If true then the Apple Silicon team is truly executing very well!
If only including the modem in a few SKUs then that would imply the modem is not on the AS SoC but a distinct chip still and not delivering the full cost saving and space saving potential.
Overall I am hopeful that they managed to include it in all SoC's for the mobile handsets and iPads.
Doesn’t look like Apple is “struggling” now. Bringing in house 5G chips is huge and another Apple in house triumph. The biggest mistake anyone can make is to underestimate Tim Cook. Great job!
Tim Cook can be considered for ultimate blame in not delivering this years ago.
He became CEO in 2011. The first mobile 5G modem launched in 2019 following 10 years of development.
Strategic goals sit firmly at his door. He goofed on Qualcomm and 5G. He got into a worldwide patent spat with a possible 5G supplier (Qualcomm, one of very few) and had no real alternative to already poor performing Intel modems.
When 5G hit mobile, Intel failed to deliver (was that a surprise seeing how their 4G modems performed?) and Apple was caught with its pants down.
Suddenly it was kiss-and-make-up with Qualcomm (basically on the steps of the court with battle about to commence), signing a deal for them to supply Apple with (back then) an older bolted-on modem (a true Yikes! moment) and Apple bought Intel's failing 5G modem division.
They've been rushing (and seemingly failing) to deliver ever since. They then signed the longer term deal with Qualcomm mentioned in the article.
Strategically, an in-house 5G modem was only considered very late in the day and as a result of earlier failed strategic decisions. They basically had to 'lick the toad' and get on with it. We can conclude that, strategically speaking, there was never an in-house 5G modem on the road map.
In terms of forward thinking I can confirm that Apple is currently working in a group in Europe including Huawei on network sensing technologies but that is for 6G.
Shortsighted, It took 13 years for Apple to replace Intel with Apple Silicon see theshambolic Intel today? Apple is well rid of them.
But where would Apple have been without Intel back then? And what about the competition that is coming through now?
Where would they have been without Motorola or IBM? They also supplied Apple with cpu's during the glory days and Apple had to dump them like Intel because of their shortsighted vision of the future wasn't Apples you know smaller, faster more powerful chips Motorola, IBM or Intel could have been out in front today instead of dead (Motorola) or comatose (IBM, Intel).
Doesn’t look like Apple is “struggling” now. Bringing in house 5G chips is huge and another Apple in house triumph. The biggest mistake anyone can make is to underestimate Tim Cook. Great job!
Tim Cook can be considered for ultimate blame in not delivering this years ago.
He became CEO in 2011. The first mobile 5G modem launched in 2019 following 10 years of development.
Strategic goals sit firmly at his door. He goofed on Qualcomm and 5G. He got into a worldwide patent spat with a possible 5G supplier (Qualcomm, one of very few) and had no real alternative to already poor performing Intel modems.
When 5G hit mobile, Intel failed to deliver (was that a surprise seeing how their 4G modems performed?) and Apple was caught with its pants down.
Suddenly it was kiss-and-make-up with Qualcomm (basically on the steps of the court with battle about to commence), signing a deal for them to supply Apple with (back then) an older bolted-on modem (a true Yikes! moment) and Apple bought Intel's failing 5G modem division.
They've been rushing (and seemingly failing) to deliver ever since. They then signed the longer term deal with Qualcomm mentioned in the article.
Strategically, an in-house 5G modem was only considered very late in the day and as a result of earlier failed strategic decisions. They basically had to 'lick the toad' and get on with it. We can conclude that, strategically speaking, there was never an in-house 5G modem on the road map.
In terms of forward thinking I can confirm that Apple is currently working in a group in Europe including Huawei on network sensing technologies but that is for 6G.
Shortsighted, It took 13 years for Apple to replace Intel with Apple Silicon see theshambolic Intel today? Apple is well rid of them.
But where would Apple have been without Intel back then? And what about the competition that is coming through now?
Where would they have been without Motorola or IBM? They also supplied Apple with cpu's during the glory days and Apple had to dump them like Intel because of their shortsighted vision of the future wasn't Apples you know smaller, faster more powerful chips Motorola, IBM or Intel could have been out in front today instead of dead (Motorola) or comatose (IBM, Intel).
That's a pretty inaccurate take on what actually happened. It is ironic that the Apple part of PowerPC was the power hog. Cooling those chips down was an issue.
However, the Motorola side actually had some truly groundbreaking embedded, low power, PowerPC chips. At one point, a Motorola exec even claimed that PowerPC was sitting in around 50% of the world's cars. Not too shabby even if they were highly specialiased chips. PowerPC (both regular and embedded versions, but radiation hardened) also went into space. And supercomputers.
