-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
AMD Polaris architecture – GCN 4.0
- Thread starter Boomstick777
- Start date
Yes, still the full possibility of the 2560 - 2816 core full version of the part, but it is more than likely veering closer to 2560 cores for 40 cu's.
This is how i see the full parts.
Part, shaders, TMU's, Rops
P11XT = 1280,80,40
P10XT = 2560,160,80
V10XT = 4096,256,128
The TMU values pretty much remain the same as previous parts, Just the SP's per ROP count has increased if the leaked 470X info is correct.
Last edited:
So... do we finally get to learn anything tomorrow, I wonder? AMD are having their Macau press debriefing ("The Vegas of the East").
So finally there should be some good leaks, even if they put all the press under NDA, someone will spill the beans
So finally there should be some good leaks, even if they put all the press under NDA, someone will spill the beans
Soldato
- Joined
- 18 May 2010
- Posts
- 12,850
Agreed, the childish stuff is so cringe worthy it makes checking this thread for solid info a form of torture.
Please let there finally be some specs soon so this speculation and crap posts can end
Please let there finally be some specs soon so this speculation and crap posts can end
Maybe because the people talking about TSMC/nVidia being 6 months behind AMD were talking rubbish and it is infact Samsung who are behind TSMC in terms of the maturity of their GPU suitable process. (Its not AMD who are the delay here).
It would be good having some official info to read on a Saturday morning whilst nursing a hangover.
No, no, don't be silly. Everyone on here knows exactly what they are talking about. Remember GDDRX5 wasn't going to be on any card until at least September....
Last edited:
Working, but to what extent? Were there yield issues, how many could hit the necessary clock speeds? Was it under performing? Did it fail to pass some tests requiring a respin?
Who knows
Or they are just a few weeks behind TSMC in terms of normal development progress.
There was a couple of articles by Micron to that effect though.
This is my feeling, and if you start trying to research GlobalFoundaries you find that AMD is taking about ramping up to volume production with GF in 2016. yet there has been no news that GF has actually managed to move to volume production yet.
I'm not saying there is delay, I thought all along Polaris would launch in july.
Pascal supply sounds good with eh AIB apparently showing off lots of boards at computex.
Samsung aren't making AMD's GPUs. So I had a look for a GF LPP volume announcement and it seems they have not actually made one...
Regarding GPU-suitable, it seems that 14nm is just not. Maybe that has something to do with it.
NH stated AMD was late when a google search revealed it was posted way earlier on the net when AMD was going to present cards, so I said, anything else you needed?
as their own post from NH wasnt accurate and if a website for tech can get it wrong, guess what?
A forum user and users here gets it wrong all the time.
and if they own a nvidia card you cant trust them either.
