Why Apple is buying Intel's 5G modem business

iPhone x
iPhone x (Image credit: iMore)

Apple is buying Intel's modem business for around $1 billion dollars. They're mainly LTE right now, but 5G is on the horizon, and they're already specing out 6G -- yeah, 6G -- and beyond that. It's a critical part of the hardware we all use every day, and Apple likes to control all of those critical parts.

The wheels have been in motion for a while but picked up recently when Apple settled with Qualcomm, the biggest modem maker in the game. Intel, who'd been providing Apple with modems during the spat, announced they'd be getting out of the business entirely, now that the other two had kissed and made up.

That left Intel's modem portfolio up for grabs, and much like Apple once bought P.A. Semi, licensed ARM's IP, and began making custom A-series chipsets for iPhones and iPads, the idea here is that Apple will buy that modem business, license Qualcomm IP and begin making custom modems as well.

To help me sort the facts from the wishful thinking, I sat down to talk with Ben Bajarin, principal at Creative Strategies, and founder of "Tech.pinions."

Rene: Ben, just before we get started, I wanted to ask you, there was this report, I'm going to use "report" for it, going around about Apple loyalty rates. I always look at these things, they get tremendous coverage but when you dive into the data, there's no data there. What was your take on that?

Ben: That's just the reality of Apple. Negative news helps. There's a lot of writers out there, I'm not going to name names, that start off with a negative bias towards Apple when they produce.

Those things do well because investors, especially in the investor community, because investors are always looking for that one nugget as to whether or not they should short the stock or be long, and so that's the reality of the news cycle. Something like that generates headlines when it's disingenuous. Smart people ask the right questions, others fall for the fad.

You didn't cover it. I didn't even tweet about it because I was like, "This is such a waste of space, of time, I'm not going to say anything." The reality is, since we're talking about it, is that no, we've seen no data and I have no research that indicates that their loyalty rates are dropping.

China is a different story. It's not that their loyalty rates are dropping. It's just that there's a buying freeze and there's dynamics there that are different, but across the board, US, parts of Europe, we're not seeing users switch to Android.

If anything, no one is switching anything anymore. They've made their bed in Android or iOS. We're seeing very little transition rates between customers at all because they're now just so stuck in each other's ecosystems.

Rene: Just to set the stage, Apple originally, if memory serves, used Infineon modems, but then they switched over to Qualcomm, was it when they went to Verizon?

Ben: Yeah, it was part of the Verizon deal, and then obviously they used Qualcomm until they switched to Intel.

Rene: Intel, in the meantime, bought Infineon, and so they started using the same base technology. How did the two technologies compare?

Ben: I think Qualcomm has always had a leadership position in modems. I don't think anybody that deeply digs into the technology specs would disagree with that. Obviously, when you look at the history of the industry, Qualcomm has been the leader. They have the most patents.

If you're an engineer wanting to go into wireless and modem, you go work at Qualcomm. That's the dream. They lead in network transitions. Obviously, they've been at the forefront of every time we move to a new G. I think undoubtedly they've been the standard for just quality modems for a lot of different reasons, but it's also a very, very tricky business.

Infineon didn't have bad stuff. When Intel picked that up, their goal was, honestly, when they started getting into this, was more because they wanted to bring connectivity to laptops and things like that. Then, it really was just smart phones.

If you want to read the tea leaves between those deals, Apple wanted to dual source. They started heavily investing with Intel on modems. In some degree, they had a lot of hands-on with Intel's modem development, which then led to the more exclusive deals that they had with Intel.

Yeah, it was all based on Infineon IP. There's not a lot of people with IP. If you want to make a modem, you don't just invent that. Apple could not have done this without some IP to do their own. I think the reality is, for them, there's a handful of places you go to get a license, patent, or a product from in modem and Infineon/Intel was one of those options.

