I suppose the question is, how feasible would it be for Intel to start producing ARM parts in its fabs?

Apple isn't going to keep x86 around just for them. But if Intel could actually produce Apple's M- and A-series chips, it might be worthwhile to increase supplier diversity, even if the ramp-up was painful...

Intel used to have the best ARM implementation when they got DEC's StrongARM as part of an acquisition. They tossed it because it conflicted with their milk cow.

If Intel can manage competitive technology node, it's not any more difficult than moving for new TSMC node.

Every time there is new possess, high-end micrarchitectures are redesigned and optimized for the new process.

20-years ago Taiwan had 7.7 ML earthquake.

Similar earthquake today and GIGAFABs go offline for months, new fabs are delayed and major Foxconn facilities are disrupted etc.

It could easily cost a year for Apple, Nvidia, Qualcomm, and many others.

Companies build in Arizona because nothing happens there from a natural disaster point of view, except that it gets hot.

A hail mary for the board room at Intel

Next earnings call: "Alright guys, I know it doesn't look good, what with our process technology still stuck in 2015, but picture this: TSMC's fab gets annihilated by a giant meteor."

The way supply chains go, everyone is hurt if Taiwan (chips, EMS) or South Korea (chips, dram) goes offline for longer periods.

Too big to fail?

Remember all those Intel clones?

Intel crushed everyone now we need to save it?

Do we have another option?

“Only the Paranoid Survive” -Andy Grove

https://www.amazon.com/Only-Paranoid-Survive-Exploit-Challen...

> And that challenge is compounded by the fact that Apple was Intel’s largest customer for premium x86 chipsets, and now Apple has moved to Arm, and x86 chips are not Apple’s future.

I'm skeptical that Apple was Intel's largest customer (I've heard that Apple was only 10% or so of Intel's business). Even if they were the largest customer, Apple wasn't important enough to make Intel bend over backwards.

But it's a flagship customer to say the least.

Just look at the percentage of high value devs using the platform. And count the MacBook Pros in any given good CS program.

I don't doubt that having your processor in developer workstations is helpful for driving use of your server processors.

I read this expecting to mock the argument but Bajarin is a smart guy.

He barely touches on perhaps the strongest argument: "partnering" with Intel (i.e. for all intents and purposes controlling, even if it doesn't have financial control) could be used to fend off antitrust attacks by the US government by transforming Apple into a national savior.

Likewise partnering with the weakest of the big three, Chrysler (perhaps Kia was a stalking horse for that?)

Perhaps following the same approach in Europe (although Intel has a large share of Europe's fab capacity already) could have the same value. Perhaps cars in Europe, chips in the USA.

I think the smart guy was Jobs.

He was the great integrator. music, movies, chips, even the gorilla glass thing. Roped in all these people.

This is seven habits stuff: dependent -> independent -> interdependent

Steve Jobs got Apple into a big web of interdependencies where everyone moved forward. Had the whole valley involved.

Now Apple's regressed back to independent.

> Now Apple's regressed back to independent

I don't really understand why you say this. For example, gorilla glass is an illustration of exactly what I mean: Apple controlled the supply and drove corning to both change its formula and ramp up production unexpectedly for Apple. Apple happily paid for this. It was driven by Cook.

Likewise Apple buying up a huge percentage of milling machines for aluminium and sending them to their suppliers.

Apple's supply chain is multinational. They get heavily involved with vendors/suppliers, not simply Hon Hai and TSMC. They are willing to work with competitors (e.g. Samsung).

They have made the switch to USB Type C both in USB (they are on the USB committee and had a lot to do with the design of Type C even though they were also working on lightning) and Thunderbolt (which they cooperated with Intel on).

They are quite far from "independent"

I think they've improved in the "here's my design, make 10,000,000 copies" sense.

I do not thing they've interoperated with other companies well at a high level. Thunderbolt was all due to jobs - he got intel to do it (andy grove?). It's now back in intel's court now.

Instead they bring more and more in house. Their hardware can't run windows anymore. Their software has gotten less and less flexible. They let other languages rot, it's obj c and swift. Their hardware has gone back to making g4 cube type mistakes. again. and again. Ask anyone about the ports or the keyboard on an apple laptop and you'll get an earful.

Just my opinion.

If i were intel I would be seriously considering taking RISC-V and running with it.

ARM still lags, way, way behind in software - i.e. where is vTune for my ARM server? Intel are good at that kind of thing, they could kill ARM on the server before it reaches puberty.

Really Apple should never Help Intel and Intel needs to spin off their foundry division and let that Division provide its services to a larger market, provided that entity can get back into the leading edge node race.

TSMC paid for it's R&D via spreading that costs amortization around an entire industry of fabless processor designers and really that's more economical for the future.

And Samsung and TSMC are already looking to build leading edge foundries on US soil. Apple's got no logical reasons for helping Intel for Intel's lack of proper investments geared toward the future. Really Intel reaps what Intel has sewn and some of that has held back innovation to the detriment of progress on more than one occasion!

Apple has worked with mainland China and Taiwan governments without much difficulty, and they seem to have no problem operating there.

TSMC isn’t the only supplier capable of producing better chips than Intel, Samsung is also another available option if TSMC were to have issues.

To be quite honest, I’m not sure Intel has anything of value to offer Apple. If TSMC fails to deliver they’d sooner start their own fabrication business than go back to Intel.

Is it even unlikely that Apple one day moves the entire company’s main headquarters to China? If you think about it, it’s very likely that this will one day make a lot of business sense.

In a way this article is floating protectionism to patch the plot holes in the story of American Exceptionalism. Perhaps it can be argued that the US Government’s neoliberal approach did this to itself.

In the neoliberal world, anything but the highest level of efficiency is determined to be the loser. Perhaps that could be why the US continually fails to protect whatever special sauce it might have through outsourcing and subcontracting.