• The forum will be offline Monday from 10am until approximately 3pm for maintainance and upgrades.
  • 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.

** The Official Nvidia GeForce 'Pascal' Thread - for general gossip and discussions **

Soldato
Joined
7 Feb 2011
Posts
7,699
Location
Stoke on Toast
HBM2 isn't likely to be available til Q2.

AMD apparently also has priority on its availability, so it's entirely possible the first enthusiast 14/16nm cards will be from them regardless. Nvidia wont want to wait around, but they might not have a choice.


Nvidia have been in this game long enough I imagine they have their own version HBM2 made by someone else.

They won't be planning a 2016 launch if they don't allready have stuff in production.
 
Soldato
Joined
7 Feb 2011
Posts
7,699
Location
Stoke on Toast
and production would still have to be underway to meet the qty's of cards they had for the 970 980 launch


I know it's important for AMD that nvidia are slow but they won't be.
 
Last edited:
Associate
Joined
4 Nov 2013
Posts
1,437
Location
Oxfordshire
There's a slight concern about the "early" 2016 Pascal launch for me.
If you look back at released cards, Maxwell/Kepler Fiji/Hawaii..you will find that the chips were all manufactured a good 6-8 months before release.
For Pascal to release Q1-Q2 they should be in the oven by now (or not far from now).
How did they tested them if only AMD has HBM right now is another question...
 
Soldato
Joined
7 Aug 2013
Posts
3,510
Nvidia have been in this game long enough I imagine they have their own version HBM2 made by someone else.

They won't be planning a 2016 launch if they don't allready have stuff in production.
I wasn't suggesting Pascal would be pushed back to 2017. But even a month or two advantage will be significant.
 
Caporegime
Joined
17 Jul 2010
Posts
26,091
They could probably bring it earlier but while high end Maxwell is thoroughly beating AMD into the ground they know they don't have to. No doubt they could also release something truly groundbreaking, possibly twice as fast as Maxwell but they won't because they can drip feed in a 20-25% increase which fanboys will lap up, release another version (or rebadge with slightly increased clocks) which will also sell.
 
Soldato
Joined
9 Nov 2009
Posts
24,993
Location
Planet Earth
Big Pascal has to be there to fight 14NM Xeon Phi,so I expect Nvidia will be trying to get it put as quickly as possible for commerical customers. For example Tianhe-2 uses Intel Xeon Phi co-processor cards and not any Nvidia or AMD ones.
 
Caporegime
Joined
20 May 2007
Posts
39,931
Location
Surrey
They could probably bring it earlier but while high end Maxwell is thoroughly beating AMD into the ground they know they don't have to. No doubt they could also release something truly groundbreaking, possibly twice as fast as Maxwell but they won't because they can drip feed in a 20-25% increase which fanboys will lap up, release another version (or rebadge with slightly increased clocks) which will also sell.

On the flip side, they know releasing something only marginally faster than the previous gen wont tempt those with current high end stuff to upgrade. If they release something big and powerful and with no competition they could completely dominate market share.
 
Caporegime
Joined
17 Feb 2006
Posts
29,263
Location
Cornwall
If they are only starting to produce the HBM2 memory in Q1 next year, then there is no way we are seeing Pascal in Q2, surely. Given that the memory needs to be shipped, products assembled, tested, refined, and shipped again.
 
Back
Top Bottom