580m vs 6990 benchmarks

ok guys how about lets get some good consistent benchmarks to compare these cards.

i can update the first post with peoples scores and we should be able to get an idea of the relative speeds.

feel free to add benchmarks (and the settings to use) to the list as well as the results themselves. remember to post with stock clocks and max OC if poss ill add my results next week. no cheating!

YOU MUST HAVE TESSELLATION NORMAL IN THE CONTROL PANEL ON ATI CARDS

Also please state if you are using voltage mods for your scores.

(volt mod denoted as *)

Added 560 for guys getting the new card also

Single Cards Only

Mafia 2 (Steam Version): 1920x1080 - V-Sync off - Max Settings, Physx Off
FPS - Username - Clock Freq. - CPU
6990

46.5/47.4/47.2 - ewitte12 - 825/1115 - 2720
​

580m

47.9/48.4/48.8 - james_2k - 740/1900 - 2720
​

560m


Unreal heaven 2.5: 1280x1024 - Default settings (dx11)
FPS - Username - Clock Freq. - CPU
6990

47.9 - rover18 - 840/1150 - 2820

41.3 - rover18- Stock - 2820
​

580m

50.6 - james_2k - 750/1800 - 2720

43.4 - mrm2x - Stock - 2720
42.6 - james_2k - Stock - 2720
42.6 - Lanscaper - Stock - 2620
42.5 - vicious2500 - Stock - 2820​

560m

3dmark 2011
Score (Graphics Score) - Username - Clock Freq. - CPU
6990

4031 (3746) - ewitte12 - 830/1150 - 2720
3903 (3609) - TostioBandito - 830/1130 - 2720

3887 (3216) - Rover18 - Stock - 2820
3494 (3295) - ewitte12 - Stock - 2720
3491 (3193) - TostitoBandito - Stock - 2720​

580m

4422 (4181) - james_2k - 840/1900 - 2720*
4296 (4050) - james_2k - 810/1900 - 2720*
4084 (3826) - james_2k - 760/1900 - 2720*

3428 (3160) - james_2k - Stock - 2720
3413 (3150) - mrm2x - Stock - 2720
3400 (3157) - Lanscaper - Stock - 2620​

560m

3dmark Vantage
Score (Graphics Score) - Username - Clock Freq. - CPU
6990

16992 (15916) - inap - 880/1250 - 2820
16578 (15504) - The Revelator - 845/1200 - 2820
16411 (15520) - ewitte12 - 850/1100 - 2720
15775 (14691) - TostitoBandito - 830 - 2720

14339 (12978) - Rover18 - Stock - 2820
14220 (12938) - inap - Stock - 2820
13871 (12561) - TostitoBandito - Stock - 2720​

580m

18285 (17796) - james_2k - 850/1900 - 2720*
17952 (17412) - james_2k - 840/1900 - 2720*
17513 (16970) - james_2k - 810/1900 - 2720*

14425 (13265) - mrm2x - Stock - 2720
14151 (13323) - Lanscaper - Stock - 2620​

560m

 

Try and have the same CPU/RAM/hard drive configurations as well before posting your scores.

 

yeah true, while ram wont make much difference, we can put your cpu in brackets

 

Just simply use the GPU of vantage / 3dM11, then you don't need to differ between the different CPUs and it's more accurate anyway.

 

Be sure to note if your looking for single or dual card benches. On the other hand a smart guy like me will assume your looking for single card benches sine the R2 doesn't officially support these cards and your in the m17x subforum.

 

Agreed. GPU scores are best for comparing cards only.

 

i7 28020qm, 8gb 1333mhz, single 6990m stock clock 3dmark vantage: 14339

and 3dmark11: 3517

 

I can't wait what to see what the scores will be when the drivers are actually released for the 6990.

 

11.7 was just released on amd's site. I used 11.7 for those tests.

 

use 11.8 beta I'm using them and they work like a charm and offer a fps increase in multiple games.

http://support.amd.com/us/kbarticles/Pages/Catalyst118DriverPreview.aspx

 

see reference clevo oem 580m oc result
http://forum.notebookreview.com/sager-clevo/595787-extreme-gtx580m-overclocking.html

 

updated scores,

disregarding non m17xr3 benchmarks to level the field as much as possible

 

I get 1608 (graphics) Vantage on my current laptop (SB integrated) lol and 19237 for the graphics score on my 560Ti.

 

overall scores please. too complicated otherwise and the cpus arnt that different

 

My 560ti scores were at 4.4Ghz on a 2600k Overall on SB laptop was P2208 on a 2720qm QS cpu.

 

I respectfully disagree, but this is your thread. Here is why GPU should be looked at.

Take this score of 14072 by 580m. Now compare it to this 14051 by a 6990.

By just looking at the overall scores we conclude the 580 and the 6990 are neck and neck, equal in most respects.

HOWEVER...

If we look at the GPU and CPU scores, we see...

the 580m got 13242 GPU, with a CPU score of 17331 (due to the 2630qm).

The 6990 got a 12675 GPU and CPU of 20834 (due to the 2820qm).

The stronger CPU on the 6990 rig brought the overall score up to equal the 580s overall score, but in fact, the 580m is looking to be 4% faster. If both rigs had the same CPU, the 580m would have scored a higher overall score. We wouldn't see this if we only looked at overall scores.

