For a while, we’ve been seeing reports that Apple could become one of the key major customers Intel signed up for chip production in its Foundry fabs—a critical factor in Intel’s long-term effort to turn the fab division around. At first, this may have been little more than Intel shareholder’s wishful thinking, but it appears the contract may indeed happen, with Intel set to manufacture Apple chips—for laptops as well as mobile phones.
While previously uncertain, the Apple deal reportedly moved significantly closer to reality in November, according to analyst Ming-Chi Kuo, whose information indicated production of one of the M-series processors for MacBooks. It should be noted that the current atmosphere in the United States provides strong incentives towards such an outcome by itself, with the Trump administration pushing for domestic manufacturing and Intel being the only major foundry manufacturer headquartered directly in the U.S.
According to newer reports, these rumors were apparently legitimate. Bloomberg and The Wall Street Journal reported this month that Apple and Intel have already entered into preliminary agreements regarding use of Intel’s chip fabs. These are not yet final binding orders for specific production volumes, but the probability that future Apple devices will be sold with chips manufactured by Intel (though no longer designed by Intel itself, as in the past) has now become very high. According to The Wall Street Journal, negotiations between the two companies have already been ongoing for more than a year.
It is still not entirely clear what Apple will manufacture in Intel’s fabs. In theory, they do not necessarily have to be flagship processors such as A19 for mobile devices or M5 for notebooks right away. Apple also produces chips for watches, cellular modems (C1, upcoming C2), and Wi-Fi/Bluetooth chips (N1). Apple could choose one of these for Intel manufacturing in theory, since maximum performance and cutting-edge power efficiency are less critical for those. On the other hand, these chips also have specialized analog circuitry requirements, which could make migration from TSMC to Intel more problematic, ironically.
M7 and A22?
Reports have emerged, however, that Apple will actually manufacture its high-profile processors or SoCs in Intel Foundry—allegedly targetting the M7 processor for notebooks as the first generation. That product is considered less critical than iPhone chips, which represent a far more important business segment for Apple. The M7 is reportedly set to use Intel’s enhanced 1.8 nm 18A-P process, which has been discussed recently. The actual product could reach the market sometime in late 2027 (with the M6 generation expected this year), in theory.
- Read more: Intel announces most advanced node to date: 18A-P technology
- Read more: Intel announces 1.4nm process, first node with high-NA technology
Reportedly, further cooperation is also already under discussion using Intel’s next-generation 1.4 nm 14A process. This colaboration could eventually produce a mobile SoC for iPhones—possibly the A22 generation in 2028 or A23 in 2029. It’s likely contingent on the M7 partnership proving viable first and demonstrating that Intel’s process technology is suitable. In other words, this appears to be a case of a lower-risk collaboration on a product with more limited reach being used as a test before Apple commits to a larger contract.
It is possible that use of Intel’s process for these products could somewhat reduce their specifications—long-term trends show Intel’s manufacturing processes remain slightly behind TSMC, whose most advanced nodes Apple currently uses. In practice, this could mean the Intel fabbed M7 and A23 will end up with somewhat lower clock speeds (and therefore performance) than would be possible with TSMC technology. Yield may also be somewhat worse.
Intel will most likely compensate Apple for this through lower manufacturing costs, however, which could ultimately make the arrangement quite profitable for Apple overall. If Intel is forced to offer Apple substantial discounts, though, the contract will naturally do less to improve Intel’s financial performance and long-term health than it would if Intel were negotiating from a stronger position without relying heavily on pricing concessions.
Sources: techPowerUp, The Wall Street Journal, Jukan
English translation and edit by Jozef Dudáš
⠀⠀
