Will Apple get its own fab?

Maybe it is starting to seem like I'm a fan of David Manners. Maybe I am. He covers the semiconductor industry in pithy posts that are easily digestable and informative.

Mr. Manners recently covered TSMC's CEO statement that it may make sense to build dedicated fabs (semiconductor manufacturing plants) for select customers. The comments are likely Morris Chang's attempt to lure Apple away from its present foundry partner Samsung. With the legal battles going on, you may think that Apple and Samsung make strange IC production bedfellows. If Steve Jobs was alive today, the lawsuits on the consumer product design front might play a more significant role in the foundry decision. I don't think that is the case today though.

This is partially out of certain fabless semicos recent challenges getting sufficient production capacity at the most advanced technology nodes. (Manners noted Apple's decision to stay at 40 nm because of difficulty with 28 nm yield and resultant available product volume.) Qualcomm would be a candidate for its own fab from amongst TSMC's present customer list.

Will Apple get its own fab? That's hard to say. I think the question goes beyond a single fabless semiconductor company. I think it depends on the fabless - or outsourced manufacturing - model and how it evloves. What's interesting is that we may be seeing the start of several trends back to the ways things were done in earlier days of integrated circuit production, and the foundry model could be just one element of this.

There are many avenues to explore on this topic. It was tempting to title this post "Part 1," but I am following the advice of head office by keeping the titles shorter. Are electronics trends coming full circle like so much fashion and pop culture?

It's either a sign of the times or my own advancing years, but yes, it certainly does seem like everything old is new again. And this is only Part 1.