R9 270x crossfire vs. R9 290x

Ausar

Honorable
Sep 30, 2014
37
0
10,540
I'm on a budget. Ill just put that up front. So money is crucial. I already have a r9 270x, so i was going to crossfire it, due to reviews saying it outperformed the 290x. But then i got counter reviews, saying that you always want to go with the single card. Now i know my rig can handle both setups (FX-6300 750w psu) and I just wanted to know which setup i should go for. Total cost of crossfire: 357.87 (including cable) Total cost of 290x: 369.99

Any and all help is appreciated!

For the love of Bacon,
Ausar
 
Solution


a9193e2e_5984_22_sapphire_radeon_r9_270x_2gb_toxic_video_cards_in_crossfire.png


And with this benchmark review,
you can assume that if the 270s in question were 270x or at...
Yes, 270x CF will be faster than a single 290x... but 290x has 4GB VRAM.
270x CF will only have 2GB VRAM.

Including all the extra problems one may encounter when running crossfire,
Single 290x would be the smarter decision, because you could then cross fire the 290x in the future if you feel the need. (after a psu upgrade).
 

Ausar

Honorable
Sep 30, 2014
37
0
10,540


So a CF doesnt combine RAM? Intresting, but VRAM is irrelevent if you have enough ram correct?
 

Ausar

Honorable
Sep 30, 2014
37
0
10,540


So a CF doesnt combine RAM? Intresting, but VRAM is irrelevent if you have enough ram correct?
 

Ausar

Honorable
Sep 30, 2014
37
0
10,540


So a CF doesnt combine RAM? Intresting, but VRAM is irrelevent if you have enough ram correct?
 

mdocod

Distinguished


How do you figure?

The 290X has 2816 shaders, 176 texture mapping units, 64 render output units.

2X 270X would have a combined 2560 shaders, 160 texture mapping units, and 64 render output units.

The core config of the 290X is actually WIDER than the PAIR of 270X's. The 270X's will average slightly higher clock speeds. Performance on paper is a wash. In the real world, the 290X will be faster most of the time because it will not suffer from scaling penalties.
 


a9193e2e_5984_22_sapphire_radeon_r9_270x_2gb_toxic_video_cards_in_crossfire.png


And with this benchmark review,
you can assume that if the 270s in question were 270x or at least over clocked, they would have out performed the 290x as they traded blows (some won, some lost) in all the benchmarks as only 270 (non-x) CF.
'http://www.ocaholic.ch/modules/xcgal/albums/userpics/New3DMarkBenched/SS/normal_FireStrikeP.png
 
Solution

mdocod

Distinguished
A_goat

That link doesn't work.

Different driver versions performed for each test in your bar chart = irrelevant data.

Admit your mistake and move on. What you are claiming is not possible. GPUs do not scale up to more than the sum of their parts.