This was kind of expected due to the shutdowns required by most countries during the chaos of COVID. Of course we need our infrastructure to be more robust, but no use crying over spilt milk.
There have been calls for some time to treat semiconductor manufacturing as in the interest of national security. The funding and policy isn't onerous to do so, and "never let a crisis go to waste." Something similar was already done for rare earths [1].
I would call it technically correct but useless only because of how ill defined and exploitable national security is as a notion to bypass all objections by turning everything into an existential threat. Everything is in some degree important to national security - entertainment is a matter of national security because bored soldiers are less vigilant.
Something similar was "thought of" for rare earths but I don't think anything practical has come of that yet. Opening new mines in the US is not a process I'd call quick. It's more in the "I'll believe it when I see it" column. You think getting a pipeline across the Midwest is difficult? Just try strip-mining in Alaska.
But yeah, if the government is serious about securing supply chains, then it's not enough to have semiconductor manufacturing here, one also needs to be able to source the inputs closer to home (quartz, etc).
And I'm not even sure it requires government to directly fund the manufacturing/mining - it would be enough if government took a "not-hostile" attitude towards the environmental and engineering approvals, and then incentivized domestic sourcing.
The relevant part is kind of buried but tl;dr: new mines do take a while (but are in the works). Luckily, there are old mines that can be productively restarted:
> “Our mission as a company is to restore the full supply chain,” said James Litinsky, CEO of MP Materials, the new owner of the U.S.’s only rare-earth mine, located in Mountain Pass, in California’s Mojave Desert.
> “We are probably further ahead than people realize.”
> MP Materials restarted the mine in 2017; the previous owner, Molycorp had declared bankruptcy a few years prior.
> “It was mismanaged for some time,” Litinsky said. Mountain Pass’ reserves are particularly rich in rare earths and now supply, in unpurified form, 15% of the rare earths consumed globally each year.
How about you spend some of that cash hoard and do it yourself? I would rather the government start dealing with cyber attacks because that's going to affect the east coast something fierce if this crap keeps going.
If the government gives money then it better talk to these companies about where their items are produced.
My exact reaction as well. Or how about all that money used for stock buy backs over the last decade? Intel especially. It's been mismanaged and now it's putting its hand out for government money while having spent billions on stock buy backs. No more but of course it'll happen because the politics of it. Socialism for the rich and corporations, capitalism for the rest of us.
Stock buy backs are just more tax efficient dividends for investors. If the company is suffering from mismanagement, dividends are good, because it returns capital to investors rather than letting managment destroy it.
Intel has been building fab capacity pretty much always, it's just that their 10nm process doesn't work as intended, and they haven't really been able to fix it, so their development is stalled and their production numbers aren't great and they haven't been able to stop making processors at 14nm to do other things with those fabs. So far, I think we've been hearing of delays on their 7nm node as well, so no good news there.
That said, if there's a market for a 14nm or more fab in the US, Intel has shown they can build that, but they'd probably prefer to spend their money getting 7nm to work than building an old tech fab; spending other people's money on a new 14nm (or whatever) fab in the US is still good for Intel though, so it's no wonder they'd put their hand out to do that.
Stock buybacks end up being nothing other than a transfer of wealth from taxpayers to shareholders because anytime something happens the government steps in bails them out because they're too big. Now everyone knows that so there's no incentive for companies to act responsibly. The fact that it's now accepted and normalized is incredibly disheartening to say the least. No government subsidies. Let the capitalists figure it out. Otherwise let's just nationalize everything instead of privatizing profits and socializing losses.
> How about you spend some of that cash hoard and do it yourself?
Because they are for-profit entities with shareholders interest as top priority, and it is much more profitable to use TSMC/Samsung for them. With subsidies, the situation may change.
The reason these items are not produced at home is because of government policies.
I agree that they should be the ones to invest in these fabs, but it also makes sense for the government to make sure that is an internationally competitive move for them to make, rather than forcing them to pour water uphill.
Right. Other governments are always going to subsidize though (unless you work that into the trade deals I suppose).
Obviously any barriers to the success of US fabs should be removed, but it's not clear if there's a market for more expensive US chips - if there were, you'd think someone would build it without the need for a subsidy
I've said it before: there's lots of low-hanging fruit in semiconductor manufacturing. Even ignoring the present shortage, it's been the case for a while that there are gains to be had for a competent tech giant with deep enough pockets to take on manufacturing themselves.
Too many outsiders tend to err on giving the benefit of the doubt, thinking that there's some sort of Chesterton's fence (you can see this in the comments here), but it's just not true. So much of what causes delays that I've seen firsthand comes down to unbelievably inefficient business processes plus the less-than-stellar pool of candidates that the industry overwhelmingly prefers to hire from.
Didn't know that! I always search for links before submitting and many times I've seen a previous submission with little to no traction (and then subsequently not posted it to avoid dupes).
Thanks for searching first! If it's a good article and hasn't had attention yet, we certainly hope you will post it - that's why we allow reposts, to give good articles multiple cracks at the bat.
protip: the funds are in on-shore banks under offshore business names. this isn't controversial, just something many don't understand as readily available and commonplace.
I didn't make any statement on that and you are assuming that the lack of signaling means acceptance.
I said onshore banks open accounts for many entity types that are not limited to just humans or just businesses from the 50 states.
The prior poster assumed a bank in another country was necessary, which can be very difficult to open and maintain for a US citizen owner of an offshore company.
> Some of the world’s biggest chip buyers, including Apple Inc, Microsoft Corp and Alphabet Inc’s Google, are joining top chip-makers such as