Lower 3DMark06 score with 4850 than with 3870+3850 Crossfire

hitokiri1

Distinguished
Jun 30, 2008
37
0
18,530
So, is it normal that i'm getting a lower score with my 4850 in 3Dmark06 than with my original Crossfire setup. (mind you things do look like they are running a heck of alot smoother) I was however shocked to see this:

3DMark06 4850 (basic 3dmark Settings) 13890 3dmarks

3DMark06 3870+3850 (on Asus P5K-E Wifi/AP 1x 16xPCIE/1x 4xPCIE in crossfire) 15500 3dMarks

Nothing has changed on my end in terms of CPU speed or Hardware. Also i have the 4850 running at : GPU Clock 660MHz Memory Clock 1080MHz running stable.

All other info on my PC is in the sig.
 
G

Guest

Guest
.... explain to me the point of your "upgrade"

your probably getting 5 - 10 % better performance with the 4850...

and 3dmark doesn't count... its heavily cpu bottlenecked
 
.... explain to me the point of your "upgrade"

your probably getting 5 - 10 % better performance with the 4850...

and 3dmark doesn't count... its heavily cpu bottlenecked
The crossfire wouldn't run as smoothly as the single card. Normally, I would advise against such an "upgrade" as well, but in this case, the price is low enough that it could be worth it, depending on how picky you are.
 

hitokiri1

Distinguished
Jun 30, 2008
37
0
18,530
.... explain to me the point of your "upgrade"

your probably getting 5 - 10 % better performance with the 4850...

and 3dmark doesn't count... its heavily cpu bottlenecked


Wellll the first reason is 1 graphics card performs almost as well as both combined with the added bonus that even if the score is showing lower the performance visually is noticeably improved. Noticeable enough that it justifies the upgrade.

Second reason would be that I returned my very recently purchased 3870 for the 4850 at the same price :p and gave my 3850 to my little brother :p.

Third, on the Asus P5K-E Wifi/AP Mobo the Crossfire support is gimped forcing the second PCIE slot to run at 4x mode. an obvious waste of performance albeit a noticeable improvement over a single card by a factor of about 15-20%.

And lastly, I can always throw another 4850 in the computer later on.

As for the other responses. Yes there is an improvement on how the games run. I also upgraded from a 19" 1440x900 display to a 22" 1680x1050 display. And at max settings everything still runs smoooth and sharp. Grid/Bioshock/COD4/yet to try Crysis as my copy seems to have corrupted.

The reason why I though this was going to be a much larger score improvement and was somewhat unsure about what was happening is that based on alot of the reviews i've seen it seemed that when they showed the 3dmark scores and vantage scores for this card versus the 9800GTX and + or the 8800 or the 3850 x 2 / 3870 x 2 or X2 it was for the most part either superior or on par (that being the X2 or dual 3870's in a proper crossfire setup.

I just wasnt sure why my weaker crossfire was actually outperforming in scores for 3dmark06 and vantage. compared to those of the newer 4850.

End of comment :p lol.
 

hitokiri1

Distinguished
Jun 30, 2008
37
0
18,530


Yeah :( My Q6600 doesn't compete.
 

hitokiri1

Distinguished
Jun 30, 2008
37
0
18,530
In the end I still feel that this was a worthwhile upgrade and there is always the option of adding a second 4800 series card down the road to improve it even further if i want to. Things just feel smoother and are way more stable at higher resolutions


Also something to note, I could never get Crysis to run properly in a crossfire setup the game would just be completely screwed up until i disabled crossfire. Then it would run fine. So this was also easiest way to get best performance out of a single GPU.