Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Apple will gimp their own products just to stop using Qualcomm.

When Apple were using Intel chips in previous iPhones you could notice performance was a lot slower. Every test reviewers ran Intel came out second best.

I noticed an immediate performance boost when they started using Qualcomm modems again.
 
The benefits are mostly in a SoC with radio?. Mmmhhh. Anyone already provides that?, is it efficient?.
 
I doubt it. I think when Apple released an M1 chip we didn’t see any type of price reduction. Unfortunately, the price of the iPhone will only go up :(. Also, we are years away from the first phase of the transition.
Because Apple became greedy. Costs should have fallen dramatically shifting to Apple Silicon. The foundation of the CPU’s were already there, they didn’t design the chip from the ground up. It’s a ARM chip that Apple tweaked a bit themselves.

Instead we have spiralling price increases with no justification. They blamed inflation for the most recent price rises and the rising dollar in relation to the pound. The pound has now recovered slightly but I don’t see any price reductions in sight in the UK.
 
Why does everyone seem to think these will suck? Apple Silicon was a triumph right out of the gate.

Apple has a development facility here in San Diego and I assume they chose this location specifically to poach Qualcomm's best and brightest.
 
Qualcomm is really holding everyone by the neck, seriously there is no other competitor globally?
 
Why does everyone seem to think these will suck? Apple Silicon was a triumph right out of the gate.

Apple has a development facility here in San Diego and I assume they chose this location specifically to poach Qualcomm's best and brightest.
Good thinking!
 
If modems were easy to make, there would be much more competition than there is.

Modems are a high volume low margin business, which is why nobody cares about them and there's no competition. Plus patents on all the wireless standards mean that it's literally impossible for a new entrant to appear, because they'd get sued out of existence.

That's why Apple can do it: they have enough patents in their arsenal that they can play the cross-licensing game.
 
  • Like
Reactions: Madd the Sane
That's what I was thinking as well, unless the Intel patents that Apple purchased can skirt around Qualcomm. Making your own proprietary radio chips seems a lot trickier (legally) than making your own SoCs.
Thats also partially why the Intel LTE modems sucked, because Qualcomm already had the patents on the better technology.

I smell this is gonna be an expensive failure for Apple.
 
AGAIN?! LOL can’t wait to see the cluster —— that the alternative will offer. So far, outside of Asia, no one has been able to really release themselves from Qualcomm’s grip.
If there is one company that could do it, that would be Apple. And would it really be a good thing if Qualcomm continued to be the dominate supplier for these parts? I’m sure Apple looked down the road and saw that this could become a problem in the future. Given the disagreements the two companies had in the past, this isn’t a long term relationship.
 
  • Like
Reactions: Madd the Sane
I actually think they have a lot of opportunity with this. The license fees from Qualcomm are sooo high. Everyone wishes they could get out from under it. You have to pay a percentage of the entire device price not a simple component unit price. There was a recent poll that came out saying most users don't understand the need for 5G, the speeds for 4G was fine for everyday smart phone use. I think while 5G is looking for a killer app and a reason for existing that even if Apple's modem isn't as fast as Qualcomm most regular users won't care.

That's because nobody understands 5G. Currently with 4G we are seeing an increasing problem with congestion in urban areas and over the coming years it'll get worse. 5G will help mitigate this and keep average speeds up, never mind max speeds advertised for marketing purposes. Especially the sub-6 bands will be useful for that, mmWave only really in smaller areas.
 
  • Like
Reactions: xpxp2002
One of the rumors has been that the project hasn’t “failed”, which is a pretty strong term. If it failed, Apple would abandon it. The fact that they haven’t shows that while it’s delayed, it hasn’t failed. What I've heard us that Apple wants to bypass having just a 4G, 5G modem, and to lead with a 6G model. that would take extra time.
 
  • Haha
Reactions: JPack
An Apple Modem may be designed around Globalstar's Band 53/N53 spectrum. This spectrum is at 2.4Ghz. As Band 53/N53 designated by 3GPP it can be used in TD-LTE networks which is what the original FCC ATC authority allowed. The location of this spectrum is right in both Wifi and Bluetooth range. Globalstar's original request was to be granted terrestrial use of this spectrum in WiFi networks. The original testing in 2014 was done successfully using iPhone and Android phones using WiFi. The biggest push back came from the Bluetooth community backed by Microsoft due to the proximity of Globalstar's 2.4Ghz spectrum and Bluetooth use. In the end the FCC granted use in TD-LTE networks which is how Qaulcomms X65 modem will use Band 53/N53. The Apple SOS feature is not using Band 53/N53. The SOS feature is using Globalstar's 1.6Ghz spectrum as uplink to the satellites from the phone.
 
  • Wow
Reactions: gusmula


Apple wants to start replacing the Qualcomm modem chips that it uses in iPhones as soon as 2024, according to a new report from Bloomberg. Apple has been working on in-house modem chip technology for several years now, with the aim of reducing its reliance on Qualcomm.

