It’s official. Apple is buying Intel’s smartphone modem business. Note, Intel still has their modem business for products like PCs and tablets and the IP they were selling was specific to smartphone modems. Apple has been working with Intel for years now integrating their modems, so they had a very good sense of the underlying technology and if they could build on it over the long-haul.
This was relatively predictable. I started writing about this possibility in 2014 and then kept hinting at it even as Intel and Apple got closer in their joint modem efforts. It became very clear, quickly, that Apple would be Intel’s only smartphone modem customer and Intel was bleeding cash trying to keep up with Qualcomm and failing. It simply did not make sense for Intel to keep investing in smartphone modems, really modems overall, in my opinion. The only question I had, with regard to Apple buying this business, came after their license and chipset deal with Qualcomm. Apple appeared to be in a position to utilize Qualcomm technology, which is better than Intel’s, and it’s unclear if Apple can develop a better solution than Qualcomm. Their partnership with Qualcomm came off as a win-win for both companies.
In this article, I’m going to cover two points. The first being Apple’s long game here for building their own modems, including how 5G fits in. The second is what it means for Qualcomm.
Apple’s Long Game
The main point I want to make here is that Apple is yet to ship a product that includes a modem on their own designed custom A-series processors. Apple has always had to include a thin-modem (a dedicated chip to just modem) in all its products containing A-series processors like iPhone, iPad, and Apple Watch. There are great benefits to having a complete SOC (system-on-a-chip) that includes CPU, GPU, modem, and other things. The modem was the big missing piece, and Apple can get a number of benefits by building a total chip solution that includes the modem and fine-tuning their hardware and software for this new fully integrated chip.
Things like battery life, connectivity optimization, even security, and more are all benefits to tuning a system with a complete SoC. Qualcomm sells a complete solution to everyone, but Apple (they just sell Apple the thin-modem) and Qualcomm always highlights the benefits they get in overall system performance by having a completely integrated SoC. In fact, many of Apple’s competitors benefit from this same dynamic, and now that Apple has the capabilities to integrate their own modem designs onto their own SoCs they will get a great deal of system optimization and performance benefits. I’m not saying Apple’s modem will be faster than Qualcomm’s, only that they can tune the system better than they could before.
This is the reason why I’ve known for a long time that Apple’s end goal was to build their own modem. But you don’t just wake up one day and build a modem from scratch since baseband patents, and IP has the whole industry covered. With Intel’s IP, Apple is now able to forge a path to do their own modem designs with minimal license and IP fees to other companies. More on this in a bit.
Will Apple’s First Internal Modem be 5G?
No, I don’t think so. Primarily because Apple has other products that are low-hanging fruit for 4G/LTE. Particularly iPad and Apple Watch. Both these products do not need 5G in the next few years but would also benefit massively from a fully integrated Apple chipset with a modem integrated onto the SoC. Apple Watch, in particular, is my likely bet for the first product to see an Apple-designed modem, but it will be 4G/LTE, not 5G.
About Apple Watch, the products I’ve always figured were the primary reason Apple was working on designing their own modem is their wearable product line. Not just Apple Watch but AirPods and whatever else Apple has in the five-year pipeline. Remember I said Apple needed a separate chip for their modem connectivity. This separate chip takes up valuable space on the motherboard, and when you are talking about shrinking electronics to small sizes that we will wear, you need all the space on the motherboard you can get. Integrated the modem onto the SoC is essential for Apple’s wearable business and product roadmap. This is the ultimate endgame for a fully integrated Apple SoC.
Where Does this Leave Qualcomm?
Apple buying Intel’s smartphone modem business and IP is not necessarily all bad for Qualcomm. The deal Apple struck with Qualcomm includes a multi-year chipset license and a long term deal to access Qualcomm’s IP portfolio. The chipset deal is likely 5G modems for iPhone until at least 2022. And the license to IP is where Qualcomm may have some technology that Apple can use to build its own 5G modem.
Intel was nowhere close to building a 5G modem. In fact, I’m not even sure they would have got there anytime soon. The deal Apple made with Qualcomm is because, at a technical level, Qualcomm is the only 5G game in town for the next few years. Apple may not even be able to do a 5G modem on their own, and having a licensing relationship with Qualcomm can help them fill the gaps. Obviously, if what I’m suggesting pans out, Apple relationship with Qualcomm could be a long one, but it also won’t be as financially lucrative as Qualcomm would like.
During this 5G transition, Qualcomm is well-positioned and a strategic partner for Apple. Come to the 2023-2025 timeframe, 5G will be quite mature, and perhaps this is the proper timing for Apple with their own designed modem for iPhone. The bottom line here is Qualcomm still has some of the best wireless technology in the world, and Apple has a long-term license deal to that technology. If for whatever reason, Apple can’t do better than Qualcomm’s IP then I’d like to see them still leverage it somehow.
The last point, I’m watching Apple to make a move on an RF antenna provider like Qorvo or Skyworks. RF is critical to modem design, and Apple has worked both these companies before. Verticalizing on RF could help Apple even more efficiently optimize and tune their hardware, software, and services.
