You actually right, my second phone is pixel with latest Android 16, and he works stunning. What I need on Android it is Apple home kit and I will switch like without thinking. IOS26 is a joke
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Apple's Return to Intel Rumored to Extend to iPhone
- Thread starter MacRumors
- Start date
- Sort by reaction score
Just like a lot of huge companies different Intel divisions are almost different companies on their own. This deal is with Intel’s foundry business, they’re not buying x86 procs, they’re having Intel fab their procs
No, but the fact that the whole flipping article was about chips kinda makes that argument redundant.
Intel make chips - various sorts of chips, but most famously x86 CPUs - also most famously the "Intel chips" that Apple has used in the past which would also be the only reason for describing it as a "return to Intel".
No way the headline "Apple's return to Intel" wasn't crafted to make people think that Apple might be going back to x86 as clickbait. It's misleading, even if you can contrive some pedantic justification.
"Intel may make future Apple Silicon SoCs for Apple" - there you go.
I understand the need for foundry time and diversity, but I think it is a bad idea bringing Intel back to Apple devices.
Sounds like a Microsoft executive pulling strings. Hopefully the new CEO if elected can make sure the new CPU line is up to par with Apple standards; its not about execution of translated data, its all about how well Intel works with Apple, knowing completely well they manufacture CPU’s for their competitor.
In 1997 Microsoft gave Apple a $150 million cash infusion by buying non-voting stock and committed to continued development of Office and Internet Explorer for Mac. The cash infusion helped fund Steve Jobs' turnaround of Apple and avoided what was at the time certain bankruptcy. So yes, it was almost 30 years ago, but Microsoft saved Apple.
TSMC also manufactures chips for Apple's competitors, namely Qualcomm's Snapdragon line, AMD's Ryzen chips, all Nvidia GPUs, and even some current generation Intel chips.
Well, first they have to make sure Intel can fab their chips. Then they buy them. It would be wiiise my friend.I wouldn't be overly surprised if Apple buys Intel in the foreseeable future
If for one second, with the success of Apple Silicon, you would even entertain the idea that Apple would return to Intel-designed chips, then I guess you'll believe anything.
Do you call call Apple's M-series chips TSMC chips? Yeah, me neither...
You have to have first been somewhere to ‘return’ there. Apple was NEVER an Intel fab customer. So they can’t be a returning customer if this would be the first time.
If you want to arm flapping , hand wave at the overall corporate blobs , Apple runs a large datacenters. They have bought zero Intel data center gear over last couple of years ( no network cards , no CPUs , no o service support , etc )? Additionally, given Apple has to currently service and support Intel Macs that highly unlikely. Apple would had to drop Intel from every possible corporate purchases to become a customer who totally left . Again, to ‘return’ a precondition is that you leave. One foot in the doorway of a house and come back in isn’t ‘returning’ .
“Returning as a major customer “ is also an insertion. The author of the headline assumed that Apple 100% left as an Intel customer.
That said it is relatively common for ‘front page’ Macrumors articles to be written “National Inquirer” style . Headlines, skewed summaries, or anyhtihing that will spark high response rates and larger adviews is par for the course. Objective accuracy isn’t the top criteria.
So why would Apple exert all of this effort to create Apple Silicon and then decide to partner u with Intel again? Tim, you got some splainin' to do. Make it make sense!
...but that's the point of sensationalist, click-bait/rage-bait headlines. At best, people read them, think "NO WAY!!!?" and click on the article - and get something fairly mundane. At worst, people do believe the headlines and, once they've latched on to those, don't bother to read or try to properly understand the details.
EG:
ANS: Because, apart from making x86 processors, Intel are a major chip fabricator with the capacity to manufacture Apple Silicon chips - to Apple's specification - in the US as a "second source" as a back-up to TSMC who currently make the chips (mainly in Taiwan) for Apple. Crack open a newspaper to see why Apple might want a US-based "second source" for key components.
Last edited:
With the rumors of an Ann Pro power affordable MacBook you would think they would line up with those cohorts. If Apple prunes off the plain Mn SoC then the other Mac at the lower ‘half’ of the product line up would be the contemporary Ann Pro that used the same micro-architecture generation. But if tariff motivated then perhaps old enough process that TSMC has brought it to AZ at that point. If Ann Pro is primarily for thinnest iPhone then keeping that on TSMC has traction . The Mac ‘hand me down’ usage can just go ‘older’ to get cheaper.
