I wish folks would test these against current gen Qualcomm modems. We're comparing a brand new modem to a 2 1/2 year old Modem. Android uses x80 and Qualcomm just released x85. So a fair comparison would be C1 vs x80.
I understand that is what the iPhone 16 uses, but that's an Apple problem they purposefully took an older generation modem from Qualcomm for their 16 series. Knowing Apple's modus operandi they probably decided to use the x71 so when they released C1 it appeared better than it really is. They did this during the Intel Modem days when they shipped that.
None the less it will be interesting to see how this modem develops further and what products it ends up in.
No, but Apple always uses older generation parts from their suppliers (modems, displays, glass, NAND, etc), not to sabotage their products, but to cut down on BOM costs and increase margins, as suppliers always ask a premium for their latest and greatest, especially since their customers never cared about the spec sheet so using the most expensive parts would be profits wasted.
I think there is another dimension there - volume. Apple volume is on another level and it takes time to scale up factories to build latest and greatest things.
I'm curious, is it the power reduction claim by Apple that is impressive, and where exactly does it come from?
Because as I understand it, transmitting a signal is the main power drain. If you transmit with less power it would mean a lesser range and lower signal quality. So Apple can't just have their phone transmit a signal with less power...
The other bigger reason why they are doing this is that it removes having to deal with licensing and using any Qualcomm provided firmware & hardware designs. It doesn't quite free them from having to license any patents probably. But other than that, they no longer have to deal with anything Qualcomm on the critical path to shipping hardware & software.
Depending on what their patent licensing with Qualcomm looks like, this also might free them up to embed 5G capabilities in just about any other device they have that isn't a phone. Combined with soft sims, there's no need for sim card slots even. Think laptops, vr goggles, apple tv's, watches, etc. That stuff is/was mostly blocked on licensing and Qualcomm taking a cut; and only partially on power requirements. I wouldn't be surprised to see the C1 in some future mac books. In fact I'd be surprised if that didn't happen.
Some Chinese phone companies like Huawei have been doing similar things and use their own chips rather than Qualcomm provided ones and of course the whole notion of Qualcomm enforcing their patents in China is not much of a thing.
It will be interesting to see what happens with 6G and beyond. That will be the first standard where Qualcomm has to compete with multiple competitors competing on quality and features. Apple won't wait for them to do whatever. And likewise, companies like Ericsson/Nokia might start talking directly to Apple about making sure their network infrastructure works great with IOS.
> Modems are speed limited by networks which evolve on a MUCH slower timescale
While true, at some point networks do catch up, and many people hang on to their phones for several years at this point. If that's the case, whether having a better modem makes a difference in the end depends on which dimension of "faster" you're looking at:
Just using more and more Hz of spectrum on a single device does indeed not help anything other than peak speeds in empty cells (something I personally don't care about in the slightest in a phone), but being able to deliver more bit/Hz/s (through spatial multiplexing, better beamforming etc.) does eventually benefit everyone in congested/spectrum-limited situations.
>High signal: 0.88 watts for the iPhone 16, 0.67 watts for the iPhone 16e
>Low signal: 0.81 watts for the iPhone 16, 0.67 watts for the iPhone 16e
For the legacy modem, why is power consumption lower in lower signal areas? I naively expected more power use, either searching for signal or amplifying more. Is the power consumption mostly a factor of data throughput?
Apple was able to get 18% power savings on their modem (a quick average).
If Qualcomm cared they could do the same thing. That's gigawatts of energy, over billions of cellphones, being wasted by modems because the vendor doesn't care.
The problem, of course, is the cost of that inefficiency is external to the vendor.
Maybe Qualcomm will improve their power performance in the next-next-generation.
> The problem, of course, is the cost of that inefficiency is external to the vendor.
How so? The impact of power efficiency is arguably as immediate to a mobile phone manufacturer as it gets. These things are battery powered, and many people are struggling to make it through a single day on one charge!
Of course, that's practically exactly what battery size is tuned to – making it through exactly one day for most users. Knowing Apple, they'll just make the next flagship half a millimeter thinner again rather than risking one and a half days of battery life...
I suspect they have some numbers on how much the ipad cellular is sold and how often that feature is used and they may be using that to determine whether it's worth it.
But now that they develop the chip in house, I don't see why they wouldn't include it. In fact, given they put an Apple Watch chip and extra display in their macbooks for a few years, I don't see why they haven't done it years ago.
edit: I googled, apparently Qualcomm's licensing fees are a percentage of the whole device the chipset is installed in. I suppose it's less of a factor now that some iPhones are more expensive than Macbooks though.
