HD7970 vs 7950 @ 5670x1080

[xXxsephirothxXx]

Hi,
I'm struggling to justify the price bump on the 7970 over the 7950. I'm running triple monitors and just wanted to know what sort of fps difference there will be once 2 7950's are over-clocked.

anyone know?

 

[LtMatt]

10-15% difference at stock clocks, 5% difference at matched clock speeds. *Roughly*

 

[Gregster]

10-15% difference at stock clocks, 5% difference at matched clock speeds. *Roughly*

+1

From looking at bench threads galore lately, the 7950 clock for clock is ~5% on average slower than the 7970 at the same speeds

 

[Diagro]

Have a look for reviews on xfire with that res.

 

[Rusty0611]

Have a look for reviews on xfire with that res.

Going to make no difference to the difference between the two. It's not like the 7970 performs proportionally better than the 7950 at high res.

 

[Diagro]

15% is substantial.....

 

[Drangueos]

I tested 2x 7970 Ghz edition vs 2x MSI 7950 TFIII Boost Edition
I used AVP benchmark, Tombraider, 3dmark11 and crysis 3 (gaming)

7 to 12% with stock speeds (960/1250 vs 1100/1500)
3 to 8% with same clocks (I tested both at 1100/1500)
10 to 16% both overclocked (1150/1500 vs 1200/1700)

I already tested myself.

I'm using 2x7950 for 3 monitors, if my MB could support 3 graphics cards I would be using 3x 7950.

with 2 cards and 3 monitors you will not be able to play some games at the highest quality.

 

[Rusty0611]

15% is substantial.....

At stock clocks. Who leaves their cards at stock clocks on here? Especially on a triple screen res

 

[xXxsephirothxXx]

I tested 2x 7970 Ghz edition vs 2x MSI 7950 TFIII Boost Edition
I used AVP benchmark, Tombraider, 3dmark11 and crysis 3 (gaming)

7 to 12% with stock speeds (960/1250 vs 1100/1500)
3 to 8% with same clocks (I tested both at 1100/1500)
10 to 16% both overclocked (1150/1500 vs 1200/1700)

I already tested myself.

I'm using 2x7950 for 3 monitors, if my MB could support 3 graphics cards I would be using 3x 7950.

with 2 cards and 3 monitors you will not be able to play some games at the highest quality.

10-16% is a fair amount especially when running 3 monitors. and thats what im struggling with. is that % worth the extra £100. hmmm

 

[LtMatt]

10-15% difference at stock clocks, 5% difference at matched clock speeds. *Roughly*

I tested 2x 7970 Ghz edition vs 2x MSI 7950 TFIII Boost Edition
I used AVP benchmark, Tombraider, 3dmark11 and crysis 3 (gaming)

7 to 12% with stock speeds (960/1250 vs 1100/1500)
3 to 8% with same clocks (I tested both at 1100/1500)
10 to 16% both overclocked (1150/1500 vs 1200/1700)

I already tested myself.

I'm using 2x7950 for 3 monitors, if my MB could support 3 graphics cards I would be using 3x 7950.

with 2 cards and 3 monitors you will not be able to play some games at the highest quality.

Fair enough. Good info thanks for posting.

 

[Rusty0611]

10-16% is a fair amount especially when running 3 monitors. and thats what im struggling with. is that % worth the extra £100. hmmm

That ain't at matched clock speeds though - they're at the clock speeds he was able to get them at. You could get a good 7950 or you could get a bad one, same with the 7970 .

 

[Tonester0011]

10% boost at unplayable FPS is still unplayable FPS....would make no difference to your gameplay and both ultimately would mean lowering settings.

 

[whyscotty]

If 7950's struggle, so would 7970's

 

[Gregster]

10% boost at unplayable FPS is still unplayable FPS....would make no difference to your gameplay and both ultimately would mean lowering settings.

Agreed.

Not sure if people misconstrued it or don't give it much thought but I would consider 25fps as unplayable and 10% on that makes it 27.5fps and still unplayable even 20% is touching unplayable at 30fps.

 

[WingZero30]

10-15% difference at stock clocks, 5% difference at matched clock speeds. *Roughly*

+1

From looking at bench threads galore lately, the 7950 clock for clock is ~5% on average slower than the 7970 at the same speeds

I tested 2x 7970 Ghz edition vs 2x MSI 7950 TFIII Boost Edition
I used AVP benchmark, Tombraider, 3dmark11 and crysis 3 (gaming)

7 to 12% with stock speeds (960/1250 vs 1100/1500)
3 to 8% with same clocks (I tested both at 1100/1500)
10 to 16% both overclocked (1150/1500 vs 1200/1700)

I already tested myself.

I'm using 2x7950 for 3 monitors, if my MB could support 3 graphics cards I would be using 3x 7950.

with 2 cards and 3 monitors you will not be able to play some games at the highest quality.

All of this is indeed verified. However some 7950s can also reach 1200 Mhz core speed looking on the net.

Some 7950s can also overclock higher than 7970s given the silicon lottery. But 7970 has the edge overall for obvious reasons

 

[Diagro]

Sigh... get what you can afford.

 

[Tonester0011]

Sigh... get what you can afford.

It's not about what you can afford, i'm sure most of us here could happily afford Titans with 3930ks...its justifying the price, 10% for an extra £100 isn't in the most part justifiable, more so when 10% extra performance = the same thing - lowering of ingame settings.

If the OP could justify the extra cost for 7970s he wouldn't have made this thread asking for opinions he would have already bought said cards.

Edit: Read this @OP
http://forums.overclockers.co.uk/showthread.php?t=18503005

The res wont effect the end result but should serve as a good read, me and Rusty were limited by *my* poor clocking 7970s not his 7950s, given maximum clocks he would have performed equally to me.