In a previous article, I went over how Taiwan Semiconductor Manufacturing Company (TSM 0.60%) has a somewhat risky dependence on smartphone giant Apple (AAPL 3.40%), particularly to support its business selling chips manufactured using cutting-edge technologies.
TSMC has done a good job over the years capturing Apple's business. It has manufactured the custom-designed A-series processors for Apple's last four iPhone generations, and has been the exclusive manufacturer for three of those four.
After reading through a transcript of TSMC's Oct. 19 earnings call with analysts, I get the impression that although TSMC expects to win some portion of Apple's A13 chip business -- perhaps even most of that business -- it's not confident that it'll have the entirety of the orders to itself.
Allow me to explain.
Image source: Apple.
A question and a wishy-washy answer
During the call, analyst Sebastian Hou asked TSMC management to provide an estimate on the company's market-share expectations for its 7-nanometer technology (which should start being mass-produced next year) as well as its 7-nanometer+ technology (which should go into mass production the following year).
TSMC's co-CEO C.C. Wei, replied that he hoped that the company's market share with those technologies would be "as high as possible," but he couldn't commit to a specific market-share percentage. The exec did say that the company expected to receive 50 tape-outs (completed chip designs for manufacturing) by the end of 2018.
Wei's noncommittal comments stand in contrast to what the company was saying awhile back about its market-share expectations for its 10-nanometer technology, used for Apple's A11 Bionic chip. On TSMC's fourth-quarter 2015 earnings call, for example, the co-CEO said that the company intended to enjoy "very high market share" with its then-upcoming 10-nanometer technology and intended "not to lose it."
What does this mean?
I don't think Wei was trying to hide expectations of significant market-share loss. Instead, I suspect that there are still some things up in the air, particularly when talking about expectations for the company's 7-nanometer+ technology. There's good reason for that.
TSMC's main rival, Samsung Electronics (NASDAQOTH: SSNLF), isn't expected to go into "initial production" on its 7-nanometer Low Power Plus (LPP) technology, which will use a new lithography technique known as extreme ultraviolet, until the second half of 2018. I suspect that chips built using this technology won't hit the market until 2019, which will be too late to manufacture Apple's A12.
However, Samsung might still be in the running to manufacture Apple's A13 chip. Apple should have two options to choose from here: TSMC's 7-nanometer+ technology (which is expected to provide both area and performance benefits over the initial 7-nanometer technology), and Samsung's 7-nanometer LPP.
My guess is that Apple will put the work in to design for both, just to give itself options, but that it'll ultimately let Samsung and TSMC know how much of the A13 manufacturing business they've won -- if any at all -- sometime next year, depending on how the technology development progresses for both chip manufacturers.
