• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

QuadFire Unigine Score for fun...

tangoseal

[H]F Junkie
2FA
Joined
Dec 18, 2010
Messages
9,778
Just wanted to post something for fun.

This is on my P67 powered Asus Max IV Extreme B3 with NF200.

1x 6990 (2GPUS)
2x 6970 (1 GPU Each)

4x GPU total for a sum of LOTS OF POWER.

6990 and 6970 are watercooled, permanent cards.
The second 6970 is air cooled but Im gonna block it. Going to run socket 2011 in 2 months anyways when it comes out and Cayman is still uber powerful.

The unigine score below was performed while running HUGE 122GB file transfer operation from my NAS to my hard drive in the background. I wanted to see what performance would be like while actually using my computer for something else at the same time. I pulled my raid card in sig to free up the slot to try quad fire so the file copy op is using a lot more CPU than with the raid card installed.

QF.JPG
 
I will do some other benches later while not running anything other than normal services in the background.
 
I have the same board with three 2GB unlocked 6950s installed. I've been wondering if selling one of them to replace it with a 6990 would give a decent performance boost but I can't find any examples. Can you please do an FPS comparison between three GPUs and four? Crysis 2 would be great as would Metro. Thanks and SICK setup!

BTW, what resolution do you normally run, and what is your CPU clocked at (as I understand it, you would need to clock at 4.6 minimum to remove the bottleneck with all that horsepower)?
 
Puny human... :p

Try it with 16X AF and 4X AA with extreme Tessellation!

Will do! Later tonight sometime. Also I'm using nf200 so I'm sure I'm having some bandwidth issues. I can't wait to put these cards on socket 2011 and 990fx with 8 real 3.0 lanes per card and 16 for the 6990. (edit* if there is going to be 3.0)
 
Last edited:
AFAIK PCI 3.0 wont be released with socket 2011?

To my understanding it's just a bios update???? I dont know how but that is what I gather.

And for the longest time Intel was advertising PCIe3 but now they have claimed it wont have it.

We will just have to see. Anyways these cards are not 3.0 certified. But having a true 8/16 real lanes per card direct to the processor should make a difference.
 
I have the same board with three 2GB unlocked 6950s installed. I've been wondering if selling one of them to replace it with a 6990 would give a decent performance boost but I can't find any examples.

From what I have seen in the past crossfire gives you up to 100% boost but usually around 85-90% boost

Trifire seems to give you another 70% boost over crossfire. When you go quad things aren't that big of a boost over trifire. You get maybe another 30% boost or so.

I can't seem to find the review where they show 2 gpu vs 3 gpu scaling but I think it was guru3d. I'll search for it tomorrow and update my post if I can find it. I think if you can get a 6990 cheap it's not a bad idea to upgrade. I was thinking about doing the same.
 
From what I have seen in the past crossfire gives you up to 100% boost but usually around 85-90% boost

Trifire seems to give you another 70% boost over crossfire. When you go quad things aren't that big of a boost over trifire. You get maybe another 30% boost or so.

I can't seem to find the review where they show 2 gpu vs 3 gpu scaling but I think it was guru3d. I'll search for it tomorrow and update my post if I can find it. I think if you can get a 6990 cheap it's not a bad idea to upgrade. I was thinking about doing the same.

I made another thread in this very forum where I compared all 3 methods.
 
From what I have seen in the past crossfire gives you up to 100% boost but usually around 85-90% boost

Trifire seems to give you another 70% boost over crossfire. When you go quad things aren't that big of a boost over trifire. You get maybe another 30% boost or so.

I can't seem to find the review where they show 2 gpu vs 3 gpu scaling but I think it was guru3d. I'll search for it tomorrow and update my post if I can find it. I think if you can get a 6990 cheap it's not a bad idea to upgrade. I was thinking about doing the same.


Thanks for the info. Now to just find a 6990 for a good price..... $700+ seems to be the going rate right now which is more than I'd like to throw down.
 
4 GPUs in the 6xxx serie need alot of CPU speed to start flexing it's muscle, and high memory speed also. Also, the 6990 and 6970 can all easily go over 1000 on the cores.

With my i7 2600K at 5.4, I see a good improvement of fps going from 3 to 4 GPUs, and with DDDR3-2400, I still get another 10% compare to DDR3-1600. And going from 880 to 1010 cores yeild another nice upgrade.

But under 4.8, and with DDR3-1600, 3 to 4 GPUs at 880/1250 won't be able to show what Quad-Fire can really do. :)
 
Thanks for the info. Now to just find a 6990 for a good price..... $700+ seems to be the going rate right now which is more than I'd like to throw down.

My system only allows for 2 cards so I could quadfire 6990s or tri fire a 6990 with a 6900 card.

IF your motherboard allows for 3 cards to be used I'd suggest gettting another 6950 like a toxic and unlock the shaders via the switch. Much cheaper, even a pair will run you just over $525.00 vs $700 for a 6990. But if your in the same boat as me, your left with little option. The price is only 1/2 the problem with the 6990, first find one in stock. :)
 
Seems a little low, can you bench the 6990 by itself? Just curious why it doesn't look like I would guess 4-way would. This is what I got from just 2 6970's stock.

CrossfireStock.jpg
 
You're not going to see optimal scaling of four or even three GPUs on a single 1080p screen, come on guys you know this already. It'd be like testing dual-card scaling at 720p. You need to run triple monitor resolution to see the best scaling with that many GPUs.
 
You're not going to see optimal scaling of four or even three GPUs on a single 1080p screen, come on guys you know this already. It'd be like testing dual-card scaling at 720p. You need to run triple monitor resolution to see the best scaling with that many GPUs.

Thats why I asked for the 6990 by itself, I'm curious how it scores against 2x 6970s. Obviously his setup is a little overpowered for one 1080p screen, but overkill is OK too.
 
Back
Top