Overclock.net banner

670GTX FTW's Sli'ed 3Damrk11 score-

3K views 41 replies 11 participants last post by  K62-RIG 
#1 ·
With the new R313.96 driver release I saw large performance increases in many games as well as benchmarks
smile.gif


http://www.3dmark.com/3dm11/5739544

Out of curiosity what are the 7970ghz editions doing crossfired? I have a hard time believing they are able to hit 20,700 gpu score
smile.gif
 
See less See more
2
#2 ·
Quote:
Originally Posted by SeanJ76 View Post

With the new R313.96 driver release I saw large performance increases in many games as well as benchmarks
smile.gif


http://www.3dmark.com/3dm11/5739544

Out of curiosity what are the 7970ghz editions doing crossfired? I have a hard time believing they are able to hit 20,700 gpu score
smile.gif
Nice work mate!

what were the cards clocked to?

any tweaking in the nvidia control panel or just a straight run?

I will bench mine tonight using the beta driver.
 
#3 ·
7970s are still getting more....

they're still beating my score on most counts.

Nice score regardless though
thumb.gif
\

edit:

tsm running 2 way cf

http://www.3dmark.com/3dm11/5757528
 
#5 ·
Quote:
Originally Posted by trippinonprozac View Post

Nice work mate!

what were the cards clocked to?

any tweaking in the nvidia control panel or just a straight run?

I will bench mine tonight using the beta driver.
I have both 670GTX FTW's running at 1293mhz core/3706mhz memory( basically +112 to the core and +603 to memory)
Just changed power management to perfer high performance in control panel.
313.96 driver
 
#7 ·
Quote:
Originally Posted by SeanJ76 View Post

^^This guy is one of 4 people breaking 20k with 7970ghz CS........he's also using the number 1 cpu on the market.
CPU doesn't matter unless it's a bottleneck... a 2500k would result in the system getting the same GPU score
 
#9 ·
Quote:
Originally Posted by deafboy View Post

CPU doesn't matter unless it's a bottleneck... a 2500k would result in the system getting the same GPU score
Wow you have a lot to learn about bench marking if you think cpu's don't matter when it comes to gpu scores LOL. My score would be altered significantly if I was using a 3930/3960 at 6ghz. vs 5ghz...........
 
#10 ·
Quote:
Originally Posted by SeanJ76 View Post

Wow you have a lot to learn about bench marking if you think cpu's don't matter when it comes to gpu scores LOL. My score would be altered significantly if I was using a 3930/3960 at 6ghz. vs 5ghz...........
Are you kidding...lol.

CPU will only effect the GPU score if the CPU is bottlenecking the GPUs. The benefit of a beefy CPU in 3DMark11 will only help with Physics and combined score. How much different is your score from mine, mine being with a 3930k at 4.8GHz. Very small, and that difference likely comes down to GPU clocks being different.
 
#11 ·
Quote:
Originally Posted by SeanJ76 View Post

Quote:
Originally Posted by deafboy View Post

CPU doesn't matter unless it's a bottleneck... a 2500k would result in the system getting the same GPU score
Wow you have a lot to learn about bench marking if you think cpu's don't matter when it comes to gpu scores LOL. My score would be altered significantly if I was using a 3930/3960 at 6ghz. vs 5ghz...........
The physics score might (will) be different with a better CPU but unless the CPU being used to do the benchmark is bottlenecking the GPU(s) then the GPU score would be almost exactly the same regardless.

deafboy is right
 
#12 ·
Quote:
Originally Posted by deafboy View Post

Are you kidding...lol.

CPU will only effect the GPU score if the CPU is bottlenecking the GPUs. The benefit of a beefy CPU in 3DMark11 will only help with Physics and combined score. How much different is your score from mine, mine being with a 3930k at 4.8GHz. Very small, and that difference likely comes down to GPU clocks being different.
Like I said you have no idea what your talking about.......
 
#13 ·
Quote:
Originally Posted by SeanJ76 View Post

Like I said you have no idea what your talking about.......
Go on, tell me why I have no idea what I am talking about. This should be interesting.
 
#15 ·
lol. Go for it.
 
#19 ·
http://www.3dmark.com/3dm11/4015758

That was @ 1.2ghz and 1635 memory on 12.6 Beta Drivers (20114 Graphics Score), as far as I'm concerned 1200 is probably around a little above average for CFX setups and considering those are on old drivers. I'm sure there are a lot of people that hit over 20,000 graphics score fairly easily, not sure why GHZ edition matters either, I thought it was shown they weren't binned and if you are going for over 20k GFX score you're going well over 1,000 on core anyway.

Just a quickly holy god edit I re read all of the other posts, I think this person is really really in love with their hardware considering the first post. Saying there's only four people who have graphics scores over 20k and saying how much CPU effects GPU score...
 
#20 ·
Quote:
Originally Posted by grunion View Post

CPU makes very little difference with 3DM11 gpu score, 3930K vs 3570K.
And with just a 300 point difference (less than 3%) it's even harder to tell whether the CPU played a role at all given that the version of the benchmarks are different, other tasks running, various latencies, different memory, and 3DMark being inconsistent in runs in general.
Quote:
Originally Posted by kfxsti View Post

LOL done
lol. nice
cheers.gif
 
#21 ·
Quote:
Originally Posted by deafboy View Post

And with just a 300 point difference (less than 3%) it's even harder to tell whether the CPU played a role at all given that the version of the benchmarks are different, other tasks running, various latencies, different memory, and 3DMark being inconsistent in runs in general.
lol. nice
cheers.gif
The RIVE run is with minimal processes, it's a test bed.
 
#22 ·
Quote:
Originally Posted by grunion View Post

The RIVE run is with minimal processes, it's a test bed.
Good to know, so that could certainly be part of the boost (among other things I am sure).
 
#23 ·
Just ran 3dmark, didn't feel like changing clocks, 1200/1675 ended up with 21439 graphics score, you have a 93mhz core clock advantage in addition. I'm sure that you could hit 20k GFX score with like 1100 clocks if I'm judging the difference between mine and TS's cards is correct. I've never heard of a card that couldn't hit 1100 if it can overvolt.

http://www.3dmark.com/3dm11/5780497
 
#24 ·
Quote:
Originally Posted by SeanJ76 View Post

With the new R313.96 driver release I saw large performance increases in many games as well as benchmarks
smile.gif


http://www.3dmark.com/3dm11/5739544

Out of curiosity what are the 7970ghz editions doing crossfired? I have a hard time believing they are able to hit 20,700 gpu score
smile.gif
Don't know how the Ghz editions are doing, looks like there are more non-ghz cards that overclock higher.
But to give you an idea, the single card world record is a 7970 with a gpu score a little over 19k (world record nvidia card has a bit over 17k gpu score.)
There are some 2 x 7970 3d11 scores with 30k + graphics scores.
Nvidia kinda lost the benchmarks battle this generation, sounds like it will turn around again when nvidia brings on the Titan!
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top