly, 1 GTX 680 is faster than "1" GTX 690
In theory the GTX 690 is less that twice as fast:
http://www.hwcompare.com/12659/geforce-gtx-680-vs-geforce-gtx-690/
And since dolphin only uses "half"... :3
I suppose there is alway the problem of micro studer that the GTX 690 won't have... (doesn't effect dolphin?)
690 will have studders like any Crossfire/SLI config.
Haven't experienced microstutter at all since probably GTX 2XX Series with SLI.
I doubt that. They aren't as bad as they once were but they're still there and easy to measure.
(05-27-2013, 01:28 PM)NaturalViolence Wrote: [ -> ]I doubt that. They aren't as bad as they once were but they're still there and easy to measure.
Microstutter is not a problem any more with SLI, I assure you.
If it was, I wouldn't be running it for the past 3 years.
Dual GPU cards might be different, I haven't run that since GTX 295 ( which was king of microstutter ).
And I assure you it is. Microstuttering is inherent to how AFR (alternate frame rendering) works. It's technically impossible to implement AFR without increasing microstuttering. And nobody has ever been able to show that their system has no measurable increase in microstuttering with AFR enabled. Over the years lots of organizations and users have collected data on this from various drivers and graphics cards all the way from the titan back to the 7 series. The conclusion has always been the same, a measurable increase.
Now whether the increase produces a level of microstuttering that is noticeable to a particular individual is a different story. Just because it exists doesn't mean that you will notice it. Clearly you aren't noticing it, but that doesn't mean that it doesn't exist.
Some people are more susceptible to noticing it then others. And some people actively look for it, which of course helps you notice it.
There are also a number of factors that contribute to the impact AFR has on microstuttering. Nvidia's implementation (SLI) seems to be much more effective at minimizing it than AMD's implementation (crossfire). Most likely due to better synchronization between GPUs. Framerate and vsync also have a big effect. Ideally you want your framerate to be above your refresh rate to minimize the effect. Increasing the number of GPUs in the chain also has a positive impact on microstuttering. Though you would theoretically need an infinite number of gpus to achieve a 0% increase in frame time variation, so the effect is still there and still measurable. And of course there are probably other things that effect it. Dual GPU cards should in theory have less of an increase due to tighter synchronization being possible. But I've never looked at any data with that comparison in mind so I don't know for sure.
I'll agree that for most people the effect probably isn't noticeable on a modern SLI setup unless the framerate is very low. But to say that the problem has been completely eliminated is just factually wrong.