3DMark06 Scores...Amazing...Part 2

Page 16 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Please post your best 06score here. Trying to see what setups work the best.

Here is my best at 3.6Ghz :

new.jpg


3.85Ghz :

newer.jpg


Ryan closed videl's thread so i decided to re-open 😉


Much appreciated :)
 



965be c3 @4.0ghz + 4870 512mb @775/1030 + 4gb 1066-DDR2

Obviously not a GPU match, but CPU should be relevant.
My CPU score is a little higher than yours, but that may be due to my NB being slightly OC'd to 2200.
SS02.jpg


 


The problem is likely your CPU. The GPU score is slightly higher than my two 5870's in CF, but I'm running an i7, which scales much higher with multiple GPU's.

Here is a benchmark of the two different CPU's when overclocked to the same frequencies: http://www.xbitlabs.com/articles/video/display/radeon-hd5870-cpu-scaling_4.html#sect1

Notice how the performance increases a lot more when the GPU is not bottlenecking. With two 5870's, the i7 advantage becomes a lot more apparent. With three 5870's, it's going to be huge. With a faster CPU, you'd probably see higher GPU scores as well as CPU scores.

From the same artical with Vantage scores: http://www.xbitlabs.com/articles/video/display/radeon-hd5870-cpu-scaling_3.html#sect2

Yours is a good chunk higher than their CF setup, so yours looks about right.
 


Damn you 260!~

Will know next week if I am ordering a 5850 or 5870, and possibly but unlikely, may be able to slip into a 5970. I don't have the LCD real estate to justify a 5970, (22" 1920x1080), so thats almost completely out of the question, unless the alpha wannabe in me jumps hard enough on my nuts about it.
 


Thanks for those links. good stuff. I wish they had 2560*1600 though. That's what I game at. I guess my scores are right where they should be. If memory serves me right, when looking into getting this Proc over the 920, I believe that the benchmarks @ 2560 evened out with crossfire. Let me see if I can find some..
 


or wait till Fermi 😉
 


Yes, that is likely going to happen, as the higher resolution chosen, the more stress is put on the GPU's and usually becoming the bottleneck. So you saved some money and get the same performance at that resolution. However, in vantage, they also test the CPU seperately, so it shows a lower score. That and it's performed at a low res.
 


Haha, yeah, I love that ol Pentium II.

Current specs

PII 965 @ 4.0 GHZ 200x20 1.5v
8GB OCZ PLatinum 1066 Mem
1 TB WD black
1 250GB Seagate Barracuda 7200.1
1 500GB WD Blue
Plus a few 1TB external WD drives (might get a DROBO to replace them soon though)
1 Powercolor Watercooled 5870 @ 875
2 XFX FAn cooled 5870s @stock 850
Asus M3A79-T Deluxe mobo
Corsair Obsidian 800D case
1200W PC Power and Cooling PSU
1 Loop 360* radiator with MP655 pump for water cooling.
Monitors: 1 3008WFP, 2 x26" VP2650wb, and 1 AcerH33H
Sound: Bose Companion 5 speakers
 


Right. it looks as though the Physx (or lack of a physx card) is the problem there with the CPU score, and I guess the benchmarks don't really mean anything anyway, I just wanted to compare it to what I should be getting. My buddy has a 9800GT that he might bring over tomorrow and I might switch out one of the 5870s and throw that in as a dedicated physx card to give it a run and see what I get...