I am extremely bullish on Intel (INTC), and it is by far my highest conviction stock holding. However, there has been significant speculation that Apple's (AAPL) next iPhone will feature either an Intel-designed SoC or that Intel will build Apple's ARM-based chips on its latest and greatest process technology.
While I can't speak to further generations beyond the iPhone 5 and the next iPad, I find it incredibly unlikely that Apple will switch to Intel for its iOS lineup of products for a number of reasons.
Intel Architecture Is Not Yet Compelling Enough
Apple's iPhone and iPad lines are not particularly in "need" of Intel architecture based chips. Intel's current "Medfield," while currently respectable against the dual core ARM (ARMH) Cortex A9 design found in Apple's current A5X chip, is certainly not compelling enough from a performance/watt standpoint to justify the massive work required to port iOS to x86 and make sure that each and every application in the iOS ecosystem works without a hitch.
Further, from a performance per watt perspective, if Apple were to attempt to be more aggressive in this area, ARM's own Cortex A15 offers a significant performance improvement over the Cortex A9 (and I do not doubt that future iterations of Apple's own "A" series chips will utilize this design). Qualcomm's (QCOM) Snapdragon S4 is also available and is currently the fastest mobile SoC available.
While I firmly believe future iterations of Intel's mobile architecture will be very compelling, especially as Intel brings its process technology advantage to the ultra-mobile side, in the near term, it just would not make sense for Apple to switch. Apple is also not particularly concerned with having the absolute bleeding edge on the performance front (and let's face it, Apple's products aren't about that), so making a major architecture shift would need to be engendered by extraordinarily compelling reasons.
Process Technology Lead: Not Yet On Mobile
While Intel holds the process technology lead in the mainstream notebook and desktop space with its 22nm tri-gate, the process technology lead has not yet extended into the ultra-mobile space. Intel's "Medfield" is built on a 32nm process, while its competitors from Qualcomm, Texas Instruments (TXN), and Nvidia are shipping or are slated to ship parts built on TSMC's (TSM) 28nm process, which represents a slightly smaller geometry than Intel's 32nm low power process.
At the upcoming Intel Developer Forum, Intel will be detailing its plans to accelerate its process technology development in the low power space. However, until this lead is firmly established against TSMC and Samsung (which is also currently shipping chips based on a 32nm low power process), then there will, once again, be no compelling reason for Apple to shift to Intel as a foundry partner for its mobile SoCs.
Does It Matter?
The big question, then, is, "does it really matter?"
I would say that an Apple shift to Intel architecture or even Intel foundries would be a significant PR win for Intel, and, especially in the former case, would be a gigantic validation of Intel's efforts in the smartphone space (and I'm sure Intel's P/E would expand quite rapidly), However, lack of such wins should not be viewed as particularly negative. Apple has its own SoC design team and Samsung seems to be doing a fine job providing advanced manufacturing technology for Apple's chips. Switching would be a hassle more than anything, and even if there were compelling technical reasons to switch (and I believe there will be within a few years), there would be significant business risk that Apple need not take.
Further, as Apple's market share will continue to shrink as Microsoft's (MSFT) Windows Phone and Google's (GOOG) Android continue to take share (this is not negative for Apple as the smartphone space is generally growing and the firm's margins are very juicy), Intel's TAM will increase in other areas (especially now the hard work of porting and supporting Android for x86 is done), and it is how well Intel addresses those opportunities that will make or break its smartphone chip play.
What Are These Opportunities?
Well, to wrap up the discussion, it is important to reiterate Intel's opportunities in the space. As I noted in my article, Examining Intel's Guerilla Approach To The Smartphone Sector, Intel is starting by gunning for low to mid range Android smartphones in emerging markets. This allows Intel to "debug" its approach before attempting to take on the more exacting developed markets such as the USA. A number of partners, including ZTE, MegaFon, Lava, and Orange, have already announced designs based on Intel's "Medfield" SoC. Even more encouragingly, Motorola Mobility has recently confirmed that its UK version of its Droid Razr M will pack an Intel chip.
Intel will likely be able to win higher end Android designs as its chips become more powerful and power efficient, but that won't be until 2013, when the first dual core Atom Z2580 with two x86 cores and LTE will be available.
Intel may also be able to find a home in Microsoft's Windows Phone, and Intel seems to be ready to perform the legwork necessary to bring x86 to Microsoft's platform if it is commercially viable.
In short, I doubt Intel will be inside the iPhone or iPad anytime soon. It would be a significant risk that Apple would not likely be willing to take in the near future. However, with Android currently holding 68.1% share and with Apple's share decreasing, an Apple win would be nice, symbolic, and "cool," but it is not critical to Intel's success in the space.
Disclosure: I am long INTC.