Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/index.php?threads/iphone-18-series-rumor-claims-2026-bound-a20-pro-chipset-could-be-built-by-intel.21538/
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2021370
            [XFI] => 1050270
        )

    [wordpress] => /var/www/html
)

iPhone 18 series: Rumor claims 2026-bound A20 Pro chipset could be built by Intel

Brady

Member
As shared by Chinese leaker Fixed Focus Digital, there are unlikely to be any significant node improvements for next year's A19 Pro chip. The current-gen A18 chipsets are built on TSMC's N3E node, and it appears the A19 generation will make the move up to the Taiwanese company's N3P node, likely alongside Qualcomm's Snapdragon 8 Elite 2 as has already been rumored. Interestingly, however, the source claims the A20 chips will ditch TSMC entirely in favor of Intel. Supposedly, rather than contract TSMC to build the A20 chip for the iPhone 18 series, Apple will look to Intel and its 2nm 20A process.

iPhone 18 series: Rumor claims 2026-bound A20 Pro chipset could be built by Intel
 
Let’s assume the obvious that this is almost 100% nonsense.

But for sake of discussion, what could Intel manufacture for Apple? Something that is less mission critical than the SoC for their literal flagship cash cow. Intel sold its mobile modem business, and Apple is purportedly building its own WiFi chip; could it be those? What else even is there?

As a leverage play could Apple dual-source like they did way back then with Samsung and TSMC? Would @Daniel Nenni say that would threaten Apple’s “lead customer” status? Tim Cook is notoriously ruthless with his supply chain, but TSMC is not like the others…
 
Let’s assume the obvious that this is almost 100% nonsense.

But for sake of discussion, what could Intel manufacture for Apple? Something that is less mission critical than the SoC for their literal flagship cash cow. Intel sold its mobile modem business, and Apple is purportedly building its own WiFi chip; could it be those? What else even is there?

As a leverage play could Apple dual-source like they did way back then with Samsung and TSMC? Would @Daniel Nenni say that would threaten Apple’s “lead customer” status? Tim Cook is notoriously ruthless with his supply chain, but TSMC is not like the others…
No matter how ruthless someone is you can't control everything in the supply chain
 
Sounds unlikely but it would not be the first time Apple tried to second source such a chip. They originally used Samsung but then they flip-floped between Samsung and TSMC with later chips.
 
Sounds unlikely but it would not be the first time Apple tried to second source such a chip. They originally used Samsung but then they flip-floped between Samsung and TSMC with later chips.

Apple had to second source with Samsung at 14nm (iPhone 6+) because TSMC did not have the capacity ready at 16nm. It was a total failure since the phones had different performance and power specs. I highly doubt Apple will second source again and certainly not at Intel 20A. TSMC N3 is the best process on the market with nothing to compare.

 
In 2026, TSMC's N2 process will be in production only in Taiwan. If Trump enacts tariff on made-in-Taiwan chips, the impact on Apple can be mitigated by a second source from Intel. Apple could even segment their phones with different model/price based on US requirements. If Apple wants to have Intel inside by 2026, it is not too early to start working on it already..
 
In 2026, TSMC's N2 process will be in production only in Taiwan. If Trump enacts tariff on made-in-Taiwan chips, the impact on Apple can be mitigated by a second source from Intel. Apple could even segment their phones with different model/price based on US requirements. If Apple wants to have Intel inside by 2026, it is not too early to start working on it already..

I was wondering how many of the top semiconductor companies supported Trump. I highly doubt Tim Cook did nor Jensen Huang or Lisa Su. So I doubt that they have any clout in the Trump administration. I do believe that Elon Musk would be against tariffs as it would effect most of his businesses especially Tesla and Xai.

Exciting times in the semiconductor industry, absolutely.
 
I was wondering how many of the top semiconductor companies supported Trump. I highly doubt Tim Cook did nor Jensen Huang or Lisa Su. So I doubt that they have any clout in the Trump administration. I do believe that Elon Musk would be against tariffs as it would effect most of his businesses especially Tesla and Xai.

Exciting times in the semiconductor industry, absolutely.
Hence, the risk is especially high for American fabless companies if Trump decides to get your "attention" by disrupting the Asia supply chain with tariff. Intel could potentially benefit the most, but it requires Pat Gelsinger developing strong rapport with the new White House. But, it's not clear if Intel showed any support to Trump before the election.
 
