The New 3DMark Benchmark - Testing Smartphones to Multi-GPU Gaming PCs

Toggle sidebar Toggle sidebar
piesquared

piesquared

Golden Member
Oct 16, 2006
1,651
473
136
Advantage AMD

Now we are starting to see some more interesting results and we can actually discuss the differences between hardware vendors. Starting with the high end, the Radeon HD 7970 CrossFire is running about 7% faster than our pair of GTX 680s in SLI while the single card results show a gap closer to 11.5%. While that means SLI is scaling better than CrossFire, it isn't enough to make up for the performance advantage of the AMD hardware.
Click to expand...

http://www.pcper.com/reviews/Genera...mark-Testing-Smartphones-Multi-GPU-Gaming-PCs

Will be interesting to see the results of AMD's DX11.1 discrete class APU's but I would imagine that the dumbing down of the features in Cloud Gate and Ice Storm will be a disadvantage, and an advantage to intel's position on graphs. Not sure how this is representative of all hardware capabilities, but if intel's results even comes close to AMD's, then there's obviously a problem. Looking forward to those tests for sure.
 
B

boxleitnerb

Platinum Member
Nov 1, 2011
2,605
6
81
Yes there is, don't confuse the 7970 with the 7970 GE. Computerbase used the 7970 GHz which is 10-15% faster than the normal 7970. At PCGH the 7970 and the 680 are basically tied. If you really feel the need to call a 1-2% difference an AMD advantage, be my guest - I find it silly. No one contests that the 7970 GE is a tad faster.
 
piesquared

piesquared

Golden Member
Oct 16, 2006
1,651
473
136
yes, I call 10-15% an advantage. pcper also used 3960x as apposed to a 3770k. What are you referring to that isn't in line with the 2 links you provided?
 
F

Fx1

Golden Member
Aug 22, 2012
1,215
5
81
Dont forget a 7970 can be clocked way past a 7970 GE with the move of 2 sliders and no voltage increases. The 680 GTX loses all ways
 
B

boxleitnerb

Platinum Member
Nov 1, 2011
2,605
6
81
piesquared said:
yes, I call 10-15% an advantage. pcper also used 3960x as apposed to a 3770k. What are you referring to that isn't in line with the 2 links you provided?
Click to expand...

I spoke about the 7970 non GE in case you forgot ;)
And this card is tied with the 680 at PCGH and CB (extrapolated since they used the GE). At pcper, it is not. So they are not in line, simple.
 
S

SPBHM

Diamond Member
Sep 12, 2012
5,066
418
126
with the free version you can't select a single test, or even skip the demos.
 
Last edited:
cmdrdredd

cmdrdredd

Lifer
Dec 12, 2001
27,052
357
126
and.......frames per second is all that matters?

I mean geeze. 2% here and a few fps there is a big deal?
 
Last edited:
3

3DVagabond

Lifer
Aug 10, 2009
11,951
204
106
cmdrdredd said:
Missed the point. Flew right by you...

I know that anyway.
Click to expand...

Nothing flew by me. You seem to be the one missing the point. In a real life app. ~2fps would make no difference. In a app that serves no purpose but to measure performance in absolute terms the score is all that matters.
 
cmdrdredd

cmdrdredd

Lifer
Dec 12, 2001
27,052
357
126
3DVagabond said:
Nothing flew by me. You seem to be the one missing the point. In a real life app. ~2fps would make no difference. In a app that serves no purpose but to measure performance in absolute terms the score is all that matters.
Click to expand...

And the point of arguing about it online is? See that's my point.

Anyway, it's not like these graphic tests translate to games either. Never have.
 
Last edited:
3

3DVagabond

Lifer
Aug 10, 2009
11,951
204
106
cmdrdredd said:
And the point of arguing about it online is? See that's my point.

Anyway, it's not like these graphic tests translate to games either. Never have.
Click to expand...

