According to DigiTimes, citing "industry sources," Apple (NASDAQ:AAPL) has outsourced the manufacturing of the Apple Watch applications processor to arch-rival Samsung (OTC:SSNLF). The report also claims that Apple is ordering between 3,000 and 4,000 12-inch 28-nanometer wafers' worth of applications processors per month for the watch.
It seems that despite all of the brouhaha around Apple trying to cut Samsung out of its supply chain, at the end of the day, Apple is willing to work with this "frenemy" to get the components it needs.
The question, then, is why would Apple choose Samsung rather than foundry giant Taiwan Semiconductor (NYSE:TSM) for its "next big thing"?
Plenty of possibilities, but here's my favorite
The first possibility, and the one I believe to be most likely, is that TSMC's 28-nanometer capacity is probably booked solid. This is supported by a recent article from DigiTimes claiming that TSMC has been telling its clients to "place wafer orders in advance so that they can secure a steady supply during the peak season and meet demand in time."
On the other hand, Samsung's foundry business is reportedly in pretty poor financial shape. According to IBK Securities' Lee Seung-woo (via WSJ), Samsung's logic chip business is expected to post an operating loss of 877 billion Korean won (about $797 million) this year. I wouldn't be surprised if a substantial part of this loss is due to non-cash excess capacity charges, similar to what Intel faced in late 2012 and through a substantial part of 2013.
It's not much of a stretch to assume that Samsung is pretty hungry to fill up its factories with something. In that case, I could see Samsung offering Apple very attractive wafer prices for the Apple Watch chip, something Samsung has reportedly done to try to win Apple's A9 business.
Game on for the next Apple Watch, it seems
While it seems that Samsung has won the chip orders for the first Apple Watch, the DigiTimes report claims that both Samsung and Taiwan Semiconductor are "expected to be in the race for orders for the second generation Apple Watch" -- at least according to the cited "industry sources."
Now, what's interesting here is that DigiTimes claims that the second-generation Apple Watch processor is likely to be built on a 14/16-nanometer FinFET processes from Samsung and/or Taiwan Semiconductor, respectively. This has some interesting implications.
What does leading-edge silicon mean for Apple Watch?
If we assume the next-generation Apple Watch will come out roughly a year after the first Apple Watch -- call it March 2016 -- that means Apple views the Apple Watch as a device worthy of fairly leading-edge technology. Of course, new manufacturing technology means lower power consumption, which leads to improved battery life. However, the design costs and likely wafer costs for 14/16-nanometer chips are likely quite high.
Apple, if this report is true, thinks the Apple Watch is so important of a product category that it will invest pretty heavily in designing leading-edge, specialized processors for it. This implies not only that Apple views the category as important, but that the company envisions dramatically enhanced capabilities for future Apple Watch models.
While I think the hype around the potential sales figures of the Apple Watch may be overblown, I must admit I can't wait to see where Apple plans to take this product category over the long term.