Apple's 'failed' 5G modem effort means iPhone 15 will be all-Qualcomm

AppleInsider

Apple's project to create a 5G modem of its own design may not be going how it wants, with Qualcomm now thought to be the sole modem supplier for the iPhone 15 launch in 2023.

Apple has been working on creating its own modem for use in its hardware for a number of years, in order to get more control over how the component functions and to reduce costs. However, while progress has been made, it is claimed that the results aren't enough for a modem introduction in the "iPhone 15."

In tweets on Tuesday, TF Securities analyst Ming-Chi Kuo said a supply chain survey indicates Apple's 5G modem development "may have failed," resulting in Qualcomm remaining as the "exclusive supplier for 5G chips of 2H23 new iPhones."

(1/4)
[Company Update] Qualcomm (QCOM.O)

My latest survey indicates that Apple's own iPhone 5G modem chip development may have failed, so Qualcomm will remain exclusive supplier for 5G chips of 2H23 new iPhones, with a 100% supply share (vs. company's previous estimate of 20%).

-- (Ming-Chi Kuo) (@mingchikuo)

June 28, 2022

While previous estimates from the company itself had Qualcomm supplying approximately 20% of modems for the 2023 models, the new estimate has Qualcomm having a "100% supply share."

Kuo still believes Apple will continue to develop its 5G modem chips, but doesn't think it will impact Qualcomm's future revenue in a massive way. "By the time Apple succeeds and can replace Qualcomm, Qualcomm's other new businesses should have grown enough to significantly offset the negative impacts caused by the order loss of iPhone 5G chips," the analyst writes.

Apple's acquisition of Intel's smartphone modem business, as well as the onboarding of some 2,200 engineers in 2019, has been a high expense for the company, with a significant potential payoff. If Kuo's tweet is true, then Apple has a little while more to wait before reaping the rewards.

Those benefits can include cost-savings, a reduction in reliance on suppliers like Qualcomm, and the ability to fine-tune the modem itself. This last point can include optimizing the modem for specific products, changing how it works in an iPhone compared to a less voice-dependent device like the iPad, for example.

Previous expectations had Apple coming up with an initial modem design in preparation for use in 2023, with long-time chip partner TSMC expected to be the producer of the component.

Read on AppleInsider

danox

Apple took 13 years to replace Intel’s crappy cpu, replacing Qualcomm, Google maps, game engines or Gpu’s all are long term necessary infrastructure projects for a company like Apple, how does one get to the so-called glorious tech future of AI/VR with all those type of things not being under one roof. Very long Apple.

It is obvious by the designs of Apple’s SOC they want to put a non Qualcomm modem inside the SOC somewhere down the line. 

wood1208

How much truth to this story ? Survey don't reveal truth to Apple's 5G modem internal development status.

spock1234

wood1208 said:

How much truth to this story ? Survey don't reveal truth to Apple's 5G modem internal development status.

Someone's trying to boost Qualcomm's share price. 

y2an

This is a very chip-in-slot centric discussion whereas Apple’s direction is all SoC, so I might think any delay here would be associated with getting a modem functional block into the SoC itself. How far they might be trying to take this would be an interesting question, ideally leaving only the analog portions external to the SoC (mixed digital/analog on the same chip being pretty hard). 

danox

y2an said:

This is a very chip-in-slot centric discussion whereas Apple’s direction is all SoC, so I might think any delay here would be associated with getting a modem functional block into the SoC itself. How far they might be trying to take this would be an interesting question, ideally leaving only the analog portions external to the SoC (mixed digital/analog on the same chip being pretty hard). 

A long term tech infrastructure project.

bluefire1

There is nothing like a QCOM modem. Apple has a high bar to overcome to develop  anything comparable to what QCOM has already achieved.

igorsky

Typical Kuo story: unconfirmed delay to an unconfirmed timeline.

rezwits

danox said:

Apple took 13 years to replace Intel’s crappy cpu, replacing Qualcomm, Google maps, game engines or Gpu’s all are long term necessary infrastructure projects for a company like Apple, how does one get to the so-called glorious tech future of AI/VR with all those type of things not being under one roof. Very long Apple.

It is obvious by the designs of Apple’s SOC they want to put a non Qualcomm modem inside the SOC somewhere down the line. 

LONGER.

But my sentiments exactly.  The modem game started at least 40+ years ago.  For Apple to come up with a brand new (small/low power/low heat and not to mention 5G or "6G" [next] wave of modem, is not gonna happen over night.  I mean we knew this right?

blastdoor

If true this is a huge win for Qualcomm, small loss for apple. 

thadec

blastdoor said:

If true this is a huge win for Qualcomm, small loss for apple. 

