Intel came out last week finally admitting the only thing they will try to do in 14nm will be a few already contractually obligated 14nm chipsets for Altera, to be built at the current real known 14nm production costs.

Altera when they saw the price tag responded by dropping all their orders for these 14nm Intel chipsets and going back to TSMC for the 20nm "near equivalent" for speed and price that they are already using.

Apple is getting a 20nm A8 chipset instead of the 14nm A9 chipset that was originally touted for this Christmas.

So what the heck is going on ???    The 14nm lines really do exist and they are running (Apple is making A9 chipsets for Christmas 2015 as we speak).

It all has to do with the slow slow slow 4x multi-pass 14nm processes with UEV power levels still not being strong enough or accurate enough to do efficient single or even dual pass production wafers.  

This all translates into a real fiscal question about the real production COSTS and the "effective" manufacturing production rates of 14nm chipsets.



Yup, you saw this chart at the start of 2014 -- nothing has really changed except they have added the 14nm node out there at its real cost/speed relationship.   And 14nm sucks too badly to really do anything but extremely expensive Apple A9 chipsets right now.

What does this really say?    Intel will stick with their 22nm production processes for another year until hopefully there is a breakthrough in 14nm that makes it more cost effective.   ARM will go down to 20nm for the high end stuff and then wait for the same 14nm breakthrough.

Apple is paying out the butt right now to do "production exploration" and to "refine the 14nm processes" at two separate vendors -- but questions are coming up now as to whether Apple also will pull the plug on 14nm production until it becomes much more cost effective.

Cost is cost -- and 14nm simply costs too much to do right now.