It's not clear from the press release, but the timeline shifted first from 2025, then to 2027 and now to 2030/2031:
> The semiconductor manufacturer originally estimated its first factory would come online in 2025. Then, it bumped the grand opening to 2027. Now, Intel estimates operations will begin at the Ohio One campus somewhere between 2030 and 2031.
The biggest issue is that there isn't sufficient demand for an Intel fab, which is causing them to slow play opening it. Ben Thompson has written at length on Intel's struggles over the years and advocates that the U.S. step in and help guarantee sufficient volume of orders from US customers to make the fab financially work:
> That leaves Intel and the need for native leading edge capacity, and this is in some respects the hardest problem to solve.
> First, the U.S. should engineer a spin-off of Intel’s x86 chip business to Broadcom or Qualcomm at a nominal price; the real cost for the recipient company will be guaranteed orders for not just Intel chips but also a large portion of their existing chips for Intel Foundry. This will provide the foundational customer to get Intel Foundry off the ground.
> Second, the U.S. should offer to subsidize Nvidia chips made at Intel Foundry. Yes, this is an offer worth billions of dollars, but it is the shortest, fastest route to ground the U.S. AI industry in U.S. fabs.
> Third, if Nvidia declines — and they probably will, given the risks entailed in a foundry change — then the U.S. should make a massive order for Intel Gaudi AI accelerators, build data centers to house them, and make them freely available to companies and startups who want to build their own AI models, with the caveat that everything is open source.
> Fourth, the U.S. should heavily subsidize chip startups to build at Intel Foundry, with the caveat that all of the resultant IP that is developed to actually build chips — the basic building blocks, that are separate from the “secret sauce” of the chip itself — is open-sourced.
> Fifth, the U.S. should indemnify every model created on U.S.-manufactured chips against any copyright violations, with the caveat that the data used to train the model must be made freely available.
> > Fifth, the U.S. should indemnify every model created on U.S.-manufactured chips against any copyright violations, with the caveat that the data used to train the model must be made freely available.
One of these things is not like the others.
Isn't that a massive giveaway to a relatively small number of model makers, at the cost of the huge number of copyright holders, including international copyright agreements?
If this was legislation, I would assume that someone bribed and/or horse-traded for it to be tacked on.
In the old testament fanfiction UNSONG, the 1 piece of international rules to survive the complete breakdown of the laws of physics and most nations on earth is copyright. Reminded me of that
Seems like Intel got caught with their gyatt hanging out. That said, Ben's plan seems unworkable: why would any company want to use worse accelerators, with the stipulation that their work is open source? Creators unilaterally have their work thrown down the skibidi toilet–but only if the chips are American? Who is going to check that?
Oh, the $280B in taxpayer dollars that already went to these obscenely rich companies wasn't enough? Nvidia with its crazy profit margins needs more of our taxpayer dollars to line their pockets?
This is so absurd it's getting to the point where it's laughably evil.
No, the US has far simpler and completely free methods to do this. For example, look at how travel works. You can't use any US federal dollars to fly on non-US airlines unless it's impossible to do so (with some other caveats we won't go into here).
What the US should do is simply declare, no US federal dollar can go to any AI chip that's not produced in the US. No US federal dollar can go into running training or inference for any model that isn't using GPUs produced in the US.
That's it. An instant massive market is created. Without taxpayer dollars. Without subsidies.
Obviously it's going to cost more. Just like tarrifs will.
But we can either make things cost a little more or give away massive piles of money while getting maybe nothing in return and have everything cost more anyway.
That plan is a massive government subsidy in exactly the same way. It's just that instead of direct payments, the subsidy is forcing the government to pay more for all its AI chip needs, because it is deliberately limiting the competition. So the US Government pays more for all of its AI chips and training and inference models under your plan than it does at present, and the size of that extra payment? That's your subsidy.
The 2025 date was never realistic. It was bandied to capitalize on the chip shortage of 2021 and 2022 to try to force CHIPS Act through Congress.
The shift from 2027 to 2030 is the first material slippage -- and reflects one or two things (or both!): 14A process technology timeline and/or soft foundry demand from the fabless firms.
TSMC is kind of a miraculous thing, decoupling the manufacture from design, letting anyone come up with a chip they want to build. I would hope they could have competition, but Intel can’t afford to give up the chip market and therefore competes with its own customers. They’re stuck in the trap created by their market dominance.
The biggest issue is that there isn't sufficient demand for an Intel fab, which is causing them to slow play opening it. Ben Thompson has written at length on Intel's struggles over the years and advocates that the U.S. step in and help guarantee sufficient volume of orders from US customers to make the fab financially work:
Source: https://stratechery.com/2025/ai-promise-and-chip-precariousn...