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 Chip Supplier Hit by Intel Delay Ahead of 'A17 Bionic' Production in 2023
- Thread starter MacRumors
- Start date
- Sort by reaction score
PowerPC was never ARM it was RISC as ARM is but was not ARM
I heard that nVidia wanted to cancel some production and TSMC said that they could push it out but not cancel it. And oil is down to $90. Are we finally see deflation?
1. Apple doesn't have that much money.
TSMC price tag is $400+ B.
Apple cash holding $40B (got another $100 or so tied up in investments. )
2. Apple could borrow but already have ~$200B in debt devoted to stock buyback, dividends and avoiding taxes. The reason why the debt service is so relatively cheap is that they are borrowing money they really already have just not structured conveniently for lowest taxes and distribution. So they are super low risk bonds because they when they already have money. The risk of default and can't grab the cash in the scramble at the end is very , very , very low.
3. Apple could just go deeper into debt ... why? At the point where debt is getting more expensive and already under intense antitrust scrutiny. These additional problems 'buy' what? Mostly little of real material value.
Apple can't buy TSMC and take its production "private". Apple isn't no where near big enough TSMC's only 'customer'. By a huge margin.
So Apple would have to buy TSMC and then service all ( or most) of TSMC current clients. Apple needs that distraction why?????????????
Apple avoids buying 100B of fab equipment and solely paying for the R&D for bleeding edge tech. Apple pragmatically makes almost nothing. That is a large contributing reason why their margins are so high. "Boring" , low margin stuff they can inflict on their subcontractors while Apple skims off most of the profit. If Apple soaks up those subcontractors then they have to suck up those lumps also.
The part of the issue is Apple's own making. They sometimes like picking one vendor and then funneling all the money into a single source. Do that too long and there are higher dependency issue. Part of the problem is that Samsung ( and Intel) have kludged their leading edge fab processes. If Samsung "4nm" "3nm" was more on track folks would likely be more spread out.
The chip shortage was essentially at medium-complexity nodes, in chips used for things like cars and consumer electronics.Wait. Wasn't there a chip shortage anyway? Or was that only man made to drive prices
N3 is used by the highest end CPU/SoC chips, but companies cannot change their plans on a dime. If Apple (and AMD, and nVidia, and Qualcomm and so on) knew a year ago what the available capacity would be, they would have made appropriate plans, but they cannot simply restructure the next two years' worth of schedule on a dime.
Rubbish, I never said I was planning on getting the M2 Pro when it is finally released.
Majority marketshare where? Aggregated the Windows PC vendors is still the majority. Apple comes out of the single digits in over $1K. Apple is no where on the top 5 list in the Workstation market. Not the majority there and hefty chunck of those are well over the $1K mark.
The CEO has been very very clear that "TSMC cannot be taken over by force" (his words), it depends on a vary large JIT web of suppliers globally, any takeover of Taiwan by China would completely idle those fabs
Switching from 5nm to 3nm isn’t as simple as just shrinking everything down to that size. It requires complex design changes, not something that can be done it 6-7 months.
Problem is no one else except Apple and Intel was looking to use their leading edge nodes this generation, so TSMC was planning their capacity and economy of scale just for them. If Intel pulls out the per unit part cost would rise by quite a bit.
Here's the chance for AMD and nVidia to step up to the plate to impress the gamers out there with a $4,000 RTX 5999 or Radeon 9999 card.
In the next news... "Apple begins construction of its own fabrication factory... starting with mobile chip"
An old friend of mine procured the original contract for providing shopping hand-baskets for Safeway (she said she was the first continental USA manufacturer of plastic hand-baskets (c. 1960/70's)), and stated that merely the "mold" cost her over USD 100K
I can only assume that not having guaranteed contracts would certainly cast doubt on the efficacy of such an investment.
Point is, Intel help to bring
So, you are basically dismissing 15 years of investment in Intel which actually brought the Mac mainstream? Just as you would dismiss all the years of PowerPC? You are acting like CPU's never existed before Apple silicon.
Before Apple switched to Intel, the only places I ever saw a Mac was in a printshop or that rare outcast user who happens to have one. I am sure among the over 100 million active Mac users out there, not all of them are running on Apple Silicon yet. Don't be so dismissive. This industry is unpredictable. Whose to tell what could end up happening 15 years from now.
Just as you are claiming Apple Silicon is based on ARM, remember, Apple bought the IP for Intels 5G modem business which they are planning to put in future iPhones and replace those from Qualcomm.
Well hopefully the lower volume of 3nm chips doesn't drive up the price for Apple.
Wow, INTEL’s chip plans aren’t going well… that’s shocking news! 😂
Hopefully, this does not cause issues to Apple's plan and that in 2023, we get A17 based on 3nm.
In addition to the other arguments against this (mostly that Apple would need to supply not just Apple but all TSMC's other customers, and that's far from Apple's core business), TSMC doesn't treat workers in its Asian plants very well (I say Asian because workers in its Western fabs, like those in the US, are much better-protected), which enables them to deliver products for the prices Apple and others need. Apple's somewhat shielded from this because TSMC is a supplier rather than a subsidiary. If Apple bought TSMC, they'd need to improve worker conditions, which would be very expensive, or have a PR nightmare on their hands.
It's actually already happening. The US National Semiconductor Technology Center plans a collaboration with a new Japanese chip research institution (unnamed in these articles) to produce 2 nm chips by (or as early as—I've seen it reported both ways) 2025. This was announced in June 2022.
US and Japan join forces to start factory for tiny 2-nanometer chips
The new chips would offer a much higher performance while using less power. It could also help ease the chip shortage.
Japan, U.S. to launch R&D for 2-nm chip mass production
Economic and diplomatic chiefs to discuss supply chains at 2+2 meeting Friday
US 'to help Japan' make leading-edge 2nm chips
Player Four has entered the game
Last edited: Tuesday at 8:27 PM
Intel denies this rumor and claims meteor lake is on track for 2023.
In a message to the Verge:
“Thomas Hannaford clarifying to The Verge that not only are they untrue, but that Meteor Lake will actually ship, launch, and be available to consumers in 2023.”
www.theverge.com
In a message to the Verge:
“Thomas Hannaford clarifying to The Verge that not only are they untrue, but that Meteor Lake will actually ship, launch, and be available to consumers in 2023.”
Intel denies Meteor Lake is delayed to 2024, says chips will ‘launch’ in 2023
Just as planned.
www.theverge.com