Bloomberg’s Mark Gurman has done some more reporting on Apple’s chip business (he has been busy lately). His sources confirmed that Apple is in fact building their own cellular modem. Apparently, the head of Apple’s chip business mentioned this in an internal Town Hall meeting, and the story knocked a few points off Qualcomm’s stock price.
Without diminishing Gurman’s solid work, this is not exactly news. Apple has been rumored to be working on its own modem for a very long time. We wrote about this a few times last year especially around the time Apple settled its long-running lawsuit against Qualcomm. In that agreement, Apple maintains the option to buy Qualcomm chips through 2026. And of course there was Apple’s purchase of Intel’s modem business for a $1 billion, and Apple’s growing campus around the corner from Qualcomm’s San Diego headquarters.
At the time of those developments, we thought 2026 was a reasonable estimate for the amount of time it would take Apple to build its own modem. This is not going to be easy. We have called Apple Silicon the best run semiconductor company in the world, but even they do not have a great track record with communications chips. They tried for years to build their own Bluetooth/GPS/Wi-Fi chip, but eventually gave in and stuck with Broadcom. Moreover, designing a modem is a very different task than building a CPU. Cellular modems have to contain all the vagaries of the wireless standards going back to 2G. Apple can afford the team to do that, and buying Intel’s IP and team help a lot. Once upon a time we would have questioned whether Apple could pull it off, but given their track record recently, our only question now is how long it will take.
Apple’s M1 has shown that by transforming a CPU into an ASIC that contains a CPU, tightly integrated with Apple’s software, they can break the traditional model for merchant CPUs. The reports of performance and battery life for the new Apple CPU are an earthquake to the 50 year old CPU market. In that light, we should think about what an Apple modem will do that market.
For years, Apple has done a tremendous job of integrating the iPhone’s ‘CPU’ (aka applications processor aka the N Series of chips) leading to big gains in battery life and performance. Expect them to extend those gains. In our series (parts one, two & three) last month looking at Apple’s motivation for building their own chips, we concluded:
Apple Silicon gives Apple a distinct competitive advantage. Or to put it in terms that Apple is more likely to use, its silicon gives its customers a better experience.
So what will an Apple modem provide its users? We can almost see the launch slides now from September 2026. They can talk about how their modem offers better reliability, faster speeds and noticeably better life. They will probably also be able to offer proprietary communications features available when two iPhones connect – maybe voice and video effects. This could be the equivalent of the blue vs green bubbles situation for text messages.
We can already hear the objections of our readers at Qualcomm. “Apple does not understand telephony. 5G is hard. 5G mmWave is harder. They don’t have [insert an ocean of acronyms].” And given the company’s past history of controlling the off hand remarks of its executives about Apple, we will likely hear some of those comments in public soon. To those people, we respond – get ready, Apple is coming, they know Qualcomm’s weaknesses, and they can outmarket Qualcomm a million times over. Not to diminish Qualcomm’s immense technical capabilities, but they are constrained by lowest common denominator problems of having to serve a large number of customers. Apple will optimize for their own needs and specialization can bring great performance advantages. Apple may fail, but Qualcomm should probably start getting the lawyers ready, because it will almost certainly come to that.