• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

First time Crossfire and disappointed....

Joined
Oct 19, 2002
Messages
618
I just set up my two 6950 Reference cards with unlocked shaders only. Both cards perform flawlessly individually. When I run 3dMark '11, the score goes from ~5500 to ~7500.

In game I notice no difference. I've checked GPU-Z, it says both cards are running in Crossfire mode. Cards are clocked at 845/1300.

Am I doing something wrong or is my processor just not up to the task?

Do I need 2 crossfire bridges or will one do? I have read where it does not make a difference. If so I need a long one and do not have another, I have like 6 short one, but the only long one came with my board.
 
I just set up my two 6950 Reference cards with unlocked shaders only. Both cards perform flawlessly individually. When I run 3dMark '11, the score goes from ~5500 to ~7500.

In game I notice no difference. I've checked GPU-Z, it says both cards are running in Crossfire mode. Cards are clocked at 845/1300.

Am I doing something wrong or is my processor just not up to the task?

Do I need 2 crossfire bridges or will one do? I have read where it does not make a difference. If so I need a long one and do not have another, I have like 6 short one, but the only long one came with my board.


First thing I would do is uninstall drivers and reinstall them. Try that and post new 3Dmark scores and see if there is any change. Run some in game benchmarks before and after as well.

As for the crossfire bridge you only need one.

What resolution are you running at?

Your sig says your 1055t is over clocked to 4ghz? You probably have a bottleneck but I don't know if it should be that much.

More info from you needed though.
 
Try that, and also try running your video cards with CFX turned on and off to see what happens. A good benchmark i've found was downloading the Unigine Heaven benchmark, which gives you all sorts of benchmarking information at the end. At least with my 5850's, my score at the end decreased significantly when I tried running it with CFX disabled in the Catalyst menu.
 
Here's my score with a couple reference 6950s in crossfire (stock) for your reference. System specs are in sig.



6950-xfire-score.jpg
 
question is, you notice no difference compared to what? single mode vs xfire? if so, what games are you using to compare? certain games don't scale well or at all in multi gpu mode. you should test several games that support xfire before drawing a clear conclusion.
 
On new 11.4 drivers I got a p8245. I guess it is working , just not as much as I had hoped for. I mainly play Metro 2033, Just Cause 2, and Bad Company 2. I may to to squeeze more out of my CPU.
 
it could be your CPU is bottlenecking..

When I was playing around with 5970, Q6600 was bottlenecking even on 3.8Ghz, which in term of raw performance in game (dual core/quad core support only) should be the same as your CPU..

Which in the end, you might see bottleneck pretty hard unless you crank up your resolution.
 
What resolution are you gaming at?
Only very high resolutions such as 2560*1600 fully benefit from 6950/6970 CrossifreX setups.
 
6080x1080 I think...

I got my processor to 4.1 and it bumped 3dmark scores up some more. But BC2 still hovers around 30 fps. Metro around 40 with DOF disabled.
 
6080x1080 I think...

I got my processor to 4.1 and it bumped 3dmark scores up some more. But BC2 still hovers around 30 fps. Metro around 40 with DOF disabled.

So what fps do you get with crossfire disabled for these games?

3Dmark scores are really CPU dependent, especially at the stock resolution. Honestly I don't think your processor is the limiting factor for BC2.

Also, make sure you don't have a web browser with flash 10.x running in the background. When I was running the 11.1 drivers that locked one of the video cards at the "flash" clock speeds and it wouldn't clock up to full speed. I think they've fixed that with the newer drivers.

MSI Afterburner matched with the LCD display on my G15 has been invaluable in helping me work out my crossfire issues. I can see exactly what the utilization and clock speed of each gpu is real time on the G15's LCD display as well as the current fps. I think you can configure afterburner to display that info on screen.
http://event.msi.com/vga/afterburner/download.htm
 
Last edited:
I also own a few AMDs and I absolutely love them but the IPC is less powerful than the Intels. You might, and I say this with out substantial backing, be running out of processor steam. The 6 Core AMD is great for Video encoding, etc.... but they will NOT hang in Instructions per clock/cycle as the new sandy's or even some of the slightly older I7-950+ chips.

Also keep in mind that you are possibly core limited also. Maybe not in Heaven benchmark, but in games etc... which most use one, maybe 2 cores at most still as far as I know. You may not show as high a performance due to AMDs slightly lower IPC performance. Once again I could be wrong but I remember reading that xfire and sli are more processor intensive as the cards themselves do not contol the bus and all the math that flows across the bus. The cpu still handles all of the card to card communications across the bus.

No not a fan boy I just own both platforms and they both have their place that is for sure.

Edit:
Here is my Pmark Score about 10 mins ago. Time stamp in the screenshot. Well of course this is on 6970s in dual xfire. 4.8ghz I7. I think the biggest hit you are taking is not from your cards but your processor.
Pmarks.JPG
 
Last edited:
Back
Top