• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

7950 Crossfire vs 680 SLI at 5760*1080 (Eyefinity/Surround)

1200 is easy. I'm only on 1.137V at 1150. This seemed high enough to rule out any random crashing while I tested but I'll bring it down when it's settled.

I benched 3DM11 @ 1250/1700 but performance was quite low suggesting some kind of issue my end as they were the same clocks roughly as a 11600 graphics score.

Looking for a benching clock of a touch over 1200 and whatever I can get on the memory. This should fulfil the brief of matching the performance of your average 7970 overclock and *should* match or be a little under the beast at 1080p. If I'm matching the beast at 1080 with a set of clocks then that's perfect because we'll be able to tell how much difference the 384 bit bus makes at triple screen res when otherwise performance is the same of at least very close.

My job is pretty flexible - you can go in when you want and leave when you want. The work is there no matter what time you get there. So I got to work at 6:30 this morning and will leave around 4:00 as there's a lot on.

Is is sad that I got in early so I could crack on with my actions and leave ASAP to continue? :D
 
Last edited:
One card I've got a feel for. Need to do the same for the other.

Can't seem to get the unlocked limits working for both cards though in MSI AB.

Also getting some BSODs which only appeared in Crossfire. Potential issue my end.
 
Card 2 is about 10c hotter than card 1 at the same voltage. Goes up to 85c on a Heaven run and that's with the volts only on around 1.14 :(. 75c I could live with but 85c is a little much.

Tempted to send it back :D.
 
SkodaMart said:
Ok.
I am genuinely looking forward to the results, infact, it is the first thread I have ever subscribed to.

Out of interest, is that with both cards in the case or just the one?

The reason I ask, is I have had differing temps on SLi setups before now.
Click to expand...

Both times with both cards in the case (to ensure consistency). I know one card will be warmer but this when when I swapped the cards around. So exactly the same scenario.

I even turned my side fan round to extract instead of blow and that raised temps another 10c to 95c.
 
SkodaMart said:
I am actually willing to stick my neck out and say, all this HD 7950 hype is purely that....hype.
It's almost like the fairytale 'The King and his New Clothes'

I will be so surprised if you manage to beat GTX 680 SLi with HD 7950 Crossfire.

'Edit' and that is by no means a pop at ATi, the price difference is not comparable.
Click to expand...

heavenc.png


That was with dual 1150/1700 clocks.

Not bad :D. Shame one of them got to 91c :eek: As the other one doesn't get above 75c in the same slot I would be inclined to say I've got a stinker.
 
That score puts me 5th highest 2 GPU score on Heaven. Not bad for a pair of £260 cards. :)

Anyway, triple screen time - I'll do BF3 as that's the most interesting then I'm going to bed... I'm absolutely knackered. Oh need to give away a Sleeping Dogs code before bed too!
 
Indeed. I decided earlier today that I'm going to complete these benches then sell two monitors and go back to single screen. This will include DSR'ing the hot card.

I just don't play enough to have £600 of monitor sitting on the desk doing nothing.

Will give everyone the info they want first though :).
 
I've got a spare monitor in the garage for that :D.

Card 2 - 65c, Card 1 - 92c.

LOOOL

Even considering Card 1 will be hotter due to card 2 dumping air on it... card 1 was at 85c on its own!
 
Drum roll please... BF3 results are in. I did it 3 times to be sure I wasn't going mad... it's a LOT faster. More than I expected.

And you have to bear in mind I can't go a MHz above 1150 on the core due to my retarded top card. Graph incoming for those interested.

Then I will then bench the rest tomorrow evening with my window open and send back the piece of ****. :D No more partial results after this.
 
bf3multicomp.png


Bearing in mind how rubbish my top card is, I would say a pair of half decent 7970's could extend my results by 10%!

I'm approximately 25% faster at this resolution with 7950 crossfire over 680 SLI. Very impressive results considering I was slower at 1080p with the same clock speeds.

There is a very big "BUT" though which I will attempt to explain tomorrow when I've completed all the games I said I would.
 
Last edited:
Tonester0011 said:
Very impressive.
Rusty, any chance you could bench CF 5760 on "Going Hunting" please? Just so i can add it to the benchmark thread graphs. That also gives my 7970s something to compare against.

Mine are 1225/1725 though, is that average or decent for CF?
Click to expand...

If I get time - wanna get it sent back to OcUK on Saturday morning and I got the other games to bench first.

That's assuming card 1 doesn't spontaneously combust before then.

Right I'm going to bed - I will conclude this all tomorrow so that everyone can read what I have to say and have a look at the numbers when complete.

On the face of it so far AMD just shouldn't have competition in the high res bracket. You could sell your 670's/680's and swap to 7950's, make money on the change and get better performance on paper.

It just makes nVidia's cards look like a bit paradoxical almost. They make it so easy to go surround with the connections on them and decent SLI drivers but throttle them back with a 256 bit bus which as my BF3 results have shown is quite the limiter at this resolution considering my good 7950 was actually slower on the same 1150/1700 clocks @ 1080p.
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom