Excuse me, but did you intentionally forgot to mention new high end FX processors that are planned for 2016 I think? The ones Jim Keller is working on with his team?
You mean this new 'first generation' of high end procs that AMD is going to launch?
We know what this means. We have heard and seen it before. They will not be continuing their 1=2 FX core setup so its a new beast. The only reason it has ANY legitimacy is because they can say 'but, we have Jim Keller'.
My prediction: it is going to fail. Where on earth are they going to get all this R&D from to get a killer cpu to market? Magical budgets? Also: who thinks that a high end CPU is going to really 'save' AMD? The market for that has dwindled on the consumer side, and the market for server is drastically changing (ARM). Even IF they make a monstrous CPU; how is any volume of sales of that going to 'carry' the company? I don't see it.
Honestly the best thing that could happen to AMD right now is that they get acquired by someone with big pockets like Samsung. A company capable of realigning the whole business and kicking out all of the rotten apples in one go. AMD has done very little these past years to really change course, they are again late to the party and they are going to repeat past mistakes with this new CPU.
R9 390 is going to be the same story - HBM is pricy so the card will have a much lower margin than its competitors or the old R9 inventory - that old R9 inventory that is getting rebranded for the 4th(!!!) time. Again, I struggle to see how AMD is going to really profit from that release.
Really what AMD should have done, but this is in hindsight of course, is bridge the gap between x86 and ARM in terms of performance/power. Think 'Big.Little' in an x86-ARM cluster. If they would have a sensible way of doing THAT on mobile devices on ARM tech, they would have lead a new market that IS viable. Right now Intel is struggling because of the lack of this bridge, and everything points towards the need for synergy between all devices.