GTX 670 FTW 3D mark results

[ridd21]

Hello, I'm wondering if this score is right for my card. I upgraded from a GTX 560 SC to the 670 ftw and when I installed it yesterday it really hasn't seemed like much of a upgrade =/ I run WoW around the same as I did on my 560 Sc granted I know wow is more cpu dependent but I would think I'd gain a very noticeable boost from a near $400 GPU that being said I ran Company of Heroes and again didn't seem like a big jump in performance. On battlefield 3 now it did stay above 60+ at pretty much all times on all settings being ultra but I still wasnt sure that the card is performing at what it should so I ran a 3D mark directx11 test on it and seeing I used the basic edition of it running its at 1280x720 because it wouldn't let me do 1920x1080 because its the free version of the software but after all the tests it ran wasn't as high as most people it seems. Now Id figure at 1280x720 this card would destroy the scores but from looking at the graph it really didn't. my monitor is 1920x1080 by the way so any help would be great not sure if the card is faulty or not. I also set the settings in the invidia control panel for it to be max performance so that shouldn't be causing it not to score as well as others and it is factory OC'ed I haven't tried to push it any further didn't think I would need to for a few years to be honest seeing from everything I'd read from reviews on it their isnt a big difference between it and the 680 in performance. so again any help would be great! thank you very much


http://www.3dmark.com/3dm11/5001150

 

[disolitude]

GTX 670 should almost double the performance of the GTX 560. I suggest run GPUZ and take a look at your GPU usage while gaming. It could be a driver issue or something else that it bottlenecking your system...

 

[ridd21]

GTX 670 should almost double the performance of the GTX 560. I suggest run GPUZ and take a look at your GPU usage while gaming. It could be a driver issue or something else that it bottlenecking your system...

That's what I would figure but right now it hasn't seemed that way I'm gonna download GPU-Z and see I'm running the newest drivers and cleared out my old 560 drivers. I would think for a Card that cost this much that it would make older games fps look like a joke especially when I read a guys benchmark on it for COH that he has 140+ fps at all times max settings and my max was 70 with a avg of like 58.

 

[jarlsuki]

My MSI 660ti card gets 9300+ in Graphics Score, so it does sound like there's a problem. You should be closer to 10.000. Having BF3 on Ultra and 1920x1080 and getting 60+ fps is not bad though, but as Disolitude said, it would be a driver og bottleneck issue. Have you tried re-installing your drivers?

Edit: we were typing at the same time, just ignore the parts you already answered

 

[ridd21]

My MSI 660ti card gets 9300+ in Graphics Score, so it does sound like there's a problem. You should be closer to 10.000. Having BF3 on Ultra and 1920x1080 and getting 60+ fps is not bad though, but as Disolitude said, it would be a driver og bottleneck issue. Have you tried re-installing your drivers?

Edit: we were typing at the same time, just ignore the parts you already answered

haha yea I used driver sweeper cleaned out the drivers actually twice did it again this morning and did a fresh install of them. I'm running BF3 at the moment in campaign watching GPU-Z it has the core clock going up to 1162 and memory clock going up to 1552 max GPU load at 99%.

 

[bystander]

The 3dmark11 score shows a proper score for your setup. It even says your system is performing normally. I don't see where the concern is.

When looking at the scores between the 2 different cards, focus only on the Graphics score. The total score (the one with the P in front) includes both CPU and GPU performance, and the CPU did not change with the upgrade.

 

[bigcyco1]

What do you get in Crysis Warhead look about right ? http://www.youtube.com/watch?v=2bsxc3YsGak&feature=channel&list=UL

 

[bigcyco1]

The 3dmark11 score shows a proper score for your setup. It even says your system is performing normally. I don't see where the concern is.

When looking at the scores between the 2 different cards, focus only on the Graphics score. The total score (the one with the P in front) includes both CPU and GPU performance, and the CPU did not change with the upgrade.

+1 i agree

 

[cooldudesubho]

To give u an idea my stock gtx 670 scores 8800 3d mark 11.So your FTW version should score about 9000.8800 is about the stcok score of gtx 670 according to reviews.

 

[ridd21]

What do you get in Crysis Warhead look about right ? http://www.youtube.com/watch?v=2bsxc3YsGak&feature=channel&list=UL

Anyway to test it? because I do not own Crysis warhead

 

[bigcyco1]

Anyway to test it? because I do not own Crysis warhead

no sorry

 

[ridd21]

no sorry

I got this card to run games like Crysis 3 hopefully maxxed doubt that will happen but ya know kinda the next Gen stuff with what is coming out with the new Unreal engine and such. I'm just wanting to make sure that I will be ok for a few years.

 

[osoben]

I did the same upgrade went from evga gtx 560 sc 2gb to 670 ftw and I can tell a big difference all three crysis games played much better frame rates and I replayed witcher 2 and metro 2033 again much better frames was worth the upgrade

 

[bigcyco1]

I got this card to run games like Crysis 3 hopefully maxxed doubt that will happen but ya know kinda the next Gen stuff with what is coming out with the new Unreal engine and such. I'm just wanting to make sure that I will be ok for a few years.