Huh? Explain. Qualcomm's discrete modem sales to Apple is a fraction of their overall business. Wasn't Apple's agreement with Qualcomm $4.5 billion over 6 years? Less than $1 billion a year for a company that had $36 billion in revenue in 2021. This is even sillier than the people who claim that losing Apple's business will cripple Intel when A. Apple is only the #4 PC maker behind Lenovo, Dell and HP and in 2020 was actually #5 behind Acer and not much far ahead of ASUS and B. the clear majority of CPUs Intel sold to Apple were the cheap Core i3 ones for the entry level MacBook Air and Mac Mini.

9secondkox2

Last I remember, there were some patent issues after Qualcomm bought a bunch. 

Apple will succeed in launching a new, superior modem and it will be a gigantic hit to Qualcomm revenues. Apple will use their own modems in all iPhone, iPads, apple watches, AR devices and posssibly even Macs. 

The iPhone is a HUGE market, iPad is significant, Mac is huge, and Apple Watch is pretty big as well. 

That’s a lot of revenue to lose and Qualcomm will wish they didn’t try to gouge apple a few years ago. 

Short term thinking sometimes wins out in the short term, not so much in the long. 

I doubt that other, silicon capable OEMs like Samsung and future Microsoft don’t have similar plans. 

9secondkox2

rezwits said:

danox said:

Apple took 13 years to replace Intel’s crappy cpu, replacing Qualcomm, Google maps, game engines or Gpu’s all are long term necessary infrastructure projects for a company like Apple, how does one get to the so-called glorious tech future of AI/VR with all those type of things not being under one roof. Very long Apple.

It is obvious by the designs of Apple’s SOC they want to put a non Qualcomm modem inside the SOC somewhere down the line. 

LONGER.

But my sentiments exactly.  The modem game started at least 40+ years ago.  For Apple to come up with a brand new (small/low power/low heat and not to mention 5G or "6G" [next] wave of modem, is not gonna happen over night.  I mean we knew this right?

Not really. The current and past modem game is a known quantity. Apple has been part of development of all kinds of new computing standards over the years including wireless transmission standards. 

When 5g hit, it was a game changer and was a new game. Apple can create something compatible with how things are now while reinventing the future game as well. People underestimate apple. They were “crazy” to take on the music industry at a time when physical media was at its peak and online music was pirate city. They were “insane” to create a new phone in the crowded phone market full of unstoppable competitors. They were “crazy” to enter the SOC game. Next thing, they are winning that game and not only transforming the industry, but spurring competitors like Intel to copy their innovations. 

Apple never enters a market it hasn’t analyzed to death and never enters without strong confidence in coming out on top. 

avon b7

9secondkox2 said:

Last I remember, there were some patent issues after Qualcomm bought a bunch. 

Apple will succeed in launching a new, superior modem and it will be a gigantic hit to Qualcomm revenues. Apple will use their own modems in all iPhone, iPads, apple watches, AR devices and posssibly even Macs. 

The iPhone is a HUGE market, iPad is significant, Mac is huge, and Apple Watch is pretty big as well. 

That’s a lot of revenue to lose and Qualcomm will wish they didn’t try to gouge apple a few years ago. 

Short term thinking sometimes wins out in the short term, not so much in the long. 

I doubt that other, silicon capable OEMs like Samsung and future Microsoft don’t have similar plans. 

I don't think it will be a huge hit and QC will get its patent dues all the same. 

All QC will really 'lose' is a business it never really had anyway (while they and Apple were dishing it out in court) as Apple was sourcing from Intel. 

Also, as 5G enters the IoT realm, QC will sweep a lot of that business up along with Huawei, Mediatek, Samsung... 

Apple is unlikely to shift anything like the numbers its competitors will as I doubt they will offer up their product to third parties. 

I also doubt it's first generation product will be at the same level as the competition. 

Japhey

igorsky said:

Typical Kuo story: unconfirmed delay to an unconfirmed timeline.

Exactly. What better way to generate clicks than to put “Apple” & “fail” in the same sentence? It’s straight out of the Forbes playbook. And what better way to claim said failure than to help create the very rumor of a fictional deadline that wasn’t ever meant to be met? 

My guess is that Apple found, and plugged, the leakers that Kuo relied upon as his sources. This new side hustle on Twitter is just his desperate attempt to remain relevant. 

mattinoz

y2an said:

This is a very chip-in-slot centric discussion whereas Apple’s direction is all SoC, so I might think any delay here would be associated with getting a modem functional block into the SoC itself. How far they might be trying to take this would be an interesting question, ideally leaving only the analog portions external to the SoC (mixed digital/analog on the same chip being pretty hard). 

