[Solved] CrossFire not working.

My system doesn't seem to be running Crysis Warhead any faster at all when I use two HD 5850's compared to single HD 5850. I had an MSI 770-G45 motherboard, and when I installed the two cards on that motherboard my performance actually took a hit due to the x4 operation slot on the second card. So I changed it out with an ASRock M3A785GXH128M motherboard, which had x8/x8 operation slots, but I still am not seeing any performance increase. Sometimes I think it runs a bit faster with single card, by 5 to 10 frames or so.

What could cause this? I have a CrossFire bridge attached to both cards, Phenom II X4 955 @ 3.8Ghz, Rosewill RX850 PSU, and 8GB DDR3 1333Mhz. I tested the performance in Crysis Warhead, and then in Bad Company 2, both did not show any performance gains. The 3DMark Vantage score did go from 15,000 to 22,000 though, but the gaming performance does not reflect that at all.

What version of Windows and what version of ATI Catalyst drivers are you using?

Unistall all ATI video drivers power the computer off and take both card's out and install one card in the primary pcie slot. Power on system and install the latest catalyst and drivers (11.1). After the drivers and catalyst are correctly installed power down the computer. Install second video card in secondary pcie slot and connect the crossfire bridge. Power the computer back on and open catalyst control panel and open the performance tab on the left side. Click on the crossfire tab and check to see if the box is checked. It should be checked automatically once you power the computer on after installing the second card.

Once you have done this in this order let me know.

I've done everything you told me above, and the crossfire box was automatically checked as you said. What's next?

I benched Crysis Warhead and with single card it runs at 60FPS, with CrossFire it runs at 64FPS, minimum of 38FPS with single card and 33FPS with CrossFire.

I tested Bad Company 2 and it was 91FPS with single card and 87 FPS with CrossFire.

In many other games I didn't really create a solid benchmark, but I felt no difference in performance.

If this is what CrossFire can do, then I've just spent the most wasteful $200 ever.

yes, it's a Rosewill RX850, I'm sure 850 watts can cover a bit of CPU OC and graphics OC, it's not like I'm throwing a third 5850 in there for goodness' sake.

I ran my games in full screen, vsync off.

I am really confused as to why there isn't any performance increase in my games while 3DMark did show a good improvement, what's a good way to test if CrossFire is working while I'm in game? Is there supposed to be some kind of "CrossFireX" logo when I start a game if CrossFire is enabled?

Well I think CrossFire may actually be working, it's probably just that my CPU is bottlenecking everything. The 3DMark Vantage scores showed improvement and I wonder how it was different compared to actual gaming. Maybe 3DMark only stressed the GPU? Because I did select only the GPU tests and unchecked the other 2 tests involving the CPU.

But a quad core Pheom II at 3.8Ghz just sounds a little out of place when I say it's bottlenecking stuff, I thought this thing would be close in performance to an i5-750 at these clocks.

And I checked the power settings in CCC, it just lists what power plan I'm using, which is "high performance."

What's the Catalyst A.I. tho? There's an option to disable it, it's set on 'standard' as default, and the other option is to set it on 'advanced'.

I run at 1280x1024, my monitor is a bit old and only goes up to that high.

But you're saying that if I run higher resolution I'll actually get more FPS because the cards will be more utilized?