The Intel vs PowerPC thing was like the Plasma vs LCD situation. Volume and pricing were critical.
PowerPC hit the skids for thermals and not being able to keep up.
Intel was ahead (with PowerPC out of the way) and then had AMD to deal with.
That's how the industry tends to move. There is zero guarantees that you will always be on top. That applies to Apple too. We know because they've already been down that road.
Doesn’t look like Apple is “struggling” now. Bringing in house 5G chips is huge and another Apple in house triumph. The biggest mistake anyone can make is to underestimate Tim Cook. Great job!
Tim Cook can be considered for ultimate blame in not delivering this years ago.
He became CEO in 2011. The first mobile 5G modem launched in 2019 following 10 years of development.
Strategic goals sit firmly at his door. He goofed on Qualcomm and 5G. He got into a worldwide patent spat with a possible 5G supplier (Qualcomm, one of very few) and had no real alternative to already poor performing Intel modems.
When 5G hit mobile, Intel failed to deliver (was that a surprise seeing how their 4G modems performed?) and Apple was caught with its pants down.
Suddenly it was kiss-and-make-up with Qualcomm (basically on the steps of the court with battle about to commence), signing a deal for them to supply Apple with (back then) an older bolted-on modem (a true Yikes! moment) and Apple bought Intel's failing 5G modem division.
They've been rushing (and seemingly failing) to deliver ever since. They then signed the longer term deal with Qualcomm mentioned in the article.
Strategically, an in-house 5G modem was only considered very late in the day and as a result of earlier failed strategic decisions. They basically had to 'lick the toad' and get on with it. We can conclude that, strategically speaking, there was never an in-house 5G modem on the road map.
In terms of forward thinking I can confirm that Apple is currently working in a group in Europe including Huawei on network sensing technologies but that is for 6G.
Shortsighted, It took 13 years for Apple to replace Intel with Apple Silicon see theshambolic Intel today? Apple is well rid of them.
But where would Apple have been without Intel back then? And what about the competition that is coming through now?
Where would they have been without Motorola or IBM? They also supplied Apple with cpu's during the glory days and Apple had to dump them like Intel because of their shortsighted vision of the future wasn't Apples you know smaller, faster more powerful chips Motorola, IBM or Intel could have been out in front today instead of dead (Motorola) or comatose (IBM, Intel).
That's a pretty inaccurate take on what actually happened. It is ironic that the Apple part of PowerPC was the power hog. Cooling those chips down was an issue.
However, the Motorola side actually had some truly groundbreaking embedded, low power, PowerPC chips. At one point, a Motorola exec even claimed that PowerPC was sitting in around 50% of the world's cars. Not too shabby even if they were highly specialiased chips. PowerPC (both regular and embedded versions, but radiation hardened) also went into space. And supercomputers.
The Intel vs PowerPC thing was like the Plasma vs LCD situation. Volume and pricing were critical.
PowerPC hit the skids for thermals and not being able to keep up.
Intel was ahead (with PowerPC out of the way) and then had AMD to deal with.
That's how the industry tends to move. There is zero guarantees that you will always be on top. That applies to Apple too. We know because they've already been down that road.
If Apple had continued to use Motorola, IBM or Intel, Apple would be done as a company none of them were good over time Apple had nothing but trouble with all of them, Intels answer to all problems is to crank up the mhz and the watts (that is their solution for their latest chip) and that was alsotrue of Motorola, IBM, AMD, and Nvidia too, even Qualcomms new soc needs a 85-100 watts to get somewhat close to Apple M2 and low end M3 chips and it does so with bad battery life performance and mediocre test bench scores and that doesn't include Microsofts third time butchering the Windows emulation software for Arm.
Don't be surprised if they eventually integrate it into the SOC.
I know that’s what Qualcomm does but I wonder if that will really make sense for Apple, especially given the potential need for more AI processing power. It might make more sense to spend the SOC transistor budget on CPU/GPU/NPU and anything else that benefits from a high bandwidth, low latency connection to RAM. Leave everything else for a separate chip.
But if they can put it all on one chip and still get decent yields, then I guess sure — why not?
Don't be surprised if they eventually integrate it into the SOC.
I know that’s what Qualcomm does but I wonder if that will really make sense for Apple, especially given the potential need for more AI processing power. It might make more sense to spend the SOC transistor budget on CPU/GPU/NPU and anything else that benefits from a high bandwidth, low latency connection to RAM. Leave everything else for a separate chip.
But if they can put it all on one chip and still get decent yields, then I guess sure — why not?
Is the R1 that chip for now?
Realtime OS variation to handle the constant data stream, be it sensors or receivers. It gives them more volume on that chip and eventually will integrate with the A and M series to bring the cost of the VisionPro down by reducing the complexity.
