4 ms·

This is the very reason why I had loaded up on significant amounts of Qualcomm stock earlier this year. Apple pretty much had no choice but to settle a deal wit

This is the very reason why I had loaded up on significant amounts of Qualcomm stock earlier this year. Apple pretty much had no choice but to settle a deal with them or get screwed on 5G. I have no doubt that Apple will likely attempt to sever their dependency on Qualcomm when they make eventually make their own modems (they've made incredible strides with their SoC's under Johny Srouji), but it will still be at least a few years before that will even be a reasonable option. The complexity of engineering that goes into a high-performance modem is nothing to sneeze at, especially when Qualcomm owns a significant portion of 5G-related patents. Even Intel threw in the towel upon the realization that the capital expenditure wouldn't be worth it, and it would be impossible to compete with Qualcomm profitably (Intel stock actually jumped up accordingly). Qualcomm will continue to collect licensing/royalties from other manufacturers like Huawei as well, and is in an excellent position frankly.

wyldfire 8y ago
> The complexity of engineering that goes into a high-performance modem is nothing to sneeze at For Apple to make a 5G modem that's competitive on power and area with Qualcomm's will be a significant feat. But Apple's got some really good design teams and I'm sure they'll figure it out eventually.
WrtCdEvrydy 8y ago
> when they make eventually make their own modems As a former person who was aware of Blackberry Mobility's Modem Firmware division by working within it.... LOL!
- brmgb 8y ago
  Apple has an extremely competent team of chip designer, decades of experience in manufacturing and plenty of money to snatch top talents. If they want to be serious about modem design and it seems they want to, they will most likely succeed.
  - WrtCdEvrydy 8y ago
    > extremely competent team of chip designer > decades of experience in manufacturing > plenty of money to snatch top talents You understand that Blackberry created their own mobile driver because the 850 released in 1999 predated chip companies providing firmware, right?
    - brmgb 8y ago
      I'm not sure I follow you here. You seemed to be implying that after witnessing Blackberry failure to properly make their own modem you found the idea of Apple doing it laughable. I was pointing to you that : 1. Contrary to Blackberry when they started trying Apple has been designing chips (CPU) for some time and has knowledge internally about how to conduct this kind of projects and how to build a team for it ; 2. Apple already understands and masters the manufacturing part of chip making ; 3. Apple is so rich and has such a brand that they can most likely buy the people mastering the knowledge they still lack (anything modem specific) from their competitors. The idea that Blackberry was doing before everyone else seems to only make my points stronger. Apple will have less R&D and engineers training to do.
      - WrtCdEvrydy 8y ago
        I'm not arguing it, I'm just saying it's a fool's errand. If you build your own chip, you have to build your own mobile stack, and then maintain it and enhance it with the newer technologies, so it becomes a mountain of technical debt while your competitors can take what Qualcomm gives them and roll out a phone. This assumes that Apple will maintain it's location as the top of the phone line and won't get eaten by someone else, if they slip, that mobile stack becomes a liability.
  - mattthebaker 8y ago
    None of these advantages invent you a novel radio technology that is competitive with 5G yet unencumbered by existing patents. Designing a modem to spec is the easy part.
cududa 8y ago
Slight modification to your narrative here (mostly all right) Intel through it in once Apple signaled there was a minuscule likelihood they’d go with one of their 5G modems anytime soon
Traster 8y ago
> Even Intel threw in the towel upon the realization that the capital expenditure wouldn't be worth it Intel doesn't seem to me like a company that is able to embark on a new technology, innovate and surpass the competition. Take a look at dGPUs, Intel has been getting smashed about by Nvidia for decades and there's a fantastically long history of Intel slowing buying up and killing hundreds of small GPU companies. If I were to pick someone to move into the Modem space and succeed it wouldn't have been Intel.
- TazeTSchnitzel 8y ago
  Intel's integrated GPUs have different goals than NVIDIA's dedicated GPUs. Power efficiency and video decoding and encoding is more important to Intel than raw gaming performance.
  - andrewnicolalde 8y ago
    I think they’re talking about Intel’s unsuccessful foray into the dGPU market, not the integrated GPUs on intel CPUs.
    - kristofferR 8y ago
      They're talking about the upcoming Intel dGPUs, not past attempts.
NicoJuicy 8y ago
I'm interested in your sources that made you decide to buy Qualcomm ? :)
- giobox 8y ago
  It wasn’t entirely impossible to reach this conclusion through sensible speculation alone. You don’t need to be Apple to know they only had three options for modems really: 1. Keep buying from Qualcomm. 2. Try and prop up Intel’s modem efforts in hope it becomes viable competitor to Qualcomm to either increase leverage or just get Intel good enough that you don’t need Qualcomm anymore. 3. Build own modems. They tried to make 2. work, and 3. as others have noted simply isn’t viable for several years at least presumably. Given Intel’s pretty poor track record at entering into new markets and succeeding, it wasn’t a bad bet that 2. and 3. was doomed to failure forcing a retreat to 1. We already knew from hardware tests that Intel’s 4G modems are not competitive with Qualcomm’s too, so not unreasonable to speculate the future 5G ones might not be as good either. Thanks to Intel’s press release the other day we can largely confirm too that their 5G modems weren’t going to be competitive. This article suggests there was also option 4, which is not something I’d have as easily speculated: 4. Try to engineer legal/market conditions that will negatively affect Qualcomm’s business to drive modem prices down. It is this option that seems to have failed this week, if this article’s analysis of Apple’s strategy is correct.