Rene: For a while, we had this uncomfortable period where Apple was really pushing back against Qualcomm's licensing model and trying not to use Qualcomm modems when they didn't have to.

For example, if they didn't have to support CDMA in a particular phone, they would try to use the Intel modem instead, and that all resulted in this huge lawsuit that was then settled. Now, Apple has two licenses with Qualcomm, both for the modems and for the IP?

Ben: Yeah. They have a chipset deal. The details of that as a part of their agreement was a multiyear chipset deal. Interpret multiyear however you will, two years, three years, whatever, and then they have a six-year license to the Qualcomm portfolio of patents, which means Apple can take any of those that they want.

It's a patent portfolio license. It doesn't even have to be to modem. It can be for camera, it can be for RF, you name it. They have the chipset license and then they have the broader license to Qualcomm's portfolio, which is a longer deal as it seems than to just the chipset deal.

Rene: Is this why, because even in the beginning, you started talking about whether Apple would move in on Intel's IP, because Intel at the same time announced that they were getting out of the modem business or at least the consumer side of the modem business.

Ben: Yeah. I've been talking about this since 2014 because, you looked at the tea leaves, Intel has one customer for their modem. One. Do you, as an Intel, keep investing in a business that's very expensive and engineer sourced for just one customer, or do you shed that business?

My belief was always that it just didn't make sense for Intel to be, in this case, the smartphone modem business. This is not to say that Intel shouldn't be in the modem business because obviously if they want to bring LTE and 5G to tablets and notebooks and whatever.

The smartphone side, they had one customer, and so that's where it made a hard business decision. Obviously, the likely person, if you liked Intel's IP, was Apple. My conviction was, they would absolutely try to acquire it if they felt that the IP was good enough because if they don't, then why move away from Qualcomm?

Arguably, it's cheaper to keep working with Qualcomm over the long haul, given the deal they got and given that Qualcomm, I think, genuinely does have better technology than what Intel had.

Unless they felt that it was quality enough that they could build upon it, start a foundation and really start to invest in it, then I thought they would buy it. If they came to the conclusion that it was not good enough, again, why move away from Qualcomm?

This is all reported. If it happens, then I think it really just signals that Apple knows they need it. They know they need it to do their own modem, and they did feel that the IP was compelling enough that it makes sense. A billion dollars, if that's the deal, is really not a lot of money in Apple's scheme of things.

Rene: Is there also a hybrid approach to this where maybe Intel by itself, even with Apple's help, couldn't become competitive with Qualcomm or at least couldn't be competitive efficiently enough with Qualcomm, but Apple building off Intel's IP or the Infineon IP with the additional of the Qualcomm license could round that out?

Ben: Yeah, absolutely. This was one of the questions that I got quite a bit -- what does it mean for Qualcomm? I don't think that this is a net negative for Qualcomm because I do think that Apple still is going to need and want a bit of their IP, particularly around 5G. I think one of main issues that's become very clear was, Intel just didn't really have a clean path to 5G.

Network transitions are really, really hard. The first two, three four years of moving to a new network technology, it's hard to build the devices, it's hard to build the modems, so it doesn't shock me that Intel was not going to be ready day one for 5G.

That being said, I think because of the relationship that they have, and obviously a billion dollars for IP plus what they're paying for a license for Qualcomm, this is not a lot of money in the grand scheme of things for Apple.

If those are continuing licenses that make sense, that they can bake into now their IP...because Qualcomm obviously licenses to competitors. They license to Samsung who makes modems. To some degree, they license to Huawei, even though that's a sketchy situation because of what's going on in China.

It's not uncommon for Qualcomm to license standard central patents to do modems to competitors. In this case, it's a customer who owns IP. This is not that out of the ordinary, is what I'm saying. Absolutely, I can see this being a hybrid approach.

That's probably likely and smart because I think Apple is going to want cutting edge technology in this modem, and I think you're still going to have to get that from Qualcomm.