Yeah for a while I felt that a valid counter-argument was "modems are iterating faster than laptop cycles so you don't want to be stuck with an older modem" – but with 5G modems and all the architecture changing over to IP with LTE years ago, now you can probably have a C1 in a laptop for the next 5 years and be perfectly content without cell phone networks phasing those out.
The issue for LTE and 5G is/was less tech evolution as it was just new bands coming online. At least in the US, you really did need a pretty recent phone to get decent performance on TMobile or even more recently C-Band 5G
Is there really a need to carry multiple cellular modems? We already have one in our pocket most of the time and can easily share it in a single click. Feels like the vast majority of use cases are already covered by that solution and it doesn't require a second expensive data plan.
I'm curious if they tested this in highly-congested cells, since that's when cellular modems have to do the most "magic" and probably hit weird implementation edge cases. I think using the C1 in a packed football stadium is the real test of whether Apple has caught up to Qualcomm.
Other than mmwave (which this modem does not support), I don't think there's much more magic to do in an empty vs. a congested cell for a baseband.
5G spectrum is managed by the network, so the only thing that will happen is that each device will get shorter/narrower time/frequency slices, but there shouldn't be any unexpected cross- or same-channel interference of the type that is common for unmanaged protocols like 802.11 or Bluetooth.
I wish folks would test these against current gen Qualcomm modems. We're comparing a brand new modem to a 2 1/2 year old Modem. Android uses x80 and Qualcomm just released x85. So a fair comparison would be C1 vs x80.
I understand that is what the iPhone 16 uses, but that's an Apple problem they purposefully took an older generation modem from Qualcomm for their 16 series. Knowing Apple's modus operandi they probably decided to use the x71 so when they released C1 it appeared better than it really is. They did this during the Intel Modem days when they shipped that.
None the less it will be interesting to see how this modem develops further and what products it ends up in.
You think Apple deliberately sabotaged their flagship model so they could make their cheap phone look better?
They've done it before: https://hothardware.com/news/apple-throttles-iphone-7-models...
If that's even true, that's still not even close to the same thing.
No, but Apple always uses older generation parts from their suppliers (modems, displays, glass, NAND, etc), not to sabotage their products, but to cut down on BOM costs and increase margins, as suppliers always ask a premium for their latest and greatest, especially since their customers never cared about the spec sheet so using the most expensive parts would be profits wasted.
Sure. Using older, cheaper parts that get the job done is just sensible engineering. That’s not at all the claim I was replying to.
I think there is another dimension there - volume. Apple volume is on another level and it takes time to scale up factories to build latest and greatest things.
>Apple volume is on another level and it takes time to scale up factories to build latest and greatest things.
Bad take. Samsung sells as many phones as Apple while using the latest and greatest parts on the flagships.
when you operate at the scale of apple a defect can become very costly. using well tested, older tech lowers this counterparty risk.
I'm curious, is it the power reduction claim by Apple that is impressive, and where exactly does it come from?
Because as I understand it, transmitting a signal is the main power drain. If you transmit with less power it would mean a lesser range and lower signal quality. So Apple can't just have their phone transmit a signal with less power...
And this alone is why they are worth it for Apple. Modems are speed limited by networks which evolve on a MUCH slower timescale
Qualcomm has to sell faster and faster modems but apple can optimize for what we actually care about
The other bigger reason why they are doing this is that it removes having to deal with licensing and using any Qualcomm provided firmware & hardware designs. It doesn't quite free them from having to license any patents probably. But other than that, they no longer have to deal with anything Qualcomm on the critical path to shipping hardware & software.
Depending on what their patent licensing with Qualcomm looks like, this also might free them up to embed 5G capabilities in just about any other device they have that isn't a phone. Combined with soft sims, there's no need for sim card slots even. Think laptops, vr goggles, apple tv's, watches, etc. That stuff is/was mostly blocked on licensing and Qualcomm taking a cut; and only partially on power requirements. I wouldn't be surprised to see the C1 in some future mac books. In fact I'd be surprised if that didn't happen.
Some Chinese phone companies like Huawei have been doing similar things and use their own chips rather than Qualcomm provided ones and of course the whole notion of Qualcomm enforcing their patents in China is not much of a thing.
It will be interesting to see what happens with 6G and beyond. That will be the first standard where Qualcomm has to compete with multiple competitors competing on quality and features. Apple won't wait for them to do whatever. And likewise, companies like Ericsson/Nokia might start talking directly to Apple about making sure their network infrastructure works great with IOS.