The other problem is the disconnect with reality in assigning Apple highest volume SoC with Intel . Intel grossly underinvested in EUV fab machines. Intel has signed up for some TSMC N2 because even if their 18A and 14A work great they don’t have enough machines to make everything they would need for their own stuff. ( 14A using more multipatterning on current Gen EUV machines isn’t going to help ).
Intel’s Ohio fab has been pushed back to 2030 so that capacity isn’t there in 3 years.
An Intel variant just for an iPhone 20e ( followed by AppleTV , budget iPad) ? Maybe. ( if ‘e’ product makes it that far) But if the high volume Mac models are already piled on Intel, then Apple probably would have soaked up a substantial amount of Intel’s uncommitted capacity. Even more likely if others are also tariff threatened ( and/or AI hype bubble demand driven ) onto Intel’s limited capacity.
TSMC has made comments about about being overcommitted relative to wafer demand. So this doesn’t necessity have to solely be due to trade war stuff . Some of that demand is likely going to spin off to Intel. Apple isn’t the only customer throwing giant piles of money around ( e.g., Open AI buying up 40% of the DDR5 memory supply), so may not be able to buy as much as they want.
Last edited:
If Samsung had their act together, then Intel probably would not get the win. It is still competition .is just that Samsung has stumbled also . Samsung had a fab in USA long before TSMC did.
Additionally, the AI bubble demand has become a competition of who can spend the most money the fastest. If TSMC doesn’t have enough wafer starts left , then Apple ( and others ( will have root obuy from somewhere else. If there is a fixed number of EUV machines and TSMC doesn’t own them all then folks will have to gpsomehere else to get wafers processed.
The higher TSMC’s wafer coasts go the more likely Apple is going to start trying other people’s design kits. Apple isn’t going to accept complete garbage yields because that would increase costs as much as tariffs .
As for politics …whose? Saber rattling about forcing change in Taiwan hasn’t helped the comcpetion of TSMC diversifying the manufacturing locations before now. The current Admistration does care who makes the chip as long as it is made in the USA. That doesn’t automatically fpget Intel a win.
There is no evidence at all that 14A is horrible. ( Intel hasn’t finished 18AP design kit so there is no way they are close to done on 14A ). The current Intel CEO said he might shelve it because there was not enough external customers , not because it was bad. For the Macs Apple is reportedly taking 18AP which is the first iterative refinement of 18A. AP’s ramp is not going to be like the 18A ramp was . Big part of Intel fab problems leading the dance of external customers and external design tool ecosystem. That just takes time ( can’t buy your way into maturity there ).
They don’t have to use two suppliers in the same product. For example Intel in iPhone 20e ( later AppleTV , entry iPad ) and TSMC in iPhone 20 . Intel doesn’t have the capacity to offload most of the wafers that TSMC does on EUV machines. Intel undervestrd here and it will take a very long time to catch up with ability to ‘out volume’ even Samsung let alone TSMC .
18A/14A probably runs better at higher clocks . TSMC N2 wins on density , probably cost( backside hosts more no matter who you are ) , and perhaps energy efficiency . The 18A Intelchip will likely be bigger ( because Apple used SRAM heavy designs ) and cost more, but may be in same ballpark as energy consumption.
https://semiwiki.com/forum/threads/...advantage-over-intels-18a-sram-density.21630/
Intel and TSMC have gone for different trade offs here ( 18A/18AP versus N2 ) . At 14A and A16 there still will be different trade offs .
Don’t tell that to the Chinese government. 😀 They insist it is China. Technically they currently don’t have final say on what TSMC does / doesn’t do , but spending tons of time, money, and effort to change that.
errr , buying the modems division does not in anyway mean moniving away from the people who created them. Those modems on Intel 10
errr , buying the modems division does not in anyway mean moniving away from the people who created them. Those modems on Intel 10nm ? ( there was an in house version in flight when they bought them ) Yes they MBP Ed on from that . Pretty sure Intel switched to selling rebadged MediaTek modems. Intel is the one who MBP Ed on from what was Infineon modems.
Apple does not have anywhere near a trillion dollars. Off by orders of magnitude. Apple stockholders collectively have a trillion range. Apple does not.
Time may or may not line up . Intel has hedged somewhat since they first talked about “5 process nodes in 4 years” pep talks.