Rene: When Apple started getting into silicon, they bought P.A. Semi and they bought a bunch of companies. They also hired a bunch of engineers and they ended up using ARM references on it, but then just getting an ARM IP license and going with their own custom silicon. Is that a path that you could see them taking similarly with modems?

Ben: Yeah. No doubt they want to make their own modems. I think that that's been clear not just from reports, but from the hiring. I think if you talk to anybody in the silicon team, doing baseband has been a high priority, but it's also been a struggle.

It's one of those things that they would have needed a license from somebody else, whether it was Intel and/or Qualcomm or whatnot. They needed that IP because the portfolio of patents for modems is just so well covered. You're going to get that from somebody if you're going to get into that business.

The reality is, they've always wanted to do this. This is one part of the puzzle to do that, but again, it makes a lot of sense. If you think about where they're going with computers that we wear on our wrists, on our faces, in our ears, all of those things will require modems.

For them to be able to control the design, miniaturize it, which I think they're the leading standard at miniaturizing technology and putting it into things like ear buds or things like smaller watch or glasses in the future, they need to do that themselves. That's the inevitable path to control the modem and all of the silicon bits to design something that small.

They control their destiny at this point for baseband. I think that's a good spot to be, but that doesn't mean that they don't need other partners and other technology to do it. At least they can now design this bit theirself.

Rene: That was going to be my next question, is, what advantage do they get by doing this? As you mentioned, it could be less expensive just to use Qualcomm, but we've seen Apple does like to control their destiny, but also, doing things like building it into the SoC or doing systems in package like for the watch.

Maybe as the watch becomes more like the phone, the AirPods become more like the watch and all of that, it doesn't just take differentiation. It's nice to have differentiated modems, but it takes a lot of really purpose-built technology.

Ben: Yeah, and I think the point that remains is that to date, Apple has not shipped an integrated modem on any product. They ship their A-series processor and then they ship what we just call a thin modem.

The opportunity is to bring that design onto the chip, which gives you greater efficiency, better battery life, oftentimes better performance. There's benefits to bringing the modem chipset on to the SoC, and Apple has never done that.

That's why I think there's this opportunity for them to design this whole complete solution, which includes the baseband, and that will inevitably yield great results across every product.

I think it's essential for wearables, is my broader point, is that they would have absolutely had to figure this out to make a path forward in wearables.

Down the road, it will be exciting to see what benefits they get with a A-series processor in something like even Watch, or iPhone, iPad, Mac, etc., that does have an integrated modem because they will get benefits from a range of different things.

Rene: I know it's really hard to predict this kind of stuff, but what do you see is a timeline for this? They went from P.A. Semi, to the A4 to the A7. It took them several years, half a decade maybe. Do you think we're looking at the same kind of timeline here?

Ben: No, I don't. I think for 5G, it's going to take longer. If we're thinking about a device that requires a 5G modem, I would say for the foreseeable future, that product is probably going to be a Qualcomm product, but remember, Apple has already been using this technology. They've got all of the expertise to make a LTE modem with Intel.

I could see them, in a much shorter time horizon, doing their own LTE-based product in something like iPad, maybe even Mac, perhaps Apple Watch, which probably doesn't require 5G. There are things that are low-hanging fruit for them to do this with in the next one to two years, absolutely, but I don't think it's going to going iPhone.

The meta point is, anything that's 5G-based I think still requires Qualcomm's chips for the foreseeable future out this multiyear chipset license. A 5G Apple modem iPhone is a longer timeline, but other stuff can be a much shorter timeline for them to do their own thing.

Rene: Do you think that Apple would, because we talked about wearables already, but a lot of people would love to have integrated cellular service on a Mac, and apple hasn't done that.

We've heard about licensing fees maybe, or that Mac OS just doesn't have the same efficient networking that iOS has, but as Apple becomes a master of their own modem destiny, perhaps those calculations change too.