> Modems are speed limited by networks which evolve on a MUCH slower timescale
While true, at some point networks do catch up, and many people hang on to their phones for several years at this point. If that's the case, whether having a better modem makes a difference in the end depends on which dimension of "faster" you're looking at:
Just using more and more Hz of spectrum on a single device does indeed not help anything other than peak speeds in empty cells (something I personally don't care about in the slightest in a phone), but being able to deliver more bit/Hz/s (through spatial multiplexing, better beamforming etc.) does eventually benefit everyone in congested/spectrum-limited situations.
> are phone modems a key source of security issues
Cellular models are also a common exploit vector: https://security.googleblog.com/2024/10/pixel-proactive-secu...
Maybe the phone does not matter that much: https://cellularsecurity.org/ransacked.html
>High signal: 0.88 watts for the iPhone 16, 0.67 watts for the iPhone 16e
>Low signal: 0.81 watts for the iPhone 16, 0.67 watts for the iPhone 16e
For the legacy modem, why is power consumption lower in lower signal areas? I naively expected more power use, either searching for signal or amplifying more. Is the power consumption mostly a factor of data throughput?
Could this be a translation issue, I wonder? The original source appears to be Chinese.
Perhaps "high signal" means not "conditions where the phone reports high RSS" but "conditions where the phone is transmitting at high power".
Apple was able to get 18% power savings on their modem (a quick average).
If Qualcomm cared they could do the same thing. That's gigawatts of energy, over billions of cellphones, being wasted by modems because the vendor doesn't care.
The problem, of course, is the cost of that inefficiency is external to the vendor.
Maybe Qualcomm will improve their power performance in the next-next-generation.
> The problem, of course, is the cost of that inefficiency is external to the vendor.
How so? The impact of power efficiency is arguably as immediate to a mobile phone manufacturer as it gets. These things are battery powered, and many people are struggling to make it through a single day on one charge!
Of course, that's practically exactly what battery size is tuned to – making it through exactly one day for most users. Knowing Apple, they'll just make the next flagship half a millimeter thinner again rather than risking one and a half days of battery life...
We would not save fully this energy, we would spend it somewhere else more useful (rebound effect / Jevons)
> t's not possible to determine from these tests how much of those longer streaming times are down to efficiency or the iPhone 16e's larger battery.
Hm.
I really hope these modems will come to MacBooks.
Man, this has been talked about for years now. I wish they'd just pull that trigger. A cell enabled macbook is the dream.
I suspect they have some numbers on how much the ipad cellular is sold and how often that feature is used and they may be using that to determine whether it's worth it.
But now that they develop the chip in house, I don't see why they wouldn't include it. In fact, given they put an Apple Watch chip and extra display in their macbooks for a few years, I don't see why they haven't done it years ago.
edit: I googled, apparently Qualcomm's licensing fees are a percentage of the whole device the chipset is installed in. I suppose it's less of a factor now that some iPhones are more expensive than Macbooks though.
Apple doesn’t even put NFC hardware into iPads. Despite the obvious use cases (nevermind platform homogeneity)
Yeah for a while I felt that a valid counter-argument was "modems are iterating faster than laptop cycles so you don't want to be stuck with an older modem" – but with 5G modems and all the architecture changing over to IP with LTE years ago, now you can probably have a C1 in a laptop for the next 5 years and be perfectly content without cell phone networks phasing those out.
The issue for LTE and 5G is/was less tech evolution as it was just new bands coming online. At least in the US, you really did need a pretty recent phone to get decent performance on TMobile or even more recently C-Band 5G
ARM MacBooks before launch were talked about for 10 years too. Same with AVP. Cannibalizing the iPad and iPhone market is a big concern for Apple
Even better if Apple started its own cellular ISP.
It would just be an MVNO I would think, but still cool.
TBH Id sign up in half a second
Is there really a need to carry multiple cellular modems? We already have one in our pocket most of the time and can easily share it in a single click. Feels like the vast majority of use cases are already covered by that solution and it doesn't require a second expensive data plan.
Does anyone know whether the C1 uses standard AT commands, or is it some custom protocol?
I'm curious if they tested this in highly-congested cells, since that's when cellular modems have to do the most "magic" and probably hit weird implementation edge cases. I think using the C1 in a packed football stadium is the real test of whether Apple has caught up to Qualcomm.
Other than mmwave (which this modem does not support), I don't think there's much more magic to do in an empty vs. a congested cell for a baseband.
5G spectrum is managed by the network, so the only thing that will happen is that each device will get shorter/narrower time/frequency slices, but there shouldn't be any unexpected cross- or same-channel interference of the type that is common for unmanaged protocols like 802.11 or Bluetooth.