If we are going to make honest comparisions between GPUs, we need to look at GPU scores, not overall.

 

true. but id prefer not to use 3dmark at all to be honest. and get some nice game benchmarks going.

and anyway, we only have 2 scores so far!

but ok, ill add graphics scores when people actually provide them

ill be doing some 580 benchmarks on wednesday next week anyway

 

I will be benching the heck out of my 6990 once it arrives.

 

cool

 

Here's a 3dMark11 run for the my 6990M (2820QM) at stock settings and using the Dell drivers that came installed on the R3 -- P3,860.



And with a mild overclock (@ 825/1025) -- P4,336.

 

That's an absolutely beast Revelator congrats!

 

@The Rev: Wow, you got your system, cool!! Congrats! Very great scores, the 6990m doesn't stop to impress.

 

Thanks guys, but based on very limited experience, I have my doubts about whether these 6990M's are going to be much different than a well o/c'ed 6970M. Maybe some better drivers will free it up some more. Also having fan trouble in this R3. They are not adjusting to increased temps very well and just seem generally erratic; sometimes it doesn't seem to even see the fans. HWiNFO64 helps some. Technically it only sees 1 fan (you can in effect select which one by changing the trigger sensor), but them it brings the other fan with it to some degree. Plus, I can't get inside yet because the battery release isn't working, and I don't want to force it for fear of breaking it. Ought to be a very civilized notebook with decent performance, but this one is a little rough around the edges so far.

 

added revelators scores and added the graphics score in brackets.

wheres the 580 results!

+ spent a while visually improving the table, but without your results it looks crap!

 

The peope with cash for the 580 probably went for m18x and did SLI/crossfire

 

no they did not

 

deleted...............

 

updated, cheers

 

my 3dmark11 score is slightly lower than the other 6990's that were just posted like revelators's 3860. I did my tests with 11.7b driver. With 11.8b, my scores are 10-15 points lower. 3dmark11 11.7b 3517

and 3dmark11 11.8b 3502

 

what settings did you guys use in catalyst?

 

here some results from my returned M17XR3 (im getting the FHD 3D-Display)

anything is factory default (like OS, clocks etc.), i only disabled the background-services like alienfx/fusion/comandcenter etc.

nVidia Verde 280.19 (beta)

3DMark06 - 19875

3DMark Vantage - P14425 | GPU13265 | CPU19560

3DMark11 - P3413 | GPU3150

Unigine Heaven 2.5 - 43.4 FPS | Score 1092


keep in mind that these scores are with optimus enabled. rumor has it that non-optimus systems score slighty higher (about 5%)

 

thanks updated

 

Can I use the GTX 580m in SLI in my M17x R2? It would be good with photoshop and solid works I presume? I need to do some design works besides playing games. Please advise. Sorry for going out of topic I just didn't feel it is necessary to make a whole new thread for a simple question.

I don't plan on overclocking the cards because of power Issues such as with the 6970s.

 

oi!

go make a new thread, its hard enough to keep track of people posting results

 

I'll be getting my laptop Tuesday so when I get it I'll post some benches. I have a 2820 & 580.

 

Working on finding that out... 8D

 

Apologies to you Sir!

@ichime: I'll wait then.

 

no worries,

in other news, more benchmarks! (and from people with normally priced cpus please! )

 

I have the same problem! If you found something that can explain this 300 difference in our score it will be great.

 

A 3dMark06 run at stock clocks using stock drivers -- 20,338 (6990M/2820QM):

 

It's probably the driver settings in Catalyst Control Center, mainly for Tesselation and Cat A.I./Texture Filtering. Go to CCC>Gaming>3d Application Settings and set as shown below. You'll probably find the 300 point difference in 3dMark11 right there. Has relatively little impact in other tests.



The screenshot is from the Cat 11.8 Preview version of CCC. Yours may look slightly different, but the settings will be the more or less the same.

 

thanks, i will have to change those settings and run benchmarks for stock clocks again. here is gpu vantage score of 15052 for 6990m @ 840/1050

 

Thanks The Revelator, now I have my 3860 score for 3dmark 2011...
I was a bit worried that my 6990m was not as good as yours, but it seems to be working just fine. I even tried a mild overclock (@ 825/1025), and it was quite good too.

 

Good show. These 6990m's seem to all be performing about the same. Here's another 3dMark11 run at 825/1100.

3dMark11 -- P4417 (6990m/2820qw)

 

yessss 3dmark11 stock clocks 3887; gpu: 3620.....sooo yes with catalyst the way you showed me, our cards are performing just as well. I will try it out at 825/1100 and post.

 

haha same overall score of 4417 at those clocks. gpu is ten less, but physics is higher. no doubt its the same performance 6990m @ 825/1100 with i7 2820

 

What do the 7/29 drivers on Dell's website compare to?

 

Those are modified Cat 11.5/11.5b's, which are a solid, well-performing set in reference form.

 

updated stock 3dmark 11 score for myself.

can some 6990 guys download unreal heaven 2.5 and run it at default settings 1280x1024?

 

Awesome scores Gents, I am especially happy got see the 825/1100 OC. I Hope mine does that well.