Apple Extends Deal With Qualcomm as Custom 5G Modem Delayed

[Velin]

If you read prior analyst reports, they conclude this is one area that is really difficult. Not only are you designing the chip, you have to constantly engineer for frequencies/bandwidth/changing standards. And to make it all work from week 0 through end of product life is a herculean task.

Two views. One is, just hand this off to Qualcomm/Intel and direct them using your design specs. Least amount of risk.

Another view is, it would be amazing to see just how good connectivity could be with an Apple designed radio focused solely on iPhone, wifi iPad, Apple Watch, etc. A dedicated modem/radio focused just on iPhone, for example, I want to see just how good it would perform.

 

[macbookguy]

I remember the sucktastic Intel modems well. Oh how I hate them but I was on AT&T at the time and stuck. I was dreading an Intel 5G modem.

I’ve been holding out hope that Apple could turn it around but we’re 4 years out now and they’re saying it could be three more. I am curious what Apple could do but it’s looking like not a lot. I certainly don’t want anything inferior to Qualcomm having lived through it.

I am curious if this will be cancelled altogether in a year or two or if they’ll figure it out.

 

[fatTribble]

That's far from reality.

Qualcomm has a rich history of developing algorithms and communications modem chips starting with two of its founders, Drs. Andrew Vitirbi and Irwin Jacobs. Long ago they pretty much wrote the book with respect to modern digital communications science. Qualcomm as a company has loads of patents protecting their work.

It's so much more than just designing chips.

Thanks for the background! It all makes sense as to why Apple is still using their modems.

 

[k1121j]

It’s better for “your” phone security that Apple does not make its own modem. As long as the modem is by another company some walls will always be up. If Apple makes their own modem there will probably be no security walls and we will just have to “trust” them that they won’t abuse this much control…

This is the most illogical comment ever.

 

[ilikewhey]

i'm all for apple's own modem so long its on the same standard as whatever QC releases during that time, i still have nightmare with my iphone 7 with intel modems.

 

[SoldOnApple]

Developing a reliable modem is obviously a lot harder than Apple thought.

 

[GeoStructural]

Are you joking? It's the other way around! If Qualcomm were to ever lose Apple as a customer, their share price would tank! A quarter of all revenue comes from 1 customer! That's not how you should run ANY business.

Apple have Qualcomm by the cahoonies on price and delivery.

They did part ways at some point... Qualcomm ended up doing fine even though it was a big hit, they sued Apple and later restored their alliance, which was the wisest decision for Qualcomm, and the best for the consumer since iPhones with other chips were experiencing poor connectivity issues.

 

[GeoStructural]

Supposedly. I travel quite a bit within the US and have yet to see mmwave.

I have no use for it as a I never download anything or stream video on my phone, but I'd like to see it anyway.

Same here, I live in a large US metro area and I have yet to see the mmWave let alone use it. In fact, more often than not I lose reception within city travel. (I use Verizon 5G btw).

 

[deconstruct60]

Apple bought the business unit from Intel in 2019, when Intel's culture was in ruins. Everything was delayed at Intel at that time. No team was executing well. It seems like Apple hasn't fixed the culture. That's unfortunate.

Intel bought that modem business from Infineon which also failed to compete with Qualcomm well over time. ( Intel bought the modem business around the same time as Apple walked away from buying Infineon modems.

Intel buys modem business 2010

Intel buys Infineon unit and expands wireless offer

U.S. chipmaker Intel Corp unveiled a deal on Monday to buy German chipmaker Infineon Technologies AG's wireless unit for $1.4 billion, as it claws its way into the booming smartphone market and cuts its reliance on personal computers.

www.reuters.com

Apple start selling Verizon iPhones 2010 ( with Qualcomm )

It has been a track record of 'bad timing' for a long time with this subdivision. )

That old Infineon business had a long track record of ignoring CDMA tech (not 'phone system CDMA' , the basic tech) and doing narrower , EU focused implementations that they thought were better. As cell phones got more "world wide' tech based they struggled. Qualcomm had one modem chip that Apple could toss into every iPhone deployed around the world. That is a very substantive reason why Infineon 'lost' and largely lead to why the same base org when inside of Intel struggled with 5G. )

There was also a big 'consolidation' trend in smartphones to integrated modems. Qualcomm , Samsung , MediaTek all made that move. Infineon/Intel was still mainly in the discrete modem biz with no contract with the largest discrete modem buyer in smartphones.

