Preferences

3D30497420 parent
> every decision the US takes could be considerate, but as likely also completely random and reversed and bent at any moment in the future.

This my main problem with this investment. I can certainly appreciate the benefit of US government investment to ensure "homegrown" production capabilities. However, this depends a lot on a level of understanding, intelligence, and planning from the US federal government which is monumentally lacking. If no one trusts Intel now, I cannot begin to imagine how anyone would view Intel plus the current US government as more trustworthy.

Just look at the current approach to tariffs as a good example for how current "industrial policy" is being carried out. Unpredictable, vengeful, and declared with little plan or forethought. Why should we expect any differently from other policies?


dpkirchner
Add to this the fact that given the government is now propping up a poor performing company there's no possible reason to try to create a new domestic competitor. If you do and you start becoming successful, the government can just increase their "investment" in your chief competitor and shut you down. It's ludicrous.
201984
The point of the article was that there's never going to be a new domestic competitor. It would take hundreds of billions in investment to get off the ground and build a fab from scratch, and they'd be in an even worse position than Intel is currently. No external customer would want to risk using them, and they don't have the internal demand of x86 to keep going on their own.
UncleOxidant
There are other alterntives to the US Gov taking a 10% stake (and it must be noted that this is for money already distributed to Intel via the CHIPS act, so it's tough to say how the gov having a 10% stake is really helping Intel here, there's no new investment involved and Intel needs ca$h). I agree with the article that there's a significant geopolitical risk to the status quo. I think it would've been better for the president & his economic staff to arrange a meeting of US companies that need foundry services and try to get them to a "come to Jesus" moment re the geopolitical risk to their current operations relying so heavily on Taiwan. The government role here shouldn't be to own Intel (or other companies) but to incentivize those large fab customers (Apple, Nvidia, AMD, Broadcom, etc) to make a significant investment into Intel foundry services. Not sure what those incentives would look like - perhaps tax incentives. There's also the possibility that some kind of 3rd party consortium could be setup to run Intel. This way they're not trying to start from scratch (which would take too long and too much capital), but also allows those other investing companies to have some kind of control at a distance.
BlueTemplar
"Never" is a very strong statement... or do you expect the USA or the need for chips to stop existing in a time frame as short as half a century ?
rickdeckard
I frankly look forward to see whether the US will actually CARE how Intel will conduct its business, instead of simply trying to just reap benefits from it.

Everything can for now be put under the umbrella of "US semiconductor sovereignty", but actually making this happen involves much more strategic planning and investment from the government.

For example, I doubt that Intel has sufficient experience as a foundry to support design-finalization for ARM, they are JUST starting NOW with this.

So who will pay for closing such gaps? Would they force e.g. Apple to use Intel as foundry and swallow all the associated cost, or would they rather accept Apple to source from a TSMC fab (which is built in US for the big customers like Apple and nVidia)

tick_tock_tick
> If no one trusts Intel now, I cannot begin to imagine how anyone would view Intel plus the current US government as more trustworthy.

Because people making these decisions aren't chronically online....

acdha
This is relevant how? You can be the most dyed-in-the-wool old school businessman who only reads physical newspapers and the first thing you’d think of is the way all of your suppliers are reporting unpredictability and your buyers are talking about reduced demand. Businesses like predictability and between the chaotic massive tax hikes and aggressively politicizing the federal reserve, they don’t have a stable environment.
slipperydippery
I actually think that if we're going to tax-privilege capital gains to a huge degree over wages, it's entirely appropriate for the government to claim some small chunk of non-voting, least-privileged (both to avoid conflicts of interest and fucking up the regular equities & debt markets) shares of every corporation over some size, probably with some other implementation details in there (graduated percentage as company revenue size grows or whatever, that kind of thing). I also think we should consider going back to things like having some parts of the defense industry outright government-operated, like we used to do, which these days might include a chip foundry or two. Not saying we should definitely do it, but I think it should be seriously considered.

... this ain't that, though. It's a one-off, not a reliable broadly-applicable policy, and it's not clear what kind of strategy it represents in the bigger picture. I also doubt the ownership structure is as hands-off as I'd prefer, though I admit I've not looked into the details (if there even are details yet—we've had a lot of reporting on things as if they've happened, that then sometimes go on to never actually happen, lately)

[EDIT] I further think it would be better than the status quo to acknowledge that we have an economy dominated by Zaibatsu now, and to use the government to leverage them for public benefit the way the "Asian Tigers" do/have, though I don't think this is that happening, either. I think we're currently picking the worst of three options, of "intentionally use them to their fullest; break them up; do nothing" (we've been on the "do nothing" track so far, having abandoned "break them up" in the '70s).

This item has no comments currently.