wyldfire 8y ago

> The complexity of engineering that goes into a high-performance modem is nothing to sneeze at For Apple to make a 5G modem that's competitive on power and area with Qualcomm's will be a significant feat. But Apple's got some really good design teams and I'm sure they'll figure it out eventually.

WrtCdEvrydy 8y ago

> when they make eventually make their own modems As a former person who was aware of Blackberry Mobility's Modem Firmware division by working within it.... LOL!

brmgb 8y ago
Apple has an extremely competent team of chip designer, decades of experience in manufacturing and plenty of money to snatch top talents. If they want to be serious about modem design and it seems they want to, they will most likely succeed.
- WrtCdEvrydy 8y ago
  > extremely competent team of chip designer > decades of experience in manufacturing > plenty of money to snatch top talents You understand that Blackberry created their own mobile driver because the 850 released in 1999 predated chip companies providing firmware, right?
  - brmgb 8y ago
    I'm not sure I follow you here. You seemed to be implying that after witnessing Blackberry failure to properly make their own modem you found the idea of Apple doing it laughable. I was pointing to you that : 1. Contrary to Blackberry when they started trying Apple has been designing chips (CPU) for some time and has knowledge internally about how to conduct this kind of projects and how to build a team for it ; 2. Apple already understands and masters the manufacturing part of chip making ; 3. Apple is so rich and has such a brand that they can most likely buy the people mastering the knowledge they still lack (anything modem specific) from their competitors. The idea that Blackberry was doing before everyone else seems to only make my points stronger. Apple will have less R&D and engineers training to do.
    - WrtCdEvrydy 8y ago
      I'm not arguing it, I'm just saying it's a fool's errand. If you build your own chip, you have to build your own mobile stack, and then maintain it and enhance it with the newer technologies, so it becomes a mountain of technical debt while your competitors can take what Qualcomm gives them and roll out a phone. This assumes that Apple will maintain it's location as the top of the phone line and won't get eaten by someone else, if they slip, that mobile stack becomes a liability.
- mattthebaker 8y ago
  None of these advantages invent you a novel radio technology that is competitive with 5G yet unencumbered by existing patents. Designing a modem to spec is the easy part.

cududa 8y ago

Slight modification to your narrative here (mostly all right) Intel through it in once Apple signaled there was a minuscule likelihood they’d go with one of their 5G modems anytime soon

Traster 8y ago

> Even Intel threw in the towel upon the realization that the capital expenditure wouldn't be worth it Intel doesn't seem to me like a company that is able to embark on a new technology, innovate and surpass the competition. Take a look at dGPUs, Intel has been getting smashed about by Nvidia for decades and there's a fantastically long history of Intel slowing buying up and killing hundreds of small GPU companies. If I were to pick someone to move into the Modem space and succeed it wouldn't have been Intel.

TazeTSchnitzel 8y ago
Intel's integrated GPUs have different goals than NVIDIA's dedicated GPUs. Power efficiency and video decoding and encoding is more important to Intel than raw gaming performance.
- andrewnicolalde 8y ago
  I think they’re talking about Intel’s unsuccessful foray into the dGPU market, not the integrated GPUs on intel CPUs.
  - kristofferR 8y ago
    They're talking about the upcoming Intel dGPUs, not past attempts.

NicoJuicy 8y ago

I'm interested in your sources that made you decide to buy Qualcomm ? :)

giobox 8y ago
It wasn’t entirely impossible to reach this conclusion through sensible speculation alone. You don’t need to be Apple to know they only had three options for modems really: 1. Keep buying from Qualcomm. 2. Try and prop up Intel’s modem efforts in hope it becomes viable competitor to Qualcomm to either increase leverage or just get Intel good enough that you don’t need Qualcomm anymore. 3. Build own modems. They tried to make 2. work, and 3. as others have noted simply isn’t viable for several years at least presumably. Given Intel’s pretty poor track record at entering into new markets and succeeding, it wasn’t a bad bet that 2. and 3. was doomed to failure forcing a retreat to 1. We already knew from hardware tests that Intel’s 4G modems are not competitive with Qualcomm’s too, so not unreasonable to speculate the future 5G ones might not be as good either. Thanks to Intel’s press release the other day we can largely confirm too that their 5G modems weren’t going to be competitive. This article suggests there was also option 4, which is not something I’d have as easily speculated: 4. Try to engineer legal/market conditions that will negatively affect Qualcomm’s business to drive modem prices down. It is this option that seems to have failed this week, if this article’s analysis of Apple’s strategy is correct.