Ben: Yes, especially if you think somewhere in their portfolio exists a opportunity for an ARM-based Mac, then that would be a great product that they could build because it could be a thicker chip. It doesn't have to be as small, so they don't necessarily have to make it such a tiny chip.

They can use more die space, which gives them performance, GPU, all of the ancillary parts they put on those chips, including space for a modem. That could work. The broad point is that they've been doing this so much with Intel, they could absolutely put their own, even if it's a thin modem, in iPad or in Mac. They could do this very quickly with this Intel technology.

What would be a little more interesting is ways that they can tie software and services architecture to this. This is just speculating, but obviously Apple feels like services is a big part of their business. Services demand connectivity, and so as a part of those types of bigger looming business opportunities, can they more efficiently tie those services together?

For example, could video be better, smoother, cleaner, whatever? Maybe there's some augmented reality benefits that they get in terms of performance and tying the connectivity to the system because Apple is a master integrator. That's what it comes down to.

They integrate technology better than anyone, and they have the foresight to plan road maps based on that integration, which is what they've done with CPU and GPU. You see all these developer tools like Metal, great stuff that takes advantage of their proprietary architectures in silicon and software.

My meta point is, can they do things like that around connectivity if they control that stack and maybe give us better services, better augmented reality because they're tying the architecture to the solution? That's, to me, the speculative potential upside. [laughs]

Rene: Looking further out, 5G, we've all seen the recent tests that you've been putting up. [laughs] You walk one block in Chicago and the 5G disappears.

In terms of the 5G roll-out meeting up with Apple's timeline, do you think they'll get there as 5G starts to become more common, after that becomes more common, as we start talking about 6G already? [laughs]

Ben: I think if you look at the timelines for these things, every time we shift to a new G, it takes roughly 10 years for us to consider it mature. That doesn't mean Apple is going to be there in 10 years. I think 5 years out from now, it will be what we would consider mature.

You'll have critical mass for phones. Networking kinks will be worked up between millimeter wave, which is what gives you very short form bursts of energy versus the sub-600, and then a range of other frequencies that will help bounce that out, so that you don't walk a block. You could be wherever you want and till get gigabyte speed.

The next few years, we're going to go through some hurdles of 5G. Yes, great, Apple probably has a 5G iPhone next year. It's going to be great in spots. It's probably not going to be all the glory it will be in four to five years.

If you think four to five years is perhaps your timeline for an Apple integrated 5G modem, yes, I think these kinks will be worked out with 5G in that four to five-year time frame. That meets up.

Rene: Beyond obviously the big Intel modem purchase, are there any other signs you're going to look for in terms of Apple going to the modem business?

Ben: Yeah, RF. Basically, there's two companies for them. There's Qorvo and Skyworks. Apple will probably use Qualcomm's RF right off the bat for 5G, just because that's really tricky stuff and it's going to work really good together.

I do think you need to control some RF if you want to be in the modem business, so I would be watching for an acquisition of someone like Skyworks or Qorvo if Apple does acquire Intel's business because RF is a key part of that solution that it makes sense for them to control.

You can follow Ben on Twitter @BenBajarin and read all of his work at Tech.pinions.

○ Video: YouTube
○ Podcast: Apple | Overcast | Pocket Casts | RSS
○ Column: iMore | RSS
○ Social: Twitter | Instagram

Rene Ritchie

Rene Ritchie is one of the most respected Apple analysts in the business, reaching a combined audience of over 40 million readers a month. His YouTube channel, Vector, has over 90 thousand subscribers and 14 million views and his podcasts, including Debug, have been downloaded over 20 million times. He also regularly co-hosts MacBreak Weekly for the TWiT network and co-hosted CES Live! and Talk Mobile. Based in Montreal, Rene is a former director of product marketing, web developer, and graphic designer. He's authored several books and appeared on numerous television and radio segments to discuss Apple and the technology industry. When not working, he likes to cook, grapple, and spend time with his friends and family.