CROSSFIRE 4870 LOW FPS

OK I HOPE AND PRAY SOMEBODY CAN HELP ME IM GOING NUTS HERE... IM RUNNING A ONE WEEK OLD DELL XPS 630I WITH TWO ATI 4870 HD IN CROSSFIRE. I HAVE A DELL NFORCE 650I MOBO.. (THATS RIGHT AND NFORCE MOBO THAT SUPPORTS CROSSFIRE ITS A CUSTOME BUILD BY DELL)
ITS GOT PCIE X16 BUT IN CROSSFIRE THE BRIDGE IS SPLIT TO X8 DID RESEARCH ON THAT AND FOUND ITS PRETTY COMMON TO SPLIT THE BRIDGE.
700WATT POWER SUPPLY. 8 GIGS OF 1066 DDR2 RAM AND WINDOWS VISTA 64 BIT.. IN 3D MARK IM ONLY BENCHING 16770 SCORE IN CROSSFIRE.
NOW MY REAL QUESTION AND I FIND THIS WEIRD IN CRISIS WITH EVERYTHING SET TO ULTRA 1600 X 900 RES AND 4X ANTI ON IM GETTING ABOUT 20/30 FPS IVE SEEN BETTER FPS ON A SINGLE CARD. IF I CHANGE THE SETTINGS TO HIGH AND AND A LOWER RES IM ONLY GETTING 30/40 FPS.. IVE OC MY CARD FROM 750 GPU TO 850 AND DAYS OF TESTING ITS STABLE TRUST ME AND MY RAM FROM 900 MHZ TO 1200 MHZ. IVE RAN CRISIS AT STOCK 750/900MHZ AND MY 850/1200 MHZ AND NOT ANY DIFFERANCE I COULD TELL... CROSSFIRE SAYS ITS WORKING BUT IM NOT SURE. CAN ANYBODY HELP ME.. PS I EVEN TRIED FLASHING THE BIOS ON BOTH CARDS TO A NEW ONE IT CHANGED THE GPU CLOCK TO 800 AND RAM TO 925 DIDNT SEE AND BETTER RESULTS.. THANK YOU

Yeah, I'd worry about learning the ability to shut off the caps lock button before going and messing with several hundred dollars of pc equipment.

i just need some help guys

yes whats wierd is even at such low fps it runs smooth.. ive seen bench marks with only one 4870 and better fps though i just dont understand. my 3d mark score seems a bit low to

[quotemsg=1950561,8,401516]What's your 3d mark score with just one 4870?[/quotem



Didn't run a bench with just one will do that when i get home and post score later. thx

I've never heard of a single 4870 running crysis at that resolution using 4x. Plus, crossfire support on Crysis blows. Try turning off AA if you're having a problem. Try some other games too.

That's called graphical horsepower mate. I doubt it would ever chunder even on the now infamous snow level but you are not going to see triple figures this side of an Nvidia buyout.

last time I ran 3dmark 06 I got like 16k, with a overclocked Q6600, never ran it in crossfire. i m getting two HD4890s this week so yeah lol

ok so i turned off crossfire and ran 1 ati 4870 and got 5/8 more fps in crysis then i did with crossfire on wtf lol.... 3d mark 06 score with one card was 15200ish only about 1500 less then in crossfire... is my mobo bottlenecking these cards when theyre in crossfire?

3.0 Ghz Intel core 2 Q9650 quadcore @ 3.1ghz
8gb of ddr2 1066 mhz ram
2 ATI 4870's running crossfire @ 850gpu mhz and 1200 ram mhz
nvidia nforce 650i made by dell to support crossfire (weird huh)
2x640 gb hdd

Well, do keep in mind that, on a base of 20 FPS, 5-8 FPS represents a 25%-40% improvement

Seriously though . . .

Nah, probably not. There's nothing there to bottleneck Check what a 295 does with it:

http://www.hardwarecanucks.com/for [...] ew-14.html

It ain't that good a game . . . turn down the details . . . beat it . . . put it back in the drawer.

So, no one who knows anything about this decides to post, i know some people are trying to help but other are just either trolling or misinformed.

Crysis, at least using the gpu benchmark shows practically 100% scaling, 2 x the performance with crossfire.

Average fps with crossfire at 19 x 12, 4 x AA, very high i get around 30fps.

With crossfire off i get around 15/16fps.

And yet no one seems to own the setup nor have tested the game using modern drivers and may confuse the OP, the comments about the game not scaling well or crossfire being faulty were not helping anyone.

You cannot really get a game that demonstrates crossfire benefits better than crysis as shown by my results.

If the OP is not getting that sort of performance his problem does not lie with crysis but his setup.

I see this often, any problems they are based on crossfire or in this case crysis being made badly. It is more often the case that there is an issue with the OP's comp.