Shortsighted, It took 13 years for Apple to replace Intel with Apple Silicon see theshambolic Intel today? Apple is well rid of them.
Apple didn't start working on replacing Intel the day after it switched from PowerPC. It was realized as possible in 2013 when the 64 bit A7 processor launched, IIRC. We're not really sure when they started, but it was probably about 5-6 years before they announced it. Apple has to roadmap its plans that far out just to make things happen at all.
Apple Silicon is powerful and will ultimately save Apple billions over Intel. We're only 4 years in and It's not done maturing.
Always interesting to see the community response to news like this...
It's not the chip, it's the IP...
Qualcomm will get paid whether it's a Qualcomm chip or Apple, and they will likely pay more to Qualcomm since they will not get rebates on the modem chipsets.
Hey, getting there is half the battle as I've been in the business for 30 years now - folks over in Santa Clara and Munich have been working hard - firmware and the protocol stack, that's been done for some time...
The delay, based on folks I know, was getting the baseband on to the A-Series chip itself, much like the efforts they're doing with BT and WiFi.
Intel was spending over a billion USD annually for the Apple modem contracts - and this was R&D - not just for development, but the larger expense with QA and certification and regulatory - it has to do everything in the spec, not just for 5G, but for LTE, UMTS, and GSM in a multi-mode environment.
When you do a Qualcomm based design, that's all included for the protocol stack - so it's just down to the RF testing...
Little know fact - the Intel Modems that supported CDMA (Sprint/Verizon) used licensed IP from Via directly - including their SW stack.
if apple can produce these for less, then hopefully they can charge less. it would be nice to see the option for cellular drop from about $130 to maybe $50.
On what planet is Apple—king of the upsell and margins so fluffy and plush it'd make a down jacket blush—going to pass along even a nickel of cost savings? Are you completely out of your mind?
This. In fact, it wouldn't surprise me if Apple were to charge more for it's own modems. Because obviously an Apple modem would be superior to Qualcomm's in every way, right?
As for comments about it taking so long, Apple does not want to pay anyone else licensing fees if it can avoid doing so, which means not using anyone else's patented technology, or buying those patents outright (or, cynically, using them anyway and hope they're invalidated). All that research takes time, and engineering a new modem to not infringe on existing patents is no small task.
if apple can produce these for less, then hopefully they can charge less. it would be nice to see the option for cellular drop from about $130 to maybe $50.
On what planet is Apple—king of the upsell and margins so fluffy and plush it'd make a down jacket blush—going to pass along even a nickel of cost savings? Are you completely out of your mind?
This. In fact, it wouldn't surprise me if Apple were to charge more for it's own modems. Because obviously an Apple modem would be superior to Qualcomm's in every way, right?
As for comments about it taking so long, Apple does not want to pay anyone else licensing fees if it can avoid doing so, which means not using anyone else's patented technology, or buying those patents outright (or, cynically, using them anyway and hope they're invalidated). All that research takes time, and engineering a new modem to not infringe on existing patents is no small task.
I don’t believe Apple would be superior in every way, unless you’re being sarcastic there. All modem makers pay other companies license fees for patents. Nobody has an exclusive on FRAND patents, or other useful, but not required patents. The biggest holders are Qualcomm, then Huawei. But there are others. Apple bought a large number and has developed others. From what I understand about their efforts, they’re very involved about future cell technologies such as 6. If so, maybe they believe that they can get ahead in some of that and be more ready to have a seat at the table. But for now, they’re very involved certainly would have to license more that a few. I’ve read that there are well over 100,000 working cert patents out there, possibly twice as many and they’re usually licensed out in tranches.
if apple can produce these for less, then hopefully they can charge less. it would be nice to see the option for cellular drop from about $130 to maybe $50.
On what planet is Apple—king of the upsell and margins so fluffy and plush it'd make a down jacket blush—going to pass along even a nickel of cost savings? Are you completely out of your mind?
This. In fact, it wouldn't surprise me if Apple were to charge more for it's own modems. Because obviously an Apple modem would be superior to Qualcomm's in every way, right?
As for comments about it taking so long, Apple does not want to pay anyone else licensing fees if it can avoid doing so, which means not using anyone else's patented technology, or buying those patents outright (or, cynically, using them anyway and hope they're invalidated). All that research takes time, and engineering a new modem to not infringe on existing patents is no small task.
