Comparison of 50 DX11 graphics cards in games and OpenCL!

CAT-THE-FIFTH · 13 Jan 2013 at 14:02

More details here:

http://translate.google.com/transla...iews/2013/50_directx_11_grafikkarten_im_test/

It looks like you need an HD7950 Boost Edition to double framerates on average over an HD5850 1GB.

opethdisciple · 13 Jan 2013 at 14:29

Almost 10 frame difference in BF3 in favour of the 680.

I thought AMD caught up?

tribz · 13 Jan 2013 at 14:35

opethdisciple said:
Almost 10 frame difference in BF3 in favour of the 680.

I thought AMD caught up?

Thats an interesting interpretation of those graphs!

TaKeN · 13 Jan 2013 at 14:43

opethdisciple said:
Almost 10 frame difference in BF3 in favour of the 680.

I thought AMD caught up?

What?

Martini1991 · 13 Jan 2013 at 14:44

opethdisciple said:
Almost 10 frame difference in BF3 in favour of the 680.

I thought AMD caught up?

7970 GHZ is 0.1 and 1.2 frame difference in the ones it loses.

opethdisciple · 13 Jan 2013 at 14:48

tribz said:
Thats an interesting interpretation of those graphs!

Look that the BF3 benchies here.

10 frame difference at 1680x1050 and 1920x1080.

CAT-THE-FIFTH · 13 Jan 2013 at 14:49

BTW,there is another review they did with budget cards a few months ago:

http://forums.overclockers.co.uk/showthread.php?t=18478585

It just shows you how careful you need to be with the memory types!!

Final8y · 13 Jan 2013 at 14:57

opethdisciple said:
Look that the BF3 benchies here.

10 frame difference at 1680x1050 and 1920x1080.

01.05.2012

opethdisciple · 13 Jan 2013 at 15:15

Final8y said:
01.05.2012

Fine doesnt change a thing.

The graphs say the difference is 10 fps. That was my only point.

We are discussing the op, so thats what I did.

Hence my original post which was:

"I thought AMD caught up?"

opethdisciple · 13 Jan 2013 at 15:15

tribz said:
Thats an interesting interpretation of those graphs!

This is not in the article.

Final8y · 13 Jan 2013 at 15:18

opethdisciple said:
Fine doesnt change a thing.

The graphs say the difference is 10 fps. That was my only point.

We are discussing the op, so thats what I did.

Hence my original post which was:

"I thought AMD caught up?"

The OP does not, only your outdated link does.
http://translate.googleusercontent....16.php&usg=ALkJrhhH-4RrKca5zQB6_loLCpSXBgJXog

tribz · 13 Jan 2013 at 15:22

opethdisciple said:
This is not in the article.

Yes it is, don't know what your looking at but that it from the OP article and the one your quoting is an entirely different one from may last year.

humbug · 13 Jan 2013 at 15:35

Different results on this test which shows the 7970 on par with the GTX 680 @ 1080P with the 7970 GE + 9 FPS.

http://www.techspot.com/review/603-best-graphics-cards/page4.html

I find that benchmark results these days depend on whose running the benchmark, not on the GPU's.

The truth is they both perform very well, beyond that its just about price.

Martini1991 · 13 Jan 2013 at 15:40

opethdisciple said:
This is not in the article.

Yes it is.
You've clicked a hyperlink to another review.

Hence the date, and the lack of GHZ edition and better drivers.

opethdisciple · 13 Jan 2013 at 15:59

Ok, my bad.

humbug · 13 Jan 2013 at 16:14

Interesting to see a benchmark for Brink, you don't see any OpenGL game benchmarks, many people do play OpenGL games, Brink is also one that i play.

http://translate.google.com/translate?sl=auto&tl=en&js=n&prev=_t&hl=en&ie=UTF-8&eotf=1&u=http%3A%2F%2Fht4u.net%2Freviews%2F2013%2F50_directx_11_grafikkarten_im_test%2F

Same with GPU compute acceleration, you don't see GPU's compared to eachother in applications.

I would like to see more people do that, owning a high powered GPU is not just about gaming, many 3D render, Photoshop, encode ect.... with their GPU's.

bru · 13 Jan 2013 at 16:36

One thing I have just been looking at is these two sets of battlefield 3 benchmarks, the ones in the opening post dated 11.01.2013 and then the ones that opethdisciple linked too from 1.05.2012

now even though they are several months apart they have been done on the same system hardware (except the memory but even then both sets have been set with the same timings) with the same map and by the looks of what the site says the same save game and 10 sec fraps run on version 1.03 of BF3.

earlier run on Nvidia driver 301.10 WHQL and AMD 12.3 WHQL

later run on Nvidia driver 310.70 WHQL and AMD 12.11 beta 11

Now just looking at the results that appear on both charts, the 680, 7970 non GHZ ed, 7950 non boost and the 580. You see that the AMD 7970 and 7950 have gained about 10% but the 680 has gained just over 5% as well and the 580 is rocking in there with about 9% gain.

Of course looking at the boosted version of the 7950 and 7970(GHZ ed) you can see that the extra clock speed has made just as big a difference as the much talked about driver improvements.

humbug · 13 Jan 2013 at 17:04

10 sec FRAPS run?

Thats a pretty tight and specific run, you can't get an accurate impression of performance in 10 seconds.

You need at least a few minutes, 10 seconds is like walking from point A to point B with point B being about 10 meters away.

Gregster · 13 Jan 2013 at 20:26

Good results and I am impressed with how good the 660TI comes out in those tests. Not far behind the 7950 and in a couple of games, over takes it.

LtMatt · 13 Jan 2013 at 20:31

Thanks for the review cat, but ill let more experienced heads skim through it and sum it up. Once i have to start clicking translate on every page and not every click of 'next page' brings me to a graph i pass.