Half the reason to benchmark is to post it online and compare with others. You go into damage control mode because AMD beat nVidia in a benchmark. :eek:

Don't worry, I'm sure nVidia will optimize drivers for it soon enough. AMD started off faster in 3DM11 too.
 
cmdrdredd

cmdrdredd

Lifer
Dec 12, 2001
27,052
357
126
3DVagabond said:
Half the reason to benchmark is to post it online and compare with others. You go into damage control mode because AMD beat nVidia in a benchmark. :eek:

Don't worry, I'm sure nVidia will optimize drivers for it soon enough. AMD started off faster in 3DM11 too.
Click to expand...

Oh? I do?

Where did I do that?

Oh that's right, you can't prove it cause you made it up. Don't wet your panties because I'm questioning a benchmark pissing match.
 
B

blackened23

Diamond Member
Jul 26, 2011
8,548
2
0
3DVagabond said:
Half the reason to benchmark is to post it online and compare with others. You go into damage control mode because AMD beat nVidia in a benchmark. :eek:

Don't worry, I'm sure nVidia will optimize drivers for it soon enough. AMD started off faster in 3DM11 too.
Click to expand...

Oh, come on. This back and forth is kinda silly.
 
Termie

Termie

Diamond Member
Aug 17, 2005
7,949
48
91
www.techbuyersguru.com
SPBHM said:
with the free version you can't select a single test, or even skip the demos.
Click to expand...

blackened23 said:
Lame.
Click to expand...

Yeah, they're killing us with "kindness". Obviously they want us to pay for it, and this may be the best way yet to force us to. No way I want to sit through the whole thing every time I run it - two of three testing modules aren't even applicable to gaming hardware (one is for phones!).

The whole run took about 15 minutes. Way too long for serious benchmarking work.

Liquid_Static said:
First run on my desktop, 660 ti @ 1293 MHz i5-3570k @ 4.4 GHz
http://www.3dmark.com/fs/14564
Quite a bit more demanding than 3DMark11
Click to expand...

Nice!

Here's my first run, before I test with a serious overclock. This is with a GTX670@1189 and a 3770k@4.0.

3dmarkfirestormbench118.jpg
 
RussianSensation

RussianSensation

Elite Member
Sep 5, 2003
19,458
765
126
boxleitnerb said:
At PCGH the 7970 and the 680 are basically tied. If you really feel the need to call a 1-2% difference an AMD advantage, be my guest - I find it silly. No one contests that the 7970 GE is a tad faster.
Click to expand...

A $450 card is tied with a $380 card? Great, I guess!

1Ghz Gigabyte HD7970 on the Egg is $380 with 2 free games.
1.1ghz Gigabyte HD7970 GE on the Egg is $430 with 2 free games.
Gigabyte GTX680 on the Egg is $450 with no free games.

It's not about a 1-2% difference in performance. You pay $80 more for a GTX680 over the 1Ghz 7970 for similar performance. Actually GTX670 is priced much closer to the 1Ghz HD7970 cards in the US. The $430 HD7970GE 1.1ghz version is faster than the 680 and still costs less, despite 2 free games.

Right now most after-market 670s are priced too close to 1Ghz HD7970s. In HardOCP's case, a GTX670 needed to be overclocked to 1.25Ghz just to keep up with a stock HD7970GE. At 1.25Ghz, the 670 used as much power as a stock 7970GE:
http://www.hardocp.com/article/2013/01/08/asus_geforce_gtx_670_directcu_ii_4gb_gpu_review/9

Unlike 3DMark11 that showed ridiculous differences between GTX680 and 580 and had little to do with real world gaming performance when comparing HD7970/GE against GTX680, the new 3DMark actually mirrors what's happening in recent DX11 games -- GTX670 @ 1.25ghz barely matches a stock HD7970GE, a situation depicted in Computerbase 3dMark scores.

Time for NV to drop prices on the GTX680 to $379-399 and for GTX670 to $319-329 and offer 2 free games. Right now a $380 1Ghz HD7970 with C3+Bioshock makes a $450 GTX680 or a $350-360 GTX670 seem overpriced, esp. the 680.
 
Last edited:
You must log in or register to reply here.

TRENDING THREADS

COMPANY
RESOURCES
FOLLOW
Top Bottom