I don’t believe Apple would be superior in every way, unless you’re being sarcastic there. All modem makers pay other companies license fees for patents. Nobody has an exclusive on FRAND patents, or other useful, but not required patents. The biggest holders are Qualcomm, then Huawei. But there are others. Apple bought a large number and has developed others. From what I understand about their efforts, they’re very involved about future cell technologies such as 6. If so, maybe they believe that they can get ahead in some of that and be more ready to have a seat at the table. But for now, they’re very involved certainly would have to license more that a few. I’ve read that there are well over 100,000 working cert patents out there, possibly twice as many and they’re usually licensed out in tranches.
Apple is very obviously preparing for 6G which is scheduled to be released in 2030 Apple almost certainly will be using their own in-house modems by then across-the-board if this report is true, a 2025 Apple modem release is almost certainly a trial run for what’s coming up in the near future, Qualcomm probably will not be in any Apple device after 2030 (which will be about 12 years and into Apples development for modems), whoops that’s one year less than the development period for Apple Silicon.
Apple’s first modem doesn’t have to be necessarily better than Qualcomm after all, how good is the Elite X SOC made by Qualcomm? Apples first modem, however needs to be 98-99% (see Googles Pixel modems) the future iterations by Apple is the most important part an Apple has shown in the last 25 years that they are willing to execute on product iteration and do the R&D necessary when compared to their competition and the same applies to what Qualcomm needs to do with their new Elite SOC their biggest problem will be getting Microsoft to feel a sense of urgency with the Windows software.
Comments
If only including the modem in a few SKUs then that would imply the modem is not on the AS SoC but a distinct chip still and not delivering the full cost saving and space saving potential.
Overall I am hopeful that they managed to include it in all SoC's for the mobile handsets and iPads.
https://news.ycombinator.com/item?id=41058791 Intel is sinking market inertia is the only thing keeping them up....
However, the Motorola side actually had some truly groundbreaking embedded, low power, PowerPC chips. At one point, a Motorola exec even claimed that PowerPC was sitting in around 50% of the world's cars. Not too shabby even if they were highly specialiased chips. PowerPC (both regular and embedded versions, but radiation hardened) also went into space. And supercomputers.
The Intel vs PowerPC thing was like the Plasma vs LCD situation. Volume and pricing were critical.
PowerPC hit the skids for thermals and not being able to keep up.
Intel was ahead (with PowerPC out of the way) and then had AMD to deal with.
That's how the industry tends to move. There is zero guarantees that you will always be on top. That applies to Apple too. We know because they've already been down that road.
If Apple had continued to use Motorola, IBM or Intel, Apple would be done as a company none of them were good over time Apple had nothing but trouble with all of them, Intels answer to all problems is to crank up the mhz and the watts (that is their solution for their latest chip) and that was also true of Motorola, IBM, AMD, and Nvidia too, even Qualcomms new soc needs a 85-100 watts to get somewhat close to Apple M2 and low end M3 chips and it does so with bad battery life performance and mediocre test bench scores and that doesn't include Microsofts third time butchering the Windows emulation software for Arm.
Realtime OS variation to handle the constant data stream, be it sensors or receivers.
It gives them more volume on that chip and eventually will integrate with the A and M series to bring the cost of the VisionPro down by reducing the complexity.
That would fit with Apples regular playbook.
Apple Silicon is powerful and will ultimately save Apple billions over Intel. We're only 4 years in and It's not done maturing.
It's not the chip, it's the IP...
Qualcomm will get paid whether it's a Qualcomm chip or Apple, and they will likely pay more to Qualcomm since they will not get rebates on the modem chipsets.
Hey, getting there is half the battle as I've been in the business for 30 years now - folks over in Santa Clara and Munich have been working hard - firmware and the protocol stack, that's been done for some time...
The delay, based on folks I know, was getting the baseband on to the A-Series chip itself, much like the efforts they're doing with BT and WiFi.
Intel was spending over a billion USD annually for the Apple modem contracts - and this was R&D - not just for development, but the larger expense with QA and certification and regulatory - it has to do everything in the spec, not just for 5G, but for LTE, UMTS, and GSM in a multi-mode environment.
When you do a Qualcomm based design, that's all included for the protocol stack - so it's just down to the RF testing...
Little know fact - the Intel Modems that supported CDMA (Sprint/Verizon) used licensed IP from Via directly - including their SW stack.
Apple’s first modem doesn’t have to be necessarily better than Qualcomm after all, how good is the Elite X SOC made by Qualcomm? Apples first modem, however needs to be 98-99% (see Googles Pixel modems) the future iterations by Apple is the most important part an Apple has shown in the last 25 years that they are willing to execute on product iteration and do the R&D necessary when compared to their competition and the same applies to what Qualcomm needs to do with their new Elite SOC their biggest problem will be getting Microsoft to feel a sense of urgency with the Windows software.
https://en.wikipedia.org/wiki/6G Wikipedia 6G