Apple’s 5G Strategy Still No Clearer despite Acquiring Intel’s Smartphone Modem Business

After some initial speculation about a potential bidder, Apple has been revealed as the company set to acquire the majority of Intel’s smartphone modem business for US$1 billion. Under the signed agreement, which is expected to close in 4Q 2019 (subject to regulatory approvals), approximately 2,200 Intel employees (including engineers) will join Apple, along with Intellectual Property (IP), equipment, and leases. Apple will acquire around 8,500 assets from Intel’s patent portfolio, including 6,000 patents related to 3G, 4G, and 5G cellular standards, comprising a combined portfolio of over 17,000 wireless technology patents ranging from protocols for cellular standards to modem architecture and modem operation, which will give the company a significant amount of 5G-related IP.

The announcement that Apple is to acquire Intel’s modem business ends an interesting 12 months of ongoing battles between the iPhone vendor and its modem suppliers. Back in July 2018, and as a consequence of its ongoing spat with Qualcomm, Apple stated it would make Intel its sole 5G modem partner in iPhones or else it would consider building them itself. In the year since then, that assertion has taken a few twists and turns: not only has Apple patched up its differences with Qualcomm, but Intel also announced it was exiting the 5G smartphone modem business. The revelation that Apple is now buying Intel’s modem business is just the latest in the ongoing saga, which may not be over just yet.

While every other vendor sources modems, application processors, and System-on-Chips (SoCs) for their smartphone portfolios from a variety of suppliers, Apple, which has been using its own processors since 2010, only relies on the supply of modems. Intel has provided these modems to Apple for years; prior to these announcements, they were on target to appear in about 70% of iPhones, with Qualcomm modems in the rest. Apple generally uses Qualcomm modems in iPhone model versions that run on legacy Code Division Multiple Access (CDMA) operator networks (mainly those in the United States, Japan, and other parts of Asia) and Intel modems for operators from the Global System for Mobiles (GSM) world.

Apple’s acquisition of Intel’s 5G modem business will allow it to not only gain world-class expertise in cellular modems, which is one of the most challenging areas in Research and Development (R&D), but also strike better deals with other major 5G patent holders. The acquisition will enable Apple to potentially negotiate better terms with its current 5G suppliers and integrate parts of the 5G modem into its existing processors. It is important to note that, by inheriting some of the key 5G patents developed by Intel, Apple will be able to differentiate its offering in what will soon be a highly competitive 5G landscape.

Through the settlement with Qualcomm—which includes a six-year patent license deal and a multi-year agreement for Qualcomm to supply smartphone modems for iPhones—Apple can take full advantage of Qualcomm’s innovation in the short term, which will help accelerate its move to 5G, while the Intel purchase enables a more strategic long-term vision and will allow Apple to further differentiate moving forward. Making use of its own modems and those from Qualcomm in the future also helps continue Apple’s strategy of using more than one supplier per component, as this approach has helped it reduce risk and increase its ability for negotiation in the past.

On the face of it, the acquisition makes perfect sense for Apple, paving the way for it to design and build its own modems, which had been the one main elements missing from its component strategy. However, while building its own processors, such as Central Processing Units (CPUs), Graphics Processing Units (GPUs), or Inertial Measurement Units (IMUs), and Wi-Fi and Bluetooth chips is one thing, manufacturing modems is another. This is an extremely complex and expensive business to develop, and will only get trickier with 5G as Apple will also have to develop its own Radio Frequency/Radio Frequency Front End (RF/RFFE) solution. As such, and owing to this increased level of complexity, Apple’s own 5G modem chips would not be ready until 2021 at the earliest. Moreover, creating a fully integrated SoC that utilizes its own Bionic application processor and new 5G modem will take Apple at least two to three years to perfect. Although the time it will take for Apple to release an SOC is an issue, such a move makes sense as integration at the chipset level brings a whole host of benefits: notably, reductions in cost, greater power efficiencies, and better functionality.

Moreover, entering the world of modems brings Apple into a whole new environment that it has not been previously exposed to; it will need to work closer with the 3^rd Generation Partnership Project (3GPP) and mobile operators to ratify its modems, while working extensively with RF vendors, building and harmonizing with the ecosystem. Up until now, Intel’s and Apple’s representations at this level have been quite limited, so Apple will need to work on building these relationships if it is to prove successful, but this comes at the price of both additional time, effort, and, ultimately, cost.

In the short term, it is quite clear that Apple will have to rely on Qualcomm’s 5G modems, especially if it wishes to launch a 5G iPhone lineup in 2020. Indeed, Apple buying Intel’s 5G modem IP does not mean that it can now suddenly produce its own modems any faster than would have been possible previously. From 2021 on, though, Apple could potentially minimize its reliance on Qualcomm through the release of its own offerings, although there is uncertainty regarding how this would impact their licensing agreement. Specifically, an important aspect of the Qualcomm settlement was that it would become a key supplier of 5G modems for Apple’s future generation of iPhones. Furthermore, it is not clear if Apple will just use the 5G modem from Qualcomm or if it will also take advantage of Qualcomm’s RFFE solution, which together can provide an integrated system design. Such a move may call Apple’s relationship with current RFFE suppliers, such as Skyworks Solutions, Qorvo, and Broadcom, into question, although it may continue to use these vendors in conjunction with Qualcomm’s solution across certain product segments in line with its multi-supplier strategy.

Either way, Apple will need to hit the ground running when its 5G iPhones are finally let loose on the market. Rumors suggest there will be three iPhone models launched in September 2020 and all will feature support for 5G, as well as sporting Organic Light-Emitting Diode (OLED) displays. This step needs to be taken so that Apple can compete more effectively at time of launch with lower-cost 5G Android smartphones, which by then will have been on the market for well over a year. Furthermore, the launch date of 5G iPhones would probably make Apple the last of the major manufacturers to release a 5G smartphone, and it is highly likely that the price of competing 5G Android smartphones will have fallen below the US$400 mark by then. This timing, though, may actually fall in Apple’s favor, as Qualcomm will have achieved a certain level of scale (it is a supplier to many of the Android vendors) and so Apple is sure to benefit from the availability of cheaper 5G modems. It remains to be seen, however, what price points Apple will achieve across its devices, especially if it sticks with a lower priced model in its portfolio, because it cannot afford to offer a portfolio where entry price for a 5G iPhone is over US$900 and expect no negative effect on sales. In addition, Apple will almost certainly have to support both mmWave and sub-6GHz spectrum somewhere across its models, the former being a major requirement to serve its vital U.S. market, which again will add to the cost.

Alternatively, there is every possibility that this is not what Apple is planning at all. It may not build its own modems, but instead mainly use Intel’s IP to negotiate better royalty rates, notably from Qualcomm, for itself and its partners. Intel has some crucial Standard Essential Patents (SEPs) in 5G, most likely including those previously held by Infineon, which would be useful if Apple finds itself in more court battles further down the line. Intel also has some interesting contributions to 3GPP that Apple may want to use to differentiate its offerings, notably 5G positioning, and could therefore outsource customized modems that are based on these differentiators.

The most likely scenario is that Apple will probably do all of these things to maximize any benefit that can be squeezed out of its modem business acquisition. Leveraging Qualcomm’s 5G expertise in the short term makes perfect sense and, initially, buys Apple some time to contemplate what it may be able to achieve on its own, making full use of its inherited IP while also creating 5G modem differentiation. What is in little doubt, however, is that this all indicates more expected points of contention, which will likely lead to further installments of the ongoing Apple-Qualcomm IP courtroom drama.