Friday Apple Rumors: Alternate Choice (Intel and not TSMC!)
Alternate Choice: Despite a growing chorus of reports that Apple will partner with TSMC to produce chips for its mobile devices, Forbes notes that many Silicon Valley experts increasingly expect it to tap Intel instead. Apple has been reportedly moving to reduce its dependence on Samsung as a manufacturing partner as the two companies have become global competitors in the mobile device market. Previous rumors have suggested the TSMC has begun preparing test runs to make Apple’s A6X mobile processors. However, MKM Partners analyst Daniel Berenbaum suggest that Intel is a more logical choice for the role since it’s greater chip-making capacity could allow it to act as an Apple chip foundry without the need for new capital investment. Rumors suggest that TSMC has been investing heavily to accommodate Apple’s anticipated orders. But Berenbaum says that TSMC’s chip-making capacity remains too restricted to meet Apple’s heavy demand.
The Intel margins are currently in the 60% range with 55% expected this quarter. TSMC margin is 45% as a silicon merchant.
TSMC capacity, yield and historical performance are risks that would be different with Intel as the vendor. Those would have value to a company that needed their CPU on time and in appropriate quantity (Intel).
Margins would probably move down from the 60% - 65% range down to the 55% to 60% range which is where they were when Intel was selling all the netbook Atom chips several years ago. It dropped to a low of 50% during the recessions but was at about 58% during the netbook peak.
There is tremdous value to Apple by having access to Intel fabs. It would help them compete better with Samsung who is chewing into their iPhone market share.
While Intel margins would move down if they became a foundry for Apple, the new revenue number would sky rocket. The lower margin % of the larger revenue number would more than offset. Intel will not fab for Apple without making a profit. Both the extra revenues and profit would be additive to Intel.
That is the bogus argument that Covello makes. One of his unspoken assumptions is that Intel revenues remain constant. If so, his margin argument makes sense. For the last two years, his argument has been wrong.