• 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.

Can you get rid of microstutter?

mof

mof

Associate
Joined
19 Jul 2011
Posts
288
According to this - http://www.overclock.net/amd-ati/410272-4870x2-vista-eliminate-microstutter.html you can get rid of microstuttering by changing the PCI latency times.
Can anyone confirm this? Does this only work with older components?
I'm using 5850x2 and a gigabyte 890fxa-ud5 motherboard.

I've also read that setting the gpu/memory clock speeds of the second card to be higher than the first card works but it didn't help for me.

I've experienced MS constantly in COD2, COD4 and HAWX and occasionally in BC2.

This link http://hardforum.com/showthread.php?t=1523306 shows someone who claims that shutting down CCC.exe or/and MOM.exe fixes the problem but I wonder if shutting down CCC would stop CF from working and thats why it appeared to have been fixed it when it hadn't.

Has anyone been successful in eliminating microstutter from a game where it was happening consistently?
 
If you look on Toms Hardware you can find a review/test regarding microstutter on both AMD and Nvidia hardware. It is apparently at its least on AMD/ATI 3 card/GPU systems.

From memory I think they had really good results with a 6870x2 (shouldn't they have called this a 6890?) and a regular 6870.

Now I am thinking about this and I am wondering if differing PCI-E bandwidth variations could be working favourably. For instance 3 independent cards would all operate at x8, a twin card at x8 per core and the second card most likely at x16.

I am wondering if it is anything like RAM slots TBH. Ram slots (generally mostly when overclocking) often need offsets between banks because physically the wiring (PCB circuit) to each bank is physically different in length. With the very high bandwidth of Modern PCI-E I could see how potentially this could lead to a similar scenario which would be different from mobo to mobo as all layouts are different. Perhaps running on a lower speed of x8 may disguise this somewhat. Sorry, that is mostly speculation but I thought it may be worth mentioning.
 
Just got a second 5850 for CF, no hint of stutter in BFBC2. I'm quite sensitive to stutter/mouse lag/tearing

Running at x8 x8 on an Asus M4A89gtd Pro with a 1055t @4.0GHz with 4GB 1600mhz DDR3. Not changed any bios settings, just plugged it in and enabled CF in Cats 11.8
 
We have had SLI and Crossfire now for 6 years and neither AMD or Nvidia (with their god like driver team as some would leave you to believe) has been able to solve the issue. I think if you go with a dual card setup microstuttering is something your going to have accept.
 
This is my first Sli setup and tbh i have not really experienced microstutter, the only game i do get noticeable microstutter is FarCry2 and because of some strange texture issues that is the only game i run on a single gpu lol, Afterburner does still show that even running on a single gpu both cards have the same memory usage tho.
 
Changing PCI latency timer will do nothing to help with microstutter, AMD optimised games suffer from microstutter just as much/little as any other, they certainly can't eliminate it on AMD setups even with AMD optimised games.

There are various things that can make microstutter worse and much more noticeable.

If you want to reduce microstutter as much as possible ideally you need a setup thats holding atleast 60fps minimum and artificially being limited to somewhere around 60-90fps. Unfortunatly VSync isn't always the best way to achieve this due to more complicated issues.
 
Yea i run Vsync is most games and get pretty much a steady 60fps so maybe that's why i don't notice it, they do say even tho a pair of low - mid range cards in sli/xfire can give better bang for buck they are not always the best option over a single high end card because thoes setups are more prone to microstutter.
 
VSync will generally almost eliminate noticeable microstutter, some other scenarios/issues aside the problem is you usually trade off with increased input latency - which some people notice and some don't.
 
Just got a second 5850 for CF, no hint of stutter in BFBC2. I'm quite sensitive to stutter/mouse lag/tearing

Running at x8 x8 on an Asus M4A89gtd Pro with a 1055t @4.0GHz with 4GB 1600mhz DDR3. Not changed any bios settings, just plugged it in and enabled CF in Cats 11.8
Since I got a second card I've definitely had an improvement in frame rate in BFBC2 and it runs smoothly most of the time but I've experienced microstutter also crashes and parts of the map going black.
amd optimised games don't microstutter on amd cards... Crysis2 microstutter was awefull on my rig so i binned the game what a load of ***** it was anyway!
Crysis 2 (also Crysis) is one of the only games where after getting a second card I've had a significant improvement in performance without any problems going along with it. Dirt 2 showed an increase in frame rate but it was already above my monitors refresh rate so I didn't notice it.
Changing PCI latency timer will do nothing to help with microstutter, AMD optimised games suffer from microstutter just as much/little as any other, they certainly can't eliminate it on AMD setups even with AMD optimised games.

There are various things that can make microstutter worse and much more noticeable.

If you want to reduce microstutter as much as possible ideally you need a setup thats holding atleast 60fps minimum and artificially being limited to somewhere around 60-90fps. Unfortunatly VSync isn't always the best way to achieve this due to more complicated issues.
On the HAWX benchmark my max FPS was over 200 and the minimum (or possibly average) was over 100 but I still notice the MS straight away with v-sync on/off.
 
Back
Top Bottom