Do they need a special block for that?
Once you take analogue out it seems a lot of the signal processing could happen in GPU. If they can make that work an Ax could be just the Mx Soc with the extra cores dedicated to being soft modem. 

y2an

mattinoz said:

Do they need a special block for that?
Once you take analogue out it seems a lot of the signal processing could happen in GPU. If they can make that work an Ax could be just the Mx Soc with the extra cores dedicated to being soft modem. 

If you’ve nothing better for the GPU to do, probably. But they need to exact top performance from the GPU to be competitive at the high end, support gaming, etc., so stealing cycles from the GPU would not make sense. Plus a lot if this is not signal processing, it’s protocol handling. It’s a highly specialised real-time activity.

onlyhope

This is the kind of crapy "reporting" that has taken over AI...just click-bait.  Sad. I'm done. 

9secondkox2

avon b7 said:

9secondkox2 said:

Last I remember, there were some patent issues after Qualcomm bought a bunch. 

Apple will succeed in launching a new, superior modem and it will be a gigantic hit to Qualcomm revenues. Apple will use their own modems in all iPhone, iPads, apple watches, AR devices and posssibly even Macs. 

The iPhone is a HUGE market, iPad is significant, Mac is huge, and Apple Watch is pretty big as well. 

That’s a lot of revenue to lose and Qualcomm will wish they didn’t try to gouge apple a few years ago. 

Short term thinking sometimes wins out in the short term, not so much in the long. 

I doubt that other, silicon capable OEMs like Samsung and future Microsoft don’t have similar plans. 

I don't think it will be a huge hit and QC will get its patent dues all the same. 

All QC will really 'lose' is a business it never really had anyway (while they and Apple were dishing it out in court) as Apple was sourcing from Intel. 

Also, as 5G enters the IoT realm, QC will sweep a lot of that business up along with Huawei, Mediatek, Samsung... 

Apple is unlikely to shift anything like the numbers its competitors will as I doubt they will offer up their product to third parties. 

I also doubt it's first generation product will be at the same level as the competition. 

Considering they have apples business now, they definitely stand to lose big. 

Apple may be looking to future standards as well - something that Qualcomm may need to pay royalties on. 5g has been going for a bit now. It won’t last forever. 

avon b7

9secondkox2 said:

avon b7 said:

9secondkox2 said:

Last I remember, there were some patent issues after Qualcomm bought a bunch. 

Apple will succeed in launching a new, superior modem and it will be a gigantic hit to Qualcomm revenues. Apple will use their own modems in all iPhone, iPads, apple watches, AR devices and posssibly even Macs. 

The iPhone is a HUGE market, iPad is significant, Mac is huge, and Apple Watch is pretty big as well. 

That’s a lot of revenue to lose and Qualcomm will wish they didn’t try to gouge apple a few years ago. 

Short term thinking sometimes wins out in the short term, not so much in the long. 

I doubt that other, silicon capable OEMs like Samsung and future Microsoft don’t have similar plans. 

I don't think it will be a huge hit and QC will get its patent dues all the same. 

All QC will really 'lose' is a business it never really had anyway (while they and Apple were dishing it out in court) as Apple was sourcing from Intel. 

Also, as 5G enters the IoT realm, QC will sweep a lot of that business up along with Huawei, Mediatek, Samsung... 

Apple is unlikely to shift anything like the numbers its competitors will as I doubt they will offer up their product to third parties. 

I also doubt it's first generation product will be at the same level as the competition. 

Considering they have apples business now, they definitely stand to lose big. 

Apple may be looking to future standards as well - something that Qualcomm may need to pay royalties on. 5g has been going for a bit now. It won’t last forever. 

They won't lose big because they never counted on having that business. 

Apple's 5G road map was with Intel. 

QC simply saw a business opportunity and negotiated that extra contract. Things will simply fall back to where they were prior to Intel failing. If Apple does not deliver in 2023, QC may even retain some of its Apple contracting. 

Either way, Apple's business with QC is simply a plus for them. Losing Apple will not knock a huge hole into their business. 

Apple may be looking to future standards and buying Intel's modem business bought them a lot of 5G patents but 6G won't be here until around 2030 and 5.5G will come before that. 

6G development is already underway but consumer facing modems are just a tiny part of ICT technology. Apple has no experience in the field so expect the established players to pull most of the strings on that front. 

That isn't to say it would be a bad idea to break into that market but look how much they screwed up on 5G. Something that started taking shape in 2009. Apple just isn't the kind of company that plans that far ahead and on such a wide scale. 

It's a CE company with zero experience in critical infrastructure. 

loopless

Everything analysts say is designed to manipulate stock prices for their clients benefit.