Cat is out of the bag!
1) 10nm tools is money spent and if they ever get production going between low yields and huge capital depreciation that will bust margins
2). Two more generations of larger die on 14nm was never in the spreadsheet. 14nm yield issues and larger die and slower server growth and faster client drops drove ACT1 and the silent ACT of 2015
3) IoT, memories, and modem are commodity. Bench mark the biggest competition there and their margins and cost structure and you will realize how uncompetitive intel is here just on cost.
The smoke and mirror about profit in IoT comes from lumping a lot of client and server in the IoT bucket if it can pass a smelly underwear test.
2017 ACT2 and 2018 ACT3 will be required for intel to even stay afloat. And as 10nm ramps and the full cost of the commodity products have to be accounted no amount of smoke and mirrors will hide BKs failures.
15-20K cuts next year and similar in 2018!
You intelers got a lifeboat?