Intel wrangled the modem contract back for the iPhone, not because Apple was unhappy with Qualcomm's technical meritis, but because Apple wanted to pay less.

Intel did screw some stuff on top of that other things going on. ( certainly didn't make it better). Pushing the modem folks to use Intel Fabs basically lead to some delays both when Intel owned the modem business and when Apple took over ( would now have to redo all designs (that survived) off Intel Fabs. )

Pretty good chance some of this is on Apple's management that likely bought off on throwing everything out the window and starting over. That isn't going to get them a timely replacement.

I blame WFH as well. Apple bought the unit in December 2019. Shortly after, everyone started working from home - wasting years of productivity and culture change.

It is also a culture change to go from "got consumers , need to ship and cross the i's and dot the t's " to some vague new customer years from now" with basically nothing ( but minor bug fixes) to do in-between.

When Apple was switching over from Arm/ImaginationTech designs to in-house stuff they did not put a long lull in between those in getting product out the door. They incrementally subsumed major portioins of the GPU from ImagationTech. The Arm replacement went from simpler to more complicated.

A large, 'Big Bang' replacement of Qualcomm would be a substantive culture shift also. And the standard to 'cover' is moving at a reasonable fast pace too. ( what Qualcomm is covering is expanding at a good rate. )

 

[1369281]

I agree with you though that right now it is clear that Apple's modem is not better than Qualcomm's modem. Integrated, power management etc are their goal. I don't know if they'll succeed, but I wouldn't count them out.

Remember that there was similar thinking about phones - "It's hard to imagine a scenario where Apple's phone is better than <Nokia, Blackberry etc>". Ditto MP3 players. They may not be first (e.g. iPods), and they may not get it right immediately (e.g. plenty of problems like AirPower, Apple Maps, Ping, MobileMe, Pippin, butterfly keyboards, Apple ///, Lisa), but they do keep trying and they often eventually get it right (e.g. Newton eventually got us to iPhones/iPads etc although it was a bumpy road). Sometimes they fail, but if you try 20 times, and have 10 failures, 9 things that are just okay and can be bundled into Apple One (e.g., I bet eventually SOS after 1 year free), and 1 thing that is a big success, you come out way ahead.

Haha, for what it’s worth, I never thought that about Nokia and Blackberry. I definitely understand what you are saying, though!

 

[NT1440]

They did part ways at some point... Qualcomm ended up doing fine even though it was a big hit, they sued Apple and later restored their alliance, which was the wisest decision for Qualcomm, and the best for the consumer since iPhones with other chips were experiencing poor connectivity issues.

Apple is a quarter of their modem business revenue. It’s going to be a big hit whenever it happens.

 

[NT1440]

Same here, I live in a large US metro area and I have yet to see the mmWave let alone use it. In fact, more often than not I lose reception within city travel. (I use Verizon 5G btw).

mmWave has ONE use case, stadiums.

Outside of that, even drywall blocks the signal so you’d need line of sight to use it at all.

Like everything else in America’s telecom marketing, it’s a scam.

 

[RickySky]

Are you joking? It's the other way around! If Qualcomm were to ever lose Apple as a customer, their share price would tank! A quarter of all revenue comes from 1 customer! That's not how you should run ANY business.

Apple have Qualcomm by the cahoonies on price and delivery.

The loss of Apple's business in 2024 was already factored into today's stock price, it's not like it was going to be a surprise. this news is just gravy for Qualcomm who is diversifying from phones and getting into connected cars (which Apple might have to go to for 5G modems for their car, LOL). Also no one seems to mention that Apple is paying ARM for use of their chips designs so it's not like Apple came up with their silicon on their own.

 

[NT1440]

The loss of Apple's business in 2024 was already factored into today's stock price, it's not like it was going to be a surprise. this news is just gravy for Qualcomm who is diversifying from phones and getting into connected cars (which Apple might have to go to for 5G modems for their car, LOL). Also no one seems to mention that Apple is paying ARM for use of their chips designs so it's not like Apple came up with their silicon on their own.

They’re not paying for ARM’s chips. They’re paying for the ability to conform to the ISA. Apple’s actual chips are their own creation and have been for years.

