In recent years Intel has signed on to an agreement with the FTC. The agreement is supposed to level the pricing of Intel chips. Now volume will likely be the only differentiator for price. Lower price will drive the OEMs in a race to the bottom. Cutting prices to consumers, competing for volume. Price will be the attraction at retail, other than shape and color of the plastic. If Intel ever gets that far in tab/phone.
If true, will be very bad for QualComm and subsequently Apple.
I'm not sure Apple is using much from Qualcomm that would be 28nm currently. The main things running on 28nm would be processors and GPUs. Apple's processors are still from Samsung, and still 40nm.
Apple is rumored to be looking at TSMC for 28nm processors for the next generation, but I think that is far enough out that the current capacity issues are not too relevant for Apple. The article claimed good capacity in another quarter or so.
The problem for Intel is that the fraction of BOM $ for processors in tablets and smartphones is way lower than what Intel views as their birthright in the PC world. Screens get the most $, and even batteries rival or surpass processors.
Yep, it would be really interesting to see how Intel could do fabing other designs. Maybe their cost is not so good. Still, we could probably assume the performance/power would be better, but how much.
Intel commented somewhat on that aspect re Medfield, by saying that if you look at Qualcomm or Apple, they pay $xx to get the part made by a foundry who has lower margins, but Qualcomm/Apple foots the bill for the IP development and also makes some margin as a result. So Intel gets both margins (IP owner + fab producer) whereas everyone else divides that.
The problem is that in the ARM SOC world, IP is pretty cheap. At the end of the day, Samsung who does IP + fab will sell anyone an ARM SOC at lower margins than Intel finds attractive, and Samsung also pockets both sides of the margins.