Apple accelerating plans to make a jump to TSMC for mobile chips

Last time we heard, Apple was going to move its mobile chip production from Samsung’s $14 billion Texas facility to Taiwan Semiconductor Manufacturing Company (TSMC), the world’s largest dedicated independent semiconductor foundry, some time in 2014. But according to Taipei Times, the company could make the jump earlier than expected as it’s now looking to tap TSMC’s fabrication capability to make next-gen processors for iPhones and iPads as soon as the second quarter of 2013, using TSMC’s 28-nanometer process…

Taipei Times has the story (via MacRumors) based on a Credit Suisse quote:

Nigam said their checks also suggested Apple might start chip production at TSMC by using the Taiwanese firm’s 28-nanometer (nm) technology, rather than its prior views on the 20nm node.

Credit Suisse and others previously expected that TSMC would land orders for mobile chips late in 2013 or by early-2014. Note that Credit Suisse is expecting any TSMC deal to be a net addition to the existing silicon deal with Samsung.

Currently, the engine for the iPhone and iPad is being built exclusively at Samsung’s $14 billion facility in Austin, Texas, seen below.

As TSMC takes over an unknown portion of iPhone and iPad chip orders, Advanced Semiconductor Manufacturing Cois is expected to package the chips, according to the report. The two firms will be ramping up capacity “for a sizeable new customer” starting some time in Q1 2013.

DigiTimes, a hit-and-miss trade publication, said recently that the rumored order from Apple has already raised concerns over TSMC’s profitability because the foundry is at risk of upsetting its major buyers such as Nvidia and Qualcomm.

Apple allegedly offered one billion dollars in exchange for exclusive access to TSMC’s production output, but the company turned down the lucrative offer because “TSMC wants to retain control of its plants, doesn’t want to sell part of itself and doesn’t need cash for investments”, per its finance chief Lora Ho.

First-world problems, no?