I was ISP'd last week and have read lots of articles on this site and elsewhere. It would be helpful to take a 10,000 foot view of the situation.
In the beginning .. there was the humble PC, then the Workstation, then the Datacenter Server ...
The market starting shifting towards smaller compute devices (PDAs, Initial Version of Smart Phone -- Palm Treo etc). Intel's initial foray into Phone SOC's was the Xscale processsor, which they sold off to Marvel in 2006. The Intel Brass initiated the design of a low power IA chip -- Silverthorne. This was initially a two chip solution a "CPU" chip plus an I/O chip. This design was started in 2006 (2005 ??) under the auspices of Paul Otelini, then Intel CEO. The Silverthorne and Lincroft designs were slated for Ultra Mobile PCs and Mobile Internet Devices (MID). Then came Penwell (all components in one SOC), Cloverview, Valleyview, Tangier, Anniedale, Cherry Trail, and Broxton (which may live on as another product). In 2013ish, Intel also started work on the Sofia platform.
The upshot... Cloverdale appeared in phones made for overseas markets, Valleyview in Tablets, Tangier in the Fossil smartwatch, Anniedale in the Asus Zenphone. So, Intel was getting some traction for their efforts. In fact, Intel sold 40 million tablet SOC's. However, due to contra revenue and R&D costs, the phone and tablet SOC divisions operated at a loss (to the tune of 3 - 4 billion for each of 2013 and 2014). The goal was to get Intel SOC's into the market , refine the process so as to make money. But, at the end of day, Intel was 7 - 10 billion in the hole on phone and tablet SOC's. At some point, you have to stanch the bleeding. There is no amount of getting around this fact.
Now, then lets stop beating up BK and Aicha Evans. The SOC bleeding cannot be laid at the foot of BK. The SOC development was started under Paul Otelini. BK made a good faith effort to continue SOC development. IA SOC development started in 2006, BK became CEO in 2013. That's 7 years. Plenty of time for things to run afoul before BK became CEO. When you've spent 7 - 10 billion on something ... its hard to beleive that could be made up in a reasonable amount of time. The release of the Qualcomm snapdragon 820 was the nail in the coffin. Read the description of the chip on Q's website.
Built from the ground up as a Phone/Table SOC.
Yes, i have spent several days pissed at being ISP'd, yet I also noted a year ago that this could happen. A year ago, you could see other areas of the company making handsome profits.
It makes sense to play to your strengths. Right now Datacenter markets, IOTG, FPGA in Xeon CPUs, and memory are the hot spots.
And lastly, stop beating up the workers from India. Its pointless. The Indian coworker in my dept knew the PM stuff cold. He could recite the registers and the required settings of off the top of his head.
And that's my take on this .....