AMD server roadmap 2014

[pablo87]

Globalfoundries, but the Reg says TSMC.



Not really Pablo. Those are not compatible processes, so if true instead of finishing a design and start working on an updated version or in a brand new design on the process, AMD engineers will have to backport everything from TSMC to Globalfoundries process just because of the WSA nuisance. That's really bad from a TTM POV.

If my speculation is correct, TTM improves both because a) AMD has adopted the ATI way of doing things and part of that is launching new products on existing and new nodes at TSMC and b) TSMC is ahead of GF.

By the way I was surprised how small the Kaveri die is allegedly...

Cost is another matter...but that's the WSA for you.

 

[pablo87]

Why does AMD market always write things they will regret later on?

Would you buy an AMD x86 server after reading this?

3 years ago I questioned AMD's server strategy over on S/A and got flamed. Just made no sense to me to carry on with such expensive development, dual sockets, etc.. and a me too strategy when Intel had a better product, way better infrastructure and superior customer engagement, not to mention more advanced fabs!!! It was probably all based on the hope (aka delusion) of getting back to 25% MSS...

Anything that accelerates the demise of that failed strategy is a positive IMO, including self inflicted...but yeah, that is some bizarre slide.

 

[CHADBOGA]

Why does AMD market always write things they will regret later on?

Would you buy an AMD x86 server after reading this?

One shouldn't need that slide to be wary about AMD x86 servers.