• 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.

7970 CF issue

Associate
Joined
22 Dec 2011
Posts
55
I've been thinking my CF 7970's haven't been performing as well as they should for a while (based on other people's in game FPS compared to mine). It wasn't until I ran the new Sky Diver benchmark that the size of the issue was made clear.

http://www.3dmark.com/compare/sd/2134856/sd/2132818

According to the comparison linked above I'm getting 25% less performance out of very similar specs. The graphics score in particular is alarming as I'm getting 50% less?

Ran Afterburner, it looks like usage on the 2nd GPU drops to ~50% as soon as the first part of the test stops, which matches up with the results I'm getting, but don't know why. :confused:

Any help you guys could offer would be amazing.
 

I updated to latest version of Afterburner and went through the set up in the link. Now have even GPU usage throughout the benchmark tests but only got a tiny bump in the score? Still getting destroyed on the Graphics Score/Tests.

Read that my MB might cause a ~10% performance hit due to the speed of the 2nd card slot but doesn't explain the 50% performance hit I seem to be taking.

Any ideas?
 
Have you tried overclocking your CPU to 4.5ghz+? Could be getting a bottleneck!

Unfortunately the chip isn't a great, getting to 4.3 required pretty high volts from what I remember. Will probably have another go at it at some point but the guy in that comparison is only 100Mhz faster and getting 2x my scores/fps in the graphics tests so don't think bottleneck is the issue.

When I first installed the second card I put it in the first slot but swapped them round as my PC was crashing due to the new card overheating. Everything seemed to be running fine once swapped so kinda forgot about it. Now thinking that there might be an issue with the newer card? Will benchmark the two cards individually tonight to see if there's something up.
 
Have benched both cards on their own and they outperform the crossfire setup. Have disabled ULPS and have the crossfire logo in the corner so should work but is instead negatively impacting my performance.

Had a closer look at Afterburner and noticed that the core clock on GPU 1 was at 501Mhz for most of the bench (GPU2 was at 1150 as its meant to be). Why would the core clock speed be so low on the primary card?

I'm thinking that this is causing my crossfire setup to run as 2x 501Mhz clock speed, rather than 2x1150Mhz. This would match up with the slightly weaker results when in crossfire compare to when I run the cards individually (where both cards run at full speed).

Will keep looking for a solution but if any one knows what the problem is I could really use the help!
 
From what I can find seems like GPU 1 is getting stuck in a 2d hardware acceleration mode?
Seem to be able to get around the issue by creating profiles in RadeonPro and force max settings when specified .exe's are running. Not ideal but seems to do the job.

http://www.3dmark.com/compare/sd/2142528/sd/2142670

Edit: this only worked once, has gone back to clocking at 501Mhz on gpu1. Really annoying!
 
Last edited:
Back
Top Bottom