Is Intel a Threat to Taiwan Semiconductor?

Intel started its Intel Custom Foundry business back in 2010. To date, this has not been a meaningful revenue source for the company, but the progress is becoming more visible.

Intel's previous CEO, Paul Otellini, was adamant in that the company's foundry business was a very selective business and that it would avoid trying to go toe-to-toe with foundry giant Taiwan Semiconductor . Under new CEO Brian Krzanich does this still hold true? Does TSMC have anything to worry about? 

Intel Custom Foundry is probably not a huge direct threat
If you look at some of the deals that Intel has publicly announced, you'll note that the companies using Intel's foundry services are generally building really big, very cost-insensitive parts. Intel has made it clear that even though it is opening its foundries up to more potential customers, it still plans on being paid for the value that it delivers with its leading-edge technologies.

Further, many of TSMC's high-volume customers -- think Qualcomm -- are unlikely to build at Intel for a variety of reasons, an obvious one being that Intel is spending billions developing products that compete with Qualcomm. Less obvious is that giving Qualcomm access to leading-edge manufacturing could indirectly affect PC sales via mobile chip cannibalization.

However, Intel itself is a threat
While Intel Custom Foundry is likely to have minimal impact on TSMC's revenues, Intel's own mobile efforts could represent a threat. Every point of smartphone and tablet share that Intel gains with its own home-grown products is a chip that the foundry landscape in general, which TSMC dominates, doesn't get to build. If enough leading-edge volume is sucked out of the foundry landscape, then it becomes difficult for these companies to invest as rapidly in leading-edge R&D and capacity expansion.

However, any material threat from Intel's mobile products is unlikely to materialize until late 2015 or early 2016. This is when Intel will bring its high end, 14-nanometer Atom applications processor for phones and tablets codenamed Broxton to market. Further, Q1 2016 is when Intel will bring its integrated apps processor and cellular solution built on 14-nanometer to market.

This is when Intel-built products could become very competitive in the high-volume tablet and phone markets, and this is when Intel could stand to gain material share from TSMC.

Foolish bottom line
Over the next year, TSMC should enjoy a number of very nice tailwinds. It's likely that it has future Apple business in the bag, and Qualcomm will probably remain unchallenged in the smartphone applications processor market during the majority of 2015. (Any share that Intel gains with SoFIA -- its first integrated apps processor and modem -- is business for TSMC anyway.)

However, as we move into 2016, Intel becomes a much more serious threat to the semiconductor foundry players with its Atom products. While only time will tell whether these 14-nanometer Atoms will be enough to win high-volume phone designs, this is something that TSMC investors should keep a closer eye on as 2015 winds to a close.