• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

SLI/CF is microstutter an issue for you?

heaven2013-03-2121-47-08-68-Time.png


My Heaven run but only the first 2 mins.
 
Im intrigued by these results, do any of you guys notice the microstutter while playing? or do you only notice it when viewing the results?

I never noticed it with my old xfire, but i dont have the cards anymore to bench them unfortunately.
 
Im intrigued by these results, do any of you guys notice the microstutter while playing? or do you only notice it when viewing the results?

I never noticed it with my old xfire, but i dont have the cards anymore to bench them unfortunately.

In the benchmarks and BL2 it was horrid. BL2 was unplayable until I used RadeonPro and set the frame render thing to 1.
Though the benchmarks and a clouple of other games (mainly metro) this doesn't cure at all.
 
Surprised to see its smoother on 7950 xfire than a 690. I thought the 690 had something that was supposed to eliminate micro stutter, frame metering or something. Is that just a gimmick?

I think the GTX 690 graph @1080p is better than the HD 7950s running the same resolution. If the HD 7950s were run @1600p that would be interesting to compare to the one for the GTX 690 @1600p.

I have noticed that workload has an effect on the graphs. Higher settings = lower framerates and worse graphs.

On the quad sli graphs I did, the results were not as bad as what most people would have predicted for 4 GPUs. I think the reason for this is 1080p on Heaven 4 is quite easy for quad sli, the graphs were worse for 1600p due to the extra workload/lower framerate.

When comparing these graphs I think its very important to use as near identical settings as possible.
 
Question regarding Gregs/Kaps/my Heaven results:

I was under the impression the spikes over 33ms was the smoothness issue and that's why the 33ms bar is always present in the FT graphs?

With the results in Heaven being over 33ms@Titan(+13)>7950 CrossFire(+37)>690(+45) in regards to our three particular setups.

I have never really looked into it that much as it doesn't affect me outwith FC3.

Im intrigued by these results, do any of you guys notice the microstutter while playing? or do you only notice it when viewing the results?

I never noticed it with my old xfire, but i dont have the cards anymore to bench them unfortunately.

The only time outside of Farcry 3(where it's was broken@launch-fixed by AMD-broken by Ubi each time they released a patch) where microstutter wasn't just noticeable-it was plain terrible, was on a stopgap 555BE that couldn't cope with twin 6950>70's, but that wasn't down to the gpus-the cpu just couldn't cope and it was a nightmare.

Put my 2500K in and it was gone, super smooth.

The aim is to make sure you have a setup to maximise the power of CF/SLi in order to eliminate all bottlenecks for maximum performance.
 
Im intrigued by these results, do any of you guys notice the microstutter while playing? or do you only notice it when viewing the results?

I never noticed it with my old xfire, but i dont have the cards anymore to bench them unfortunately.

I never realy noticed (but felt it wasn't 100% smooth) till running the fraps bench viewer. It does show any stutters and I felt 50ms wasn't smooth but it can be as low as 35ms that shows slight stutters.

I think Heaven is just an example of how it can be clearly seen on both systems. The start jutters and the part where you go through the corridor/courtyard and look over the bridge at night is another bad part and generaly the whole run doesn't feel smooth in any place.
 
Tri fire 7970's

BF3 multiplayer 64man

Vsync off and no frame limiter, no tweaks, out of the box.

capturejtz.png


2560x1600 ultra everything 4xmsaa
 
Last edited:
Back
Top Bottom