It is well-known that Apple (AAPL +1.18%) has become a very formidable semiconductor chip designer. Its A-series processors are extremely well designed, and given that the company seems to be able to attract top talent from chip design houses all over the world, I think that we're only seeing the beginning of what Apple's chip team can do.
I truly believe that Apple cares very deeply about the performance and competitiveness of its processors -- otherwise it wouldn't be spending big bucks on chip development teams. While Apple seems to have the design side of its chip manufacturing taken care of, I do worry that in the long-term battle with Samsung (NASDAQOTH: SSNLF), the latter's in-house chip manufacturing is going to prove a significant disadvantage to Apple.
For example, the upcoming Samsung Galaxy S6 is reportedly going to feature a Samsung-designed 14-nanometer FinFET chip, which should allow it significant performance/power benefits relative to the Taiwan Semiconductor (TSM +6.30%) built 20-nanometer A8 chip found in the iPhone 6/6 Plus.
That's why I would be very interested in Apple inking a chip manufacturing agreement with chip manufacturing powerhouse Intel (INTC +4.36%).
Intel would get a huge mobile presence; Apple would get amazing silicon
As has been widely talked about, if Apple were to move to Intel for chip manufacturing, then this has the potential to be a huge win/win. Intel would get plenty of leading-edge semiconductor volume inside of its plants and would generate likely a few billion in incremental revenue at decent (but almost certainly below corporate average) margins.
Apple would "win" in the sense that it would have access to more advanced chip manufacturing technology than what Samsung or foundry giant Taiwan Semiconductor (TSM +6.30%) would be able to bring into volume production in a similar timeframe. This could prove a very interesting competitive advantage against rival smartphone makers using Samsung-built or Taiwan Semiconductor-built chips.
The downside to such a deal
Intel's Sunit Rikhi -- the individual who runs Intel's Custom Foundry efforts -- was recently interviewed by Semiconductor Engineering. According to Rikhi, Intel is able to "talk to everybody" due to its attractive technology, but implied that this isn't enough to secure foundry contracts.
"It is far more than the technology we have to sell for the customers to actually make a switch," Rikhi said in the interview. "And that's something a lot of people don't appreciate and understand."
Does Intel have the required capabilities in place to support a foundry customer like Apple, particularly one with such demanding product launch timelines? Intel has certainly proven that it can perform very steep ramps of its own products, but it's still very much "to be determined" with respect to the ability to ramp high-volume foundry designs.
It takes a long time to become a viable foundry
Samsung, which is now increasingly viewed as a potential long-term threat to Taiwan Semiconductor, started its foundry business many years ago. Indeed, according to an article published on SemiWiki in 2011, Samsung had just $420 million in foundry revenue in 2010. This, according to the piece, put Samsung in ninth place as far as semiconductor foundry revenues go. This is after reportedly five-years' worth of investment.
Getting into the semiconductor logic foundry business is tough.
Intel started its Custom Foundry business in 2009, and six years later, I would guess that this business generates within rounding error of zero revenue. Samsung did get the benefit of Apple's A-chip business, which led to significant growth of the company's foundry business from the 2010 to 2013 timeframe. Intel's recently announced foundry deals -- which are likely to be quite low volume -- aren't expected to generate material revenue for a while yet.
I think it will still be a while before Intel's foundry business is generating material revenue. However, if Intel eventually builds a good foundry business, and if its technology lead remains intact by that time, I think it could be interesting for both Apple and Intel shareholders to have the latter build the former's A-series processors.
