Widescreen Gaming Forum

Hi guys,

First time here... I can see that there is a vast of information and approachable people in this forum. So I ask for some guidance or re-assuring that this is not happening just to me.

I have 3x 6970's and running eyefinity in most games. However I notice that most (not all) games that I play in crossfire and eyefinity, I get some REALY bad performance issues in comparison to just running on one card. Which is really strange, because running a program like radeaon pro, the framerates tell me that I am getting sometimes double the performance but visually I get terrible fps.. All of this while attempting the same game, same screen res (5760x1024)

Hard to describe... say for example when I play either dirt 2 or dirt 3. Playing on ONE card in 5760x1024, runs like silk. Playing in crossfire with all three cards runs like GARBAGE, if it runs at all... many times the game crashes. I must say, this is only happening with eyefinity+crossfire!.. if I switch off eyefintiy and running in 1920x1024 on one screen, no problem in the sky... runs silky smooth and no crashes with crossfire enabled.

I guess my question is after all that rambling on my part... is this just to do with game patches?

The list of games I see the same problem in is:
NFS Shift 2
Dirt 2
Dirt 3
Grid
Just Cause 2
F1 2010
Blur
GTA 4

The list of games that seem to work very well when both eyefinity and crossfire is enabled:
NFS Hot purtsuit
Call of duty black ops
Metro 2033 (damn this game is hungry for even MORE power with dx11 and everything on at 5760x1024!!!) :)
Assasins creed brotherhood
Crysis 2

Anyway, I am very happy to have found this site... and I hope I can contribute when I can..

Cheers,
Ndev

I had the same problems as you with my trifire 6950 setup.
I changed my 3x6950 with 2x6970 and the most of the problems solved.
I think the third gpu core makes a lot of problems at eyefinity mode.
I have dirt 2 and 3,f1 2010,blackops and bfbc2.
No more problems with my crossfire/eyefinity setup. :D

I have Dirt2,3 Just cause 2 and f1 2010, tri-fire flashed 6950's and mine run a lot smoother all enabled in eyefinity. I am running a higher res with 3 30's though. I know I had one driver issue in the past with dirt 2 and I did not recently start playing dirt 3 till the 11.8's

I hated running crossfire with eyefinity. I got terrible microstutter in every game i tried. Not worth it imo

I hated running crossfire with eyefinity. I got terrible microstutter in every game i tried. Not worth it imo

Was that on 5xxx series cards? I saw that first hand on 5870's, but I've never seen it on my 6950 setup.

[quote]I hated running crossfire with eyefinity. I got terrible microstutter in every game i tried. Not worth it imo

Was that on 5xxx series cards? I saw that first hand on 5870's, but I've never seen it on my 6950 setup.
Yep. I was using 2 5850's. Needless to say, I sold one since my gameplay was much smoother with 1 card.

With regards to micro stuttering, this article is well worth a read.

[#16639]

_________________
: ) Skid : )
My Multi-Monitor Focused YouTube Channel: http://www.youtube.com/c/SkidIncGaming
My Twitch channel which I've been known to stream on from time to time: http://www.twitch.tv/SkidInc
"Imagination is the only weapon in the war against reality."

With regards to micro stuttering, this article is well worth a read.

[#16639]

I was really unimpressed with their review techniques. Testing only Call of Juarez (why this game anyway?) and then showing fps plots with only numbers (average I assume? also never labeling the axes) is just not informative.

[quote]With regards to micro stuttering, this article is well worth a read.

[#16639]

I was really unimpressed with their review techniques. Testing only Call of Juarez (why this game anyway?) and then showing fps plots with only numbers (average I assume? also never labeling the axes) is just not informative.

Helps if you read the text rather then just look at the graphs, the review is extremely in-depth, they only gave the expanded details of one game because doing it for every game would of been redundant, they did however test a number of games and comment on the micro stuttering in each. They do however take it for granted the you know they are talking about fps over time, since that's the core of the subject.

_________________
: ) Skid : )
My Multi-Monitor Focused YouTube Channel: http://www.youtube.com/c/SkidIncGaming
My Twitch channel which I've been known to stream on from time to time: http://www.twitch.tv/SkidInc
"Imagination is the only weapon in the war against reality."

I did read the article. It was posted on the 23rd on the [H] forums and I spent a good amount of time going over it then. I just read through it again for good measure.

My main issues:
- Test methodology is never outlined - what software are they using to capture and create these microstutter graphs? I would love to be able to capture this data on my own machine.
- What are the units of each axes in their stutter graphs? (specifically the X-axis) When in the level/benchmark does this snippet of information occur?