Apple developing own iPhone cellular modem, hirings and insiders say

AppleInsider

A new Apple job listing is suggesting that the company is looking to break the shackles of Qualcomm and Intel for the iPhone in the future, raising the possibility again that Apple might make its own cellular modem.

Intel's iPhone modem with penny for scale

In a job listing posted in early December, Apple is seeking a cellular systems architect to work out of San Diego. Why specifically in San Diego isn't clear -- other than effectively being Qualcomm's backyard -- but sources told The Information that Apple has plans to build a new cellular modem chip out of that location.

While the job listing is new, Apple's intentions to develop its own cellular modem have been fairly clear based on recent hires. In 2014, Apple recruited at least 30 mid and senior-level RF engineers from Broadcom and Qualcomm, suggesting an intent to begin working on its own internal baseband chips.

Executives cited for the project are Bernd Adler, previously of Intel's modem team and Esin Terzioglu, a former vice president at Qualcomm and presently Apple's wireless chip head.

The report from The Information suggests that any Apple-designed chip is three years away from shipping.

wood1208

Would be nice if Apple can embedded it's own modem into A-series chips in 2021 iPhones. Apple should also develop own WiFi 6 and BT 5 chips. Has to happen to drive the components cost down and avoid IP royalties,issues.

ksec

wood1208 said:

Would be nice if Apple can embedded it's own modem into A-series chips in 2021 iPhones. Apple should also develop own WiFi 6 and BT 5 chips. Has to happen to drive the components cost down and avoid IP royalties,issues.

Apple is definitely working on their own WiFi 6 and BT 5 solution, across every Apple product uses WiFi and BT, and it make sense to bring them in. Although I think that is for 2020. I doubt Apple is making their own Modem though, they are more likely license a fully working Mediatek blueprint and fab it themselves. 

tht

I would bet a lot that Apple has been dabbling making their own cellular modems for 6 years now. There’s always going to be argument about what is the value add of doing this. The only value really is less power at comparable signal strength to other modems, more signal at comparable signal strengths, etc. If it performs the same, they aren’t adding much with productising it.

The patent licensing fees don’t really change if they make there own. Just not a lot of advantages. If Apple is able to set licensing fees based on component costs however, watch out! A custom modem than would be low single digit dollars instead of high tens of dollars. That’s big. LTE modems cost about $130 for iPads and tablets. That could drop to basically free if the licensing structure changes.

There isn’t much reason for LTE not to be in virtually all computing devices. They should be in all laptop computers, all tablets, all cars, all GPS devices, etc. The current licensing structure effectively stops this from happening as it is too expensive.

zoetmb

They should do their own, but even if they do, they would have to do so without violating any current patents, otherwise they'll still be in litigation.   And you just know that Qualcomm will be looking at whatever they do and will claim that they have violated their patents.  

Apple needs to do their own not just to escape Qualcomm because of the legal disputes but because Qualcomm is doing quite badly.   They had a net income loss in 2018 of $4.86 billion.   In 2014, they had net income of $7.54 billion and it's been trending down since.   

Market cap is still about $71 billion which is ridiculous for a company that's losing money big time.   If they were still making decent profits, I'd argue Apple should buy them.  

hodar

How does Qualcom win in this scenario?

Alienate a major player into creating his own sets of Modems, specific to his needs.  This player has more than ample cash on hand, to simply buy you.  So, you continue to harass this "customer" until he mixes your product, with your primary competitor.  In a year or two, this major source of cell phone revenue will have built his own specific modem, and due to existing technology, he can modular it such that this new product not only serves him; but also can be sold to other existing customers (now at a discount).

This seems like a stupid war, where the best that Qualcomm can hope for, is going bankrupt (or severely reduced in customer base).  Because LTE is a STANDARD, they cannot demand royalties on implementing it, only royalties on "how" the protocol is implemented.  I do not see Apple utilizing the same techniques as Qualcomm, as they likely already have in-house solutions that will accomplish the tasks required - it's simply a task to organize these patents into a modem tailored to their needs.

wizard69

Apple has been working on this for a very long time!    As such it would be really hard to guess a delivery date.   

macgui

Apple charges $130 for a modem/cellular radio for the iPad, and $100 for the Watch. I doubt very much they cost Apple anywhere near that. 

I further doubt we'd see a significant, if any price drop in retail pricing. At best, Apple might go to their previous pricing model— more features and/or performance for the same price. And I wouldn't expect that to last.

xbit

macgui said:

Apple charges $130 for a modem/cellular radio for the iPad, and $100 for the Watch. I doubt very much they cost Apple anywhere near that. 

The main cost is licensing and licensing fees are primarily based on the selling price of the device. Licensing those patents isn't $130 expensive but it's more expensive than you'd think.

tht

xbit said:

macgui said:

Apple charges $130 for a modem/cellular radio for the iPad, and $100 for the Watch. I doubt very much they cost Apple anywhere near that. 

The main cost is licensing and licensing fees are primarily based on the selling price of the device. Licensing those patents isn't $130 expensive but it's more expensive than you'd think.

The licensing costs for cellular is the primary reason why cellular modems aren’t ubiquitous in computing hardware, like WiFi or Bluetooth is, and QCOM and probably holders of the cellular patent pool want it that way. 

Apple is big enough to negotiate favorable terms with QCOM and got a rebate, while QCOM maintained its licensing structure. But if Apple wants to by from another party like Intel or MediaTek or build their own, the licensing structure makes doing it expensive. There’s a reason why Samsung decided on shipping QCOM SoCs and modems in the USA instead of their own, all these years now. 

Apple won’t discount LTE products by $130 or $100, primarily because it prices products based on price tiers and add hardware that represents that cost. If the licensing was cheaper, it may mean more RAM, more storage, bigger charging bricks in the box, better other hardware item, etc. That’s where the win is. 

Another big win is that cellular modems could be just part of the marginal cost of computing chipset, like WiFi is. Are there good reasons why cellular support isn’t part of a laptop chip sets from Intel? Other than licensing costs?

spamsandwich

I think it would be a relatively easy thing for Apple to hire away the top cellular modem chip designers in the world. Start with Qualcomm by hiring away ALL of their key personnel.

mattinoz

SpamSandwich said:

I think it would be a relatively easy thing for Apple to hire away the top cellular modem chip designers in the world. Start with Qualcomm by hiring away ALL of their key personnel.

What would they need that's not cover by the Apple designed parts of an A series chip?

They have Signal processing, CPU cores big and small, neutral and gpu. Most of the modems break out Antenna handling to a seperate chip. What's missing for Apple to make a software modem?