4890 not that great?

[xtkxhom3r]

i just got 2 4890's and oc'd the cores to 965 and ran 3dmark06.. i know 3dmark doesnt mean anything but i used it to compare my "upgrade" from 2 4870's to 2 4890's and well

with my 4870's at 820 core i got a little over 25k and with my 4890's at stock 850 i got 24k flat...now i thought these cards were supposed to be faster clock for clock?...

after i oc'd them to 965 i got about the same score as with my 4870's at 820..

so what does all this mean?? im guessing they arent worth the upgrade... now the only game i play is dawn of war and cod5 for now.. so those games were maxed out with the 4870's so i cant really compare ingame experience either

a friend told me that they probably perform a lot better at higher resolutions he suggested for me to run all the settings maxed out in 3dmark06 so i did...

i went into the settings and set the resolution at 1920x1080, anti-aliasing at 8, texture filtering at anisotropic X 16

and at those settings i got a flat 24k can someone else run those same settings with 2 4870's and post their results ?

[sandyduff]

il do the next best thing thing and do it with my x2... 2 mins.... alto i dont have an i7 so i doubt this could be credible testing...

Edit: Ok, with those settings i get just under 17k... altho my x2 is at stock clocks as well... its safe to say that the i7 plus 2 4890's in crossfire is a killer combo....

[xtkxhom3r]

what scores did you get in the sm 2.0 and sm 3.0?

[vixro]

You tested on 06 with top of the line cards. If you want a real comparison use Vantage. You are going to barely see a score fluctuation at that level when comparing those cards on the same CPU. The graphics are so low you are only bottlenecked by the CPU and you probably weren't even using the full power of the GPU.

[xtkxhom3r]

You tested on 06 with top of the line cards. If you want a real comparison use Vantage. You are going to barely see a score fluctuation at that level when comparing those cards on the same CPU. The graphics are so low you are only bottlenecked by the CPU and you probably weren't even using the full power of the GPU.

lol...i SERIOUSLY doubt im bottle necked by my cpu lol...come on now if i was running a c2d ok even a low end 775 quad but im running i7 at 4.0 with HT on and both lanes are running at 16x how the **** can i be bottle necked lol

[vixro]

Do you not understand? You are not bottlenecked by your CPU in anything other than old games. Lower and raise the clock on your cpu in 06 and watch your score go completely wonky, because it's so reliant on the cpu. Change the video card from a 4870 to a 4890, watch how the score doesn't even move! .....


This is why you should be using up-to-date benchmarks, so that you are not bottlenecked by the CPU in the score, but so you are only bottlenecked by the speed of your video cards. Vantage is going to notice less of a score change based on CPU speeds (once you cross the bottleneck), because the benchmark stresses the cards more and gives you a score based on the speed of your cards... not just the power of your CPU.



It would be like Using a 4890 on Diablo 2. Your FPS is not really going to move that much if you have any recent video card at all. But if you change your CPU speed, it's going to drastically change the score. You aren't necessarily bottlenecked by your CPU, the game is bottlenecked because your video card is already doing all it can. Even if that means your GPU is only working at 10%.

[xtkxhom3r]

haha now i understand what you were trying to say in the other post sorry for the miss understanding bro i would try vantage with my 4870's but i dont have them anymore lol but i will try with my 4890's ill post my results later