Apple vs. Qualcomm – Black magic

Apple is likely to stay away from the black art of cellular radio.

• News articles and recent hires by Apple suggest that Apple may be moving into developing its own basebands.
• This could have significant negative implications for Qualcomm for whom Apple makes up a substantial portion of sales.
• This would imply that Apple wishes to integrate its in house applications processor with a baseband in order to produce a smaller more power and cost efficient component.
• Given Apple’s desire to vertically integrate its products, this move would make some sense but I think it extremely unlikely for the following reasons:
• First: A lot of the value in silicon has moved from the baseband into the applications processor. Hence the cost saving of moving to internally source a baseband would not be that great.
• Baseband-only chips are way down the rate cards of Qualcomm, MediaTek and Broadcom.
• Furthermore, much of the integration with hardware and software in smartphones occurs in the application processor where Apple’s solution is already well developed.
• Second: RF is really difficult.
• Cellular radio is a black art where experience often counts for more than pure engineering expertise.
• This will make organic growth of an in house baseband a very long, expensive and laborious process.
• Furthermore, it is unlikely to ever really catch up with Qualcomm and MediaTek in terms of product quality and cost.
• Third: If Apple wanted to get into basebands why did it not outbid Broadcom for the Renesas baseband business?
• This team originated from Nokia and consequently has a wealth of RF expertise and most importantly an excellent LTE solution.
• This team is now the last hope of Broadcom’s baseband business but it could easily have been the core of an in house baseband at Apple.
• This is an area where I suspect it will be cheaper and better to buy in.
• Fourth: The pedigrees of the two hires that we know about, suggests expertise in the components that sit around the baseband (such as transceivers) rather than the baseband itself.
• There are some difficult problems coming as LTE becomes more mainstream that relate to radio bands and filtering which will be important for Apple to have some expertise in as things get more complex.
• Hence, I think these hires are about Apple moving to understand the radio complexities of its own products rather than the cellular protocols themselves.

Take Home Message

• At a high level, Apple’s history suggests greater vertical integration but inclusion of the baseband in that strategy makes no sense.
• In all probability, an in house baseband would probably end up costing Apple more and resulting in significantly worse performance.
• Hence, I suspect that Apple will concentrate on that which it knows best and leave the dark art of radio to those that have been doing it for 20 years.
• Any resulting weakness in Qualcomm represents an opportunity.