Apple-5G-Modem-Feature-Triad.jpg

Apple initially wanted to introduce its own modem chips as early as 2023, but Apple analyst Ming-Chi Kuo said in late 2022 that Apple would need to continue to rely on Qualcomm in the near future as Apple's development efforts "failed." At the time, Kuo said that Apple would continue working on its 5G chips, but that development would not be done in time for the launch of the 2023 iPhones, which Bloomberg's report agrees with. Modem chip development has seen delays, and Apple will use a slow rollout to end its reliance on Qualcomm. Apple will start with using its own modem chip in a single device before expanding the rollout to other devices. The transition away from Qualcomm could take up to three years.

In addition to transitioning to its own in-house chips starting in 2024, Apple wants to stop using wireless components from Broadcom in 2025. Apple is working on a WiFi and Bluetooth chip that would replace the components that it is currently sourcing from Broadcom. Apple in 2020 signed a three and a half-year deal with Broadcom for wireless components and modules, with that deal set to expire midway through 2023.

While separate components to replace hardware sourced from Qualcomm and Broadcom are in development right now, Bloomberg says Apple is also working on a chip that will combine the cellular modem, WiFi, and Bluetooth functions into a single component. Broadcom also supplies Apple with radio-frequency chips and chips for wireless charging, both of which Apple is working on replacing.

Apple designs its own A-series chips for the iPhone and M-series chips for the Mac, and replacing modem chips and wireless components has been a priority ever since its Qualcomm dispute. Apple launched a legal battle against Qualcomm in 2017, accusing Qualcomm of unfairly collecting royalties for technologies it had nothing to do with. Apple wanted to transition away from Qualcomm with the shift to 5G to stop paying Qualcomm's fees and instead use Intel technology, but Intel was not able to manufacture 5G chips that met Apple's standards.

Apple was forced to settle its lawsuit with Qualcomm and has since used Qualcomm 5G modem chips for the iPhone and iPad lineups. Apple in 2019 purchased Intel's modem chip business to get a head start on chip development, and 2024 may be the soonest that the technology is finally advanced enough for Apple to phase out Qualcomm.

Article Link: Apple to Start Transitioning Away From Qualcomm Modem Chips in Late 2024 or 2025
I think Apple is trying to run into being a chip manufacturer too fast. Apple and its users don't understand making chips isn't the same as software/OS R&D and chip design. Making and testing chips is a long process and planning slow and long term.

I wonder if Apple's long term goal is more AI chip and support software than devices.
 
I doubt it. I think when Apple released an M1 chip we didn’t see any type of price reduction. Unfortunately, the price of the iPhone will only go up :(. Also, we are years away from the first phase of the transition.
Considering inflation, we've gotten significant price decreases for the last two years. But even leaving that aside:

2020 MacBook Air i7 8GB/256GB original price: $1249
2020 MacBook Air M1 8/7 8GB 256GB original price: $999

2020 MacBook Pro 13" 2Tbolt i7 1.7 8/256 price: $1599
2020 MacBook Pro 13" 2Tbolt M1 8/8 8/256 price: $1299

And yes, I know there were lower spec models using Intel that were the same price as the M1s, but they were not even remotely comparable. In fact, the Intels I compared are an embarrassment compared to the M1s, but that's the closest I could get.

A similar comparison could be made for the 14/16" models.

Nobody's going to call Macs inexpensive, but on the low-end and mid-range, the value proposition has improved dramatically since the 2020 transition.

Lastly, everyone was grousing about how prices would rise for the iPhone 14 Pro. And yet... they did not.
 
The majority of the modem chip area is dedicated to analog circuitry which doesn't depend on small transistors to reduce power and area. Some of these are still using 65nm tech, etc. But there is a portion of the chip that does use digital circuitry and this could theoretically be integrated into their SOC on n4/n3/etc. But then that removes the flexibility of a single modem working for various different SOCs. You would have to redo new versions of all the chips to support it. So we will see what Apple goes with. The SOC does already have the control logic built-in to handle the connections to and from the modems. And most likely since the portion of digital logic that could be integrated is small it wouldn't pay for itself in power savings/performance/die area cost

Your initial facts are correct but I think your argument is wrong. There is no loss of flexibility.

I fully expect that as Apple takes more I/O in-house, they will integrate the MACs into their Mx (and Ax) chips, building separate PHYs. Or, possibly, separate chiplets, though that's really just a question of engineering->economics (is it worth it; my guess is, not in the next year or two, but I'm not expert in this).

This is actually very common in the industry. For example, Intel "chipsets" (really, just a single chip these days) have an integrated MAC for wifi, but you still need a separate PHY.

It's probably no more actual work to take an already-designed MAC and integrate it into a new chip than it is to add some generic interface (presumably PCIe) to that chip that you'd use to link it to a separate chip containing MAC and PHY. I'm guessing that the total transistor and energy budget is also better with the integrated MAC (due to a generic I/O interface never being as optimal, need for more clock conversions, etc.) but I can't say for sure. If name99 sees this he can probably answer it.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.