Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Good to hear about this. I think initially the entry level products will be using this chip before it expands to more devices. It will be another 2 to 3 years before devices with this will be available.
 
  • Like
Reactions: mganu
Made-up drivel. No way Apple gets Intel to make anything for them if it’s inferior to their existing products.

Not necessarily. If the intel-manufactured chips are good enough, and cheaper than TSMC chips, Apple has products that can use them.

For example, Apple could (hypothetically) have an A21 chip made by Intel for use in the standard iPhone and iPhone e, and an A21 Pro and A21X made by TSMC for use in Pro iPhone and iPads.
 
This decision is entirely political.

Not entirely. TSMC doesn’t need Apple like they used to (plenty of AI customers), so Apple will be finding it harder to drive a good deal on pricing than they have in the past.

And it’s just good business sense to have multiple suppliers for critical components if you can. You can play suppliers off against each other on pricing. And if one supplier screws up, you can order more from the other to make up for it.
 
Nothing newsy here.

Apple would like to diversify if they can rather than rely on one foundry, TSMC, and Intel gets some extra business, win-win.

Apple will never go back to x86.
 
  • Like
Reactions: TenSeven
Intel has to get revenue somehow, not a horrible idea to be a manufacture.
Intel's stock price dropped 15% today. Correction, 16% and still falling.

Intel's stock is down primarily because in the last 6-7 months it was hyped up beyond rational threshold n the first place. Trump threw taxpayer money at Intel ... it have to 'win'. Blah blah blah

Intel has a big 'mess' to clean up and it is simply just going to take time. A bunch of the "what are you going to do for me next quarter" folks dropped out. Surprise, surprise. Those folks bought the hype that Intel was going to be 'back on top' super soon and are no dropping out.

A 'correction' says lots more about the folks who bought the stock than it does about the company.
 
  • Like
Reactions: TenSeven
The question: can Intel's fabs produce 3 nanometer chips that meets Apple's power requirements?

Intel's 18A is about 24-30% more power efficient than Intel 3. If using last Intel chips used by Apple in 2020 as a reference point then Intel 3 is two generations past what those chips were. If looking at TSMC N3B/N3E then Intel is competitive.

Power efficiency is not a momental stumbling block and large disconnect from where Apple designs target. Apple uses more cache heavy designs. So decent chance Intel chip would be bigger.. Bigger chips come with lower yields so ends up costing more. ( Apple doesn't like to pay more than they have to). If there is a wafer queue bidding war at TSMC and those wafers go high, then those chips cost more also.

If Apple wanted to make a "fast enough for most users" A-series SoC for the iPhone xx-e (and entry iPad) product, that would not be extremely hard.


Is Intel 18A going to be more power efficient than TSMC N2 ? No. But N2 is going to be substantively more expensive than N3E was/is. If Apple is considering iterating on N3P longer (to save money) versus Intel 18A (or 14A) it gets closer.


P.S. Non technology issue , how much volume could Apple purchase is something that gets questionable if trying to loop in the full non-Pro iPhone line up. Unless, Intel's product footprint gets smaller that is a question if they have enough 'extra' capacity in the 2027-2028 timeframe to handle that. Less than 10M units of a small-medium die maybe. high 10's of millions would be believable with the Ohio Intel fab open , but it won't be.
 
Last edited:
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.