Why do people jump to the conclusion this could not happen? 1) Apple has shown before they will use a 2nd source if needed, 2) Intel 18A should be well matured in 2026 when this is supposed to happen. 3) Who knows what Trump will do. If there are tariffs, then Apple could be paying a higher price for TSMC N2. 3) The Taiwan government has recently made it law that TSMC's newest process node manufacturing cannot be done outside of Taiwan. This means, by law their Arizona site will never be their leading node. This all points to Apple at least laying the groundwork for this possibility. Is it reasonable to think Apple is considering it? Yes. Will it happen? Don't know. Also with Intel 14A which will be coming online in 2026/2027, Intel will have a lead over TSMC since Intel has at least a year headstart with ASML high NA.
 
Last edited:
Why do people jump to the conclusion this could not happen? 1) Apple has shown before they like having more than one source. 2) Intel 18A should be well matured in 2026 when this is supposed to happen. 3) Who knows what Trump will do. If there are tariffs, then Apple could be paying a high price for TSMC N2. 3) The Taiwan government has recently made it law that TSMC's newest process node manufacturing cannot be done outside of Taiwan. This means, by law their Arizona site will never be their leading node. This all points to Apple at least laying the groundwork for this possibility. Is it reasonable? Yes. Is it reasonable to think Apple is considering it? Yes. Will it happen? Don't know. Also with Intel 14A which will be coming online in 2026/2027, Intel will have a lead over TSMC since Intel has at least a year headstart with ASML high NA.

Sorry, that is not reasonable. Intel is a much higher risk for Apple than TSMC. Nor do they have the yield/capacity Apple requires. Apple produces complex SoCs for mobile devices, not chiplet based designs. Apple has not shown before that they like having more than one source. That is absolutely false.
 
Sorry, that is not reasonable. Intel is a much higher risk for Apple than TSMC. Nor do they have the yield/capacity Apple requires. Apple produces complex SoCs for mobile devices, not chiplet based designs. Apple has not shown before that they like having more than one source. That is absolutely false.
Yep...changed my wording to reflect that Apple has shown they are willing to 2nd source if need be. I don't know the intracies of SoC vs chiplet, etc... I'm just pointing out there are likely many changes ahead with Tariffs and relationships amongst countries and companies. If TSMC is not allowed to have any leading edge fab outside of Taiwan, it would be wise for Apple to at least start thinking about using Intel. That's all I'm trying to convey.
 
Yep...changed my wording to reflect that Apple has shown they are willing to 2nd source if need be. I don't know the intracies of SoC vs chiplet, etc... I'm just pointing out there are likely many changes ahead with Tariffs and relationships amongst countries and companies. If TSMC is not allowed to have any leading edge fab outside of Taiwan, it would be wise for Apple to at least start thinking about using Intel. That's all I'm trying to convey.

Sorry again, that is not true. Apple second sourced one time and it failed miserably so they are not "willing" to do it again.

I seriously doubt Trump will cripple the US economy with Taiwan tariffs. It would set us back in the AI arms race against China. I also believe TSMC could make leading edge semiconductors in the US. Without the US market TSMC will fail. Without TSMC Taiwan's Silicon Shield will fail. It really is as simple as that. It is in the world's best interest to keep Taiwan independent and tariff free, absolutely.
 
US has a substantial trade deficit with Taiwan. In 2023, it was $48B. In 2024, it will likely reach $70B due to strong growth in AI-related semi trade. We can expect the AI-led trade gap with Taiwan to widen further in coming years. Actually, the current trajectory will likely lead to Taiwan becoming the third highest trade surplus nation by 2030 - only behind China and Mexico.

So, the question is not what Apple and Nvidia would or might do. It is a question about how the new White House sees the significance and urgency to address Taiwan's growing trade surplus as well as the geopolitical risk related to Taiwan-based AI hardware manufacturing.
 
Last edited:
Assuming it would be 18A there is zero chance of them having all the IO needed by an Apple chip on that process. Look how Intel plan to use 18A on their own chips, the 18A is purely logic and then the IO (and SRAM) is on a Intel3 substrate. In principle that could interest Apple but the difficulty of getting volume production at low prices makes it implausible.

For it to make sense the A20 would need to be a huge functional shift, for example growing a large AI unit, that makes partitioning into two chips (pure logic + IO/low power) worth considering. Are we there yet?

Simplest explanation is bad faith deliberate rumor.
 
Back
Top