Competitors use ARM’s Cortex chip series (except some Samsungs), Apple does not. That’s an important distinction.

 

[jz0309]

Apple is a quarter of their modem business revenue. It’s going to be a big hit whenever it happens.

It is a financial loss for sure, but, QCOM has made quite some inroads into automotive and other areas to offset the Apple loss, they will do fine without that business.

 

[cocky jeremy]

Apple should pull itself out of business. Qualcomm is the market leader and so it will remain for the time being.
Apple brings 5G modem out:
Tim Cook:WOoOoW wHaT a Big leAd fOr AppLe!!!!!1
Meanwhile Qualcomm brings 6G out 😂😂

Because Apple couldn't possibly be working on 6G along side 5G, right?

 

[NT1440]

It is a financial loss for sure, but, QCOM has made quite some inroads into automotive and other areas to offset the Apple loss, they will do fine without that business.

Oh Qualcomm is a behemoth, modems are just one revenue stream so it’s definitely not doom and gloom for them like you’ve said.

 

[jz0309]

Modem and RF design are hard. I’d you’re good at CPU design, does not translate that you’re good at modem design.
The same holds true for a foundry, intels foundry was catered solely for CPU design, squeezing a modem in there, well, doesn’t work that easy.
Apple needs to catch up on what QC has done thus far, then for top of the line iPhones, it needs to be better than what QC offers at that point… that’s a long road

 

[xpxp2002]

mmWave has ONE use case, stadiums.

Outside of that, even drywall blocks the signal so you’d need line of sight to use it at all.

Like everything else in America’s telecom marketing, it’s a scam.

That's simply not true. I've used mmWave from a mast outside of a building before and achieved 3 Gbps downlink performance. Neither drywall nor outside wall materials completely attenuate mmWave, at least not n261. I've also used n260 inside a building through glass without a problem, maxing out one of AT&T's paltry 1 Gbps backhaul deployments.

mmWave can provide a meaningful capacity boost to central business districts, large parks, college campuses, and even museums and conference/exposition centers. That being said, mmWave is often deployed in dense environments like arenas (indoor or outdoor), stadiums, and other venues that frequently hold thousands of people within a confined area for good reason, too. Today, and over the past decade, carriers have to deploy lower mid-band to provide an LTE PCC off of their large venue DASes, usually B2 and/or B66. Those bands, even at low power, still propagate and reflect more than engineers would like in open air in those types of environments. This lowers the SNR considerably, especially as the number of UE rises into the tens of thousands, and makes controlling the noise floor very difficult. The poorer SNR further diminishes the per-device and overall performance, as well as the channel reuse opportunities in these venues. Moreover, carriers who don't have strong mid-band holdings suffer more challenges finding enough capacity to redeploy across a venue and might even have to resort to a low-band underlay.

mmWave solves all of those problems. The bands designated for mmWave are high enough in frequency to propagate within open air and through one or two walls or physical barriers without significantly raising the noise floor of adjacent rooms or areas. The carriers can be provisioned and aggregated widely enough to provide plentiful capacity. And the higher numerology/subcarrier spacing allows for more performance to be extracted from the carrier. mmWave-only SA support, higher orders of modulation, along with MIMO improvements for mmWave due to come in the near future will enable existing defined mmWave bands to deliver even more capacity than the common 2x2 MIMO 8x100MHz carriers pushing ~4 Gbps aggregate that you've probably seen or heard about today.

I won't pretend that mmWave is a good solution for every capacity situation. But the increasing demands on the limited sub-6 spectrum that's available for cellular use are not sustainable; and the spectrum that's allocated isn't ideal for confined, dense environments. mmWave will be necessary as time goes on and providers need a way to deliver consistent usable service to more devices within a defined space. Some areas will likely never see mmWave deployed, but the likelihood of more city CBDs, performance and conference venues, and dense residential and commercial environments getting a mmWave capacity layer added will unquestionably increase over time as demand continues to necessitate it.

 

[jz0309]

mmWave has ONE use case, stadiums.

Outside of that, even drywall blocks the signal so you’d need line of sight to use it at all.

Like everything else in America’s telecom marketing, it’s a scam.