“…
The first one is, Intel still has the option to have either a Low-NA or a High-NA solution on our 14A technology, and its design-rule compatible, there will be no impact to the customers, depending on the path that we choose. Second, High-NA EUV is performing to the expectations, and we will introduce it at the right time, … ”
Fairly decent chance that Intel will be heavily multipatterning just like TSMC in that three year time frame. High-NA also means lower limits in the reticle size for chips. So far Apple’s quasi-chiplet streaky has been a big mismatch to that.
The large die, expensive datacenter customers don’t want to give up those big dies if they don’t have to. There is a tension of drawing smaller ‘lines’ and multipatterning faster . Intel has bet heavily on first and TSMC has bet heavily on the latter. Pretty decent chance Intel may win the boutique , ‘smallest possible’ market over the longer term. There is also decent chance that also won’t be the highest volume market
This situation is a bit like the hype where folks were proclaiming that Apple bought up all the N3B capacity when it was far more so that they were they only ones buying while Intel time shifted to substantially later. Now Intel is N3B wafer constrained ) . Intel got the early EUV machines also and then flip-flopped into mega multipatterning on DUV for Intel 10 .
It makes sense for Intel because they don’t really have a high volume legacy EUV process to sell .Nor multiple ,super high volumes external clients to service on a legacy EUV process. But it is a gamble.
The fallback to regular EUV option for 14A is a more likely park for “round 1” . It will probably be seen by external folks as less risky.
Since design rule compatible Intel probably won’t be pressing High-NA super hard in that first targeting. It is only future years out where they may get a gap. Intel needs to find the zone of not moving too fast , but not too slow.
Some of Apple’s chips/chiplets could be a god fit for where Intel is going , but probably not all of them.
he just took apple to heights nobody could have dreamed of before, but you for sure know it better 😂
This move is an obvious political ass kiss imo. The government invests in Intel and suddenly Apple wants to use Intel again. Gee, I wonder why.
Leaving or threatening to leave? quick search turns up some very new Gurman rumblings within last day or so. Is there something older than that? Gurman's rumors seems to be more like 'match this competitive offer I got and I'd like some new terms'.
That would start to smell a bit like the new designate future CEO is has some friction with folks and folks are not happy.
Srouji also reportedly bristled at chasing more custom chips for glasses/Vision Pro. If Apple puts the dies onto multiple fab processes also then being the head of all of this looks more like 'management" (lots more people spread over lots more 'mature' projects) than of "management of technology".
Depends. If Apple Silicon is going for bigger and broader set of products they may need a better person to manage the set up and more lower very good tech leads over the subset of areas.
For example, someone like Jim Keller couldn't run a place like AMD or Intel well. That is not his wheelhouse.
Part of Apple's past problem is that they promoted Jony Ive into a high level executive and handed him too much stuff. ( Peter principle. One promotion too many. ). Ive picked a dubious person to run Human Computer Interaction in part because he himself was a dubious choice to have influence over HCI.
Apple was almost certainly pictched using Intel foundary before Trump was President. (Gelsinger certainly would have done more than just 'pitch to Apple' via the press
Intel CEO Hopes to Win Back Apple by Making Better Chips Than Apple
In a new episode of Axios on HBO shared by MarketWatch, Intel CEO Pat Gelsinger sat down with Axios' Chief Technology Correspondent Ina Fried to discuss various topics, including Apple's transition to its own custom-designed silicon chips across its Mac lineup. When asked if Intel has given up...
www.macrumors.com
Open AI ran out and bought up 40% of all the DDR5 memory. What if they took their drunken sailor spending money and go off and buy up 15-20% of the all the TSMC wafers. Apple would have a volume access problem. The crazy stupid money sloshing around on the AI hype train is disrupting logistics here about as much as the White House is ( The WH is certainly not making things more stable and sane. )
But one and only one supplier than several folks with super deep pockets are throwing money at ... Apple doesn't like that kind of supplier situation. They want some control on pricing and availability. And very often want more than one supplier.
Last edited:
Intel only relatively recently finished 'firewalling' the systems for the fab business so it is technically far more a solely owned subsidiary ( there are some loans/funding they have done for some of the Irish fabs which mean revenue sharings) . The upside with TSMC is that they don't sell any finished chips themselves. Intel would but in another silo. Samsung is similar in that the display folks are a different subsidiary than the phone folks , but they seem to launch look alikes pretty fast. ( so matter of "how opaque" the firewall separate is. )