AMD: Southern Islands (7*** series) Speculation/ Rumour Thread

9th of January seems to be wrong, unless reviewers usually have over 3½ weeks of time to do the reviewing. :)

EDIT: obviously wrong month in the original post.
 
Last edited by a moderator:
That's half a year review time :p

Maybe they've finally understood that reviewers can't do proper reviews in less than a week :LOL:

But anyway, hoping I understood few things right, and get to tell couple ppl "[nelson]ha-ha[/nelson] Told you so!" :D
 
slajd2rvkky.png


OBR said:
I have almost FULL performance info ...
Okay, i have NOT a cards (no one have them now) but i have some performance numbers. I have real results from Live presentation (with benevolent staff there) in 3D Mark 11, Unigine Heaven, Battlefield 3, Dirt 3 and Crysis 2.

Plus have numbers from slides (unreal PR stunts here) from slide above. I can say how HD 7970 perform, i know it from these few real tests ... stay tuned, all will be here soon.

If i am telling, performance is only few percents above HD 6970 (in Games not 3D11 and Unigine) it is NOT speculation, but info from real GAME results ...
 
Last edited by a moderator:
So, keeping the same setup rate as in Cayman, but now with cached stream-out gives something like ~50% performance increase?!
 
I have strong reason to believe 7970 will actually be released this year :eek:

If they wanted to release it this year they should have already released it, with full availability of course. Launching it now, without very high volume availability, will only decrease sales of their current gen cards and the Christmas sales are too important to risk that. Just my guess though, I'm not well versed with marketing and sales and everything that entails.
 
I just wonder, is the geometry processing (pre-setup) is distributed like in Fermi or still goes through a dedicated narrow pipeline -- two in this case.

Looks like GCN will still be behind Fermi in peak geometry rate.
 
I just wonder, is the geometry processing (pre-setup) is distributed like in Fermi or still goes through a dedicated narrow pipeline -- two in this case.

Looks like GCN will still be behind Fermi in peak geometry rate.

Maybe this can help... is it distributed, right?

010.jpg


slajd2.png



:D:D:D:D
 
Last edited by a moderator:
9th of January seems to be wrong, unless reviewers usually have over 3½ weeks of time to do the reviewing. :)

EDIT: obviously wrong month in the original post.

actually, it's still kinda right. but that will become clear soon.
 
Back
Top