I also think Apple will turn to Intel for the fab work. Let's compare the two options:
- $10billion+ for TSMC to do fab work on 28nm vs. $0 for Intel to do fab work on 22nm
- TSMC 28nm planar vs. Intel 22nm trigate
- TSMC 28nm or 20nm yield problems vs. Intel's proven high yields and 22nm
- TSMC has no clear roadmap vs Intel clear roadmap for 14nm next year and 10nm and 5nm starting in 2015
- TSMC cannot help with apple's chip design vs. Intel proven chip design superiority who can help design and integrate all of Apple's products into same platform
Let's see - they are just starting to explore the idea so with planning, construction, testing and ramping they might get product out the door in three or four years. By that time Intel will be doing 10nm out of five facilities. Since Steve Jobs is gone, Apple will still be selling iPads and iPhones in three or four years. Does anyone think Apple can make enough incremental improvements in the existing products for them to be relevant in three or four years?
So, if they go with TSMC, they should have old fabrication to go with old products in three or four years. Sounds like recipe for disaster to me. With its absence of new products and the ARM foundry fabrication crisis, Apple's future does not look bright.
Well, a deal with Intel would take care of one of the problems...