Samsung (NASDAQOTH: SSNLF) has been a wonderful chip supplier to Apple (AAPL +0.11%) as far as delivery and execution goes. As far as just about anybody can tell, Apple has never had supply issues related to its custom-designed applications processors built exclusively by Samsung. And since most of Samsung's logic foundry business comes from Apple, it has been in Samsung's best interests to keep Apple happy.
Unfortunately, there is a massive conflict of interest here. Not only is Samsung a direct device competitor to Apple, but it is aggressively pursuing its own chip-design efforts, and, to some extent, looking to sell chips to other mobile-device vendors.
For a while, all eyes were on Apple potentially inking a deal with Intel (INTC +4.85%) for the manufacture of its processors. But a recent stream of rumors seem to indicate that Apple may use TSMC (TSM 0.83%) for the 20-nanometer node, and then it will split the orders between Samsung and TSMC for the 14/16-nanometer node. This raises the question: "Does Apple really need Intel to escape Samsung's clutches?"
Why would Apple want to use Intel?
The bottom line is that Intel has the best semiconductor manufacturing technology in the world. This isn't a secret and even the most ardent of Intel bears will grudgingly admit this. While Intel has shipped hundreds of millions of chips built on it 22-nanometer FinFET process, the semiconductor foundries have yet to ship their first one. And, by the looks of it, it'll be at least another year before one of them does in any meaningful quantity.
In the semiconductor business, transistor performance and power consumption is critical to the power and performance of the actual chip -- after all, a design is only as good as its building blocks.
Apple would most certainly benefit from bringing its designs to Intel's foundries not simply for performance and density reasons, but for reliability reasons. Intel has a superb track record of ramping transistor technologies at high volume with high yields well before any of its competitors.
Intel has plenty of capacity and is probably the most reliable supplier of leading-edge technologies. In 2012, while TSMC faced yield and capacity problems on its 28-nanometer planar process, Intel was flooding the PC market with excellent 22-nanometer FinFET based chips. So, the advantages for Apple would be well beyond simply having leading edge performance -- it would have guaranteed supply.
Intel unlikely to take the business, however
There is a pretty reasonable strategic case for Intel to take Apple's business. First and foremost, this would take a good chunk of semiconductor business away from Samsung -- an emerging chip competitor to Intel. At the same time, it would be accretive to Intel's top and bottom lines. This would also likely strengthen Intel's ties with Apple, which, at least for its Mac lineup, is quite a good customer of Intel's.
On the downside, Intel would probably see gross margins from this deal well below the corporate average -- think in the realm of the low 40% range. Further -- and likely more importantly -- there is a very real opportunity cost here. As long as Intel can sell a wafer of chips designed by its own in-house team to Apple's competitors for 50%-plus gross margins, why should it take a much lower-margin foundry deal?
Intel's whole business is built on being paid for both the value of Intel's designs as well as its manufacturing prowess. As long as it can fill the factories with its own designs, building designs for an indirect competitor seems like an admission of failure.
TSMC and Global Foundries remain viable options
TSMC is most likely to be Apple's long-term chip supplier. It is a neutral foundry that does not compete with its customers, which significantly mitigates the risk of Samsung, or even Intel, having a "peek" at what Apple is doing with its own silicon.
Global Foundries, too, is an option, but given its track record thus far, it would need to prove itself as a viable supplier on leading-edge nodes to catch Apple's interest. Apple would probably want two sources for its chips, and TSMC and Global Foundries would be much better long-term option than TSMC and Samsung, for obvious reasons.
Foolish bottom line
Would it be nice for Apple to have access to Intel's latest and greatest process technology? Absolutely. Is it likely to happen? No, unfortunately not. However, if Apple can ditch Samsung and establish a long-term relationship with TSMC, then it stops enabling its largest competitor and eliminates a number of competitive risks. Those hoping for an Apple/Intel manufacturing deal will almost assuredly be disappointed.
