It would be a new chip design. And not only would Zen have to deliver. Polaris would have to as well. And dont quote PR numbers, they are meaningless.
And then all this would have to sell in a high enough volume at a right price to actually make money on it.
So far the next APU have always been a bigger flop than the previous in sales. Despite that some people always think the next APU is going to be the savior and big seller.
I think, with ongoing time and increasing capabilities in automation, simulation, test, verification, etc., chip design goes away from hit or miss to closing in on design targets. And having Polaris 10 and 11 in their hands, RTG didn't step back and say, that they'll continue to sell Fury instead. So energy efficiency mist have improved at least by 10%.
The flopping APU series (ignoring latest JPR numbers here), used these process geometry metrics since 2010: -, 32nm (and 40nm), 32nm, 32nm, 28nm, 28nm, 28nm
Intel at the same time: 32nm, 32nm, 22nm, 22nm, 22nm, 14nm, 14nm
Until 2016, AMD has to use ST-weak construction cores, which didn't leave too much power for the iGPU, except maybe with BR, which has the power management improvements. And it still has to use the previous gen GCN.
So why do you make projections, which seem to falsily be based on the continuation of these 3 series? Would you also have predicted future housing prices based on their development from 1990 to 2007?