I would certainly agree that it is overrated.
It is based on the hype that you can download a 4K movie in seconds, and while I can see some niche cases where that is helpful, the average consumer doesn’t need it…

 

[NT1440]

That's simply not true. I've used mmWave from a mast outside of a building before and achieved 3 Gbps downlink performance. Neither drywall nor outside wall materials completely attenuate mmWave, at least not n261. I've also used n260 inside a building through glass without a problem, maxing out one of AT&T's paltry 1 Gbps backhaul deployments.

mmWave can provide a meaningful capacity boost to central business districts, large parks, college campuses, and even museums and conference/exposition centers. That being said, mmWave is often deployed in dense environments like arenas (indoor or outdoor), stadiums, and other venues that frequently hold thousands of people within a confined area for good reason, too. Today, and over the past decade, carriers have to deploy lower mid-band to provide an LTE PCC off of their large venue DASes, usually B2 and/or B66. Those bands, even at low power, still propagate and reflect more than engineers would like in open air in those types of environments. This lowers the SNR considerably, especially as the number of UE rises into the tens of thousands, and makes controlling the noise floor very difficult. The poorer SNR further diminishes the per-device and overall performance, as well as the channel reuse opportunities in these venues. Moreover, carriers who don't have strong mid-band holdings suffer more challenges finding enough capacity to redeploy across a venue and might even have to resort to a low-band underlay.

mmWave solves all of those problems. The bands designated for mmWave are high enough in frequency to propagate within open air and through one or two walls or physical barriers without significantly raising the noise floor of adjacent rooms or areas. The carriers can be provisioned and aggregated widely enough to provide plentiful capacity. And the higher numerology/subcarrier spacing allows for more performance to be extracted from the carrier. mmWave-only SA support, higher orders of modulation, along with MIMO improvements for mmWave due to come in the near future will enable existing defined mmWave bands to deliver even more capacity than the common 2x2 MIMO 8x100MHz carriers pushing ~4 Gbps aggregate that you've probably seen or heard about today.

I won't pretend that mmWave is a good solution for every capacity situation. But the increasing demands on the limited sub-6 spectrum that's available for cellular use are not sustainable; and the spectrum that's allocated isn't ideal for confined, dense environments. mmWave will be necessary as time goes on and providers need a way to deliver consistent usable service to more devices within a defined space. Some areas will likely never see mmWave deployed, but the likelihood of more city CBDs, performance and conference venues, and dense residential and commercial environments getting a mmWave capacity layer added will unquestionably increase over time as demand continues to necessitate it.

If properly deployed in a city, it’s useful. The way the marketing makes it appear is a different story entirely for the vast majority of the population.

 

[mfram]

mmWave has ONE use case, stadiums.

Outside of that, even drywall blocks the signal so you’d need line of sight to use it at all.

But when you're there, it works great. I pulled 1447 Mbps down and 77.5 Mbps up at Petco Park in San Diego with my iPhone 14 Pro!

 

[laptech]

Correct me I am wrong but wasn't there a iphone model that used Intel modems but there was such a huge consumer backlash because the modem signals kept dropping out or was just not reliable enough so Apple went back to using Qualcomm modems.

 

[Shirasaki]

I'm excited to see what apple can produce. I wish Apple would just start their own cell carrier service.

Definitely can only take off in US. Hell No way it can expand to anywhere else, unless Apple is MVNO.

Apple bought the business unit from Intel in 2019, when Intel's culture was in ruins. Everything was delayed at Intel at that time. No team was executing well. It seems like Apple hasn't fixed the culture. That's unfortunate.

I blame WFH as well. Apple bought the unit in December 2019. Shortly after, everyone started working from home - wasting years of productivity and culture change.

Oh yeah let’s beat That horse that has never been tired of winning and we human scramble to defend while it run rampant From 2020 to 2022. Or are you going to be the one responsible of losing unspecified number of key engineers because you hate WFH?

So, you're blaming something of which you have no knowledge and for which you have no reason to put that blame on?

LMAO.

 

[NT1440]

But when you're there, it works great. I pulled 1447 Mbps down and 77.5 Mbps up at Petco Park in San Diego with my iPhone 14 Pro!

Oh, definitely when it works it works great. The idea that it’s a major selling point. In the advertisements is kind of ridiculous though because it’s only in major cities running those ads in Iowa in my mind is tantamount to fraud. is full of **** to begin with with deceptive advertising.