Yeah i know what you mean that is why i bought two 670's :lol:

 

[RussK1]

I'd say something is wrong...

http://www.3dmark.com/compare/3dm11/5001150/3dm11/4994165

My result isn't valid because my drivers aren't approved (12.11 beta)

http://www.3dmark.com/3dm11/4994165

Extreme:

http://www.3dmark.com/3dm11/4994259

 

[bigcyco1]

http://www.overclock.net/t/1265110/the-gtx-670-overclocking-master-guide Try this tutorial, it goes over everything needed for GTX 670, and takes into account that each chip is different. I followed this and got great results.My best so far with one card

1418 Mhz

 

[bystander]

I'd say something is wrong...

http://www.3dmark.com/compare/3dm11/5001150/3dm11/4994165

My result isn't valid because my drivers aren't approved (12.11 beta)

http://www.3dmark.com/3dm11/4994165

Extreme:

http://www.3dmark.com/3dm11/4994259

I don't know if those are doctored, something with the newest beta drivers or you have extreme OCing, but those are WAY higher than you should get.

Go around and look at the reviews with 3Dmark11 scores, and in the high 7000's to low P8000's is all you find with that card.

http://www.overclock3d.net/reviews/gpu_displays/his_hd7950_review/4
http://www.hardwarecanucks.com/forum/hardware-canucks-reviews/53456-gigabyte-hd-7950-windforce-msi-hd-7950-twin-frozr-iii-review-5.html
http://www.vortez.net/articles_pages/powercolor_pcs_hd_7950_review,17.html
http://www.neoseeker.com/Articles/Hardware/Reviews/AMD_HD_7950/7.html

 

[RussK1]

I don't know if those are doctored, something with the newest beta drivers or you have extreme OCing, but those are WAY higher than you should get.

Go around and look at the reviews with 3Dmark11 scores, and in the high 7000's to low P8000's is all you find with that card.

http://www.overclock3d.net/reviews/gpu_displays/his_hd7950_review/4
http://www.hardwarecanucks.com/forum/hardware-canucks-reviews/53456-gigabyte-hd-7950-windforce-msi-hd-7950-twin-frozr-iii-review-5.html
http://www.vortez.net/articles_pages/powercolor_pcs_hd_7950_review,17.html
http://www.neoseeker.com/Articles/Hardware/Reviews/AMD_HD_7950/7.html

Nothing doctored... default values in CCC and OC'd to 1000/1450... maybe it's the new Virtu.

 

[bystander]

Nothing doctored... default values in CCC and OC'd to 1000/1450... maybe it's the new Virtu.

http://i1081.photobucket.com/albums/j357/russwood8814/20121115_063253.jpg
http://i1081.photobucket.com/albums/j357/russwood8814/heaven_2012_11_20_21_51_56_047.jpg

What do you mean by Virtu?

 

[bigcyco1]

Nothing doctored... default values in CCC and OC'd to 1000/1450... maybe it's the new Virtu.

http://i1081.photobucket.com/albums/j357/russwood8814/20121115_063253.jpg
http://i1081.photobucket.com/albums/j357/russwood8814/heaven_2012_11_20_21_51_56_047.jpg

WTF!How does a 7950 beat a 670 your score is higher then mine

 

[bystander]

WTF!How does a 7950 beat a 670 your score is higher then mine

I think he has some settings in the CCC turned down, personally. I cannot find anyone with scores in that range with out crossfire. Those Heaven 3.0 scores are 50% higher than a 680. Those 3Dmark11 scores are also way higher than normal. Either those beta drivers turned on some serious power never present before, or there are some setting changes in the CCC.

 

[RussK1]

What do you mean by Virtu?

Lucidlogix -----> GPU Virtualization Software

Maybe offloads some of the workload onto the IGP. Also repeatedly beats a 6990 (mine) in every test... same settings, everything.

Funny, the first thing that comes to peoples minds is cheating... no doubt ones with nVidia hardware.

or there are some setting changes in the CCC.

Which one's? There aren't many atleast not like all the messy odds and ends in the nVidia control panel. Needless to say, everything is set to "application controlled" and Catalyst AI and Texture Filtering is set to high quality... AA method is set to MSAA.

Also, don't you think 3DMark would know if it settings had been over-ran? Disable tessellation via control panel and the score will be void due to tessellation modification.

 

[bigcyco1]

I think he has some settings in the CCC turned down, personally. I cannot find anyone with scores in that range with out crossfire. Those Heaven 3.0 scores are 50% higher than a 680. Those 3Dmark11 scores are also way higher than normal. Either those beta drivers turned on some serious power never present before, or there are some setting changes in the CCC.

o.k. that makes sense.

 

[bigcyco1]

Lucidlogix -----> GPU Virtualization Software

Maybe offloads some of the workload onto the IGP. Also repeatedly beats a 6990 (mine) in every test... same settings, everything.

Funny, the first thing that comes to peoples minds is cheating... no doubt ones with nVidia hardware.

I never said you cheated

 

[bystander]

Lucidlogix -----> GPU Virtualization Software

Maybe offloads some of the workload onto the IGP. Also repeatedly beats a 6990 (mine) in every test... same settings, everything.

Well then that means your system isn't something the OP can compare. That is not a 7950, that is some sort of hybrid not at all the same, though from what I've read on that software, it was rarely giving benefits, but at times gave large ones. It would be interesting on how it behaves in a game by game analysis, maybe they've got it working a lot better than a few months ago.

Anyways, this is off topic and really of no value to the OP's non-problem. His scores are the same or higher than anything I found online and 3dmark11 says his scores are normal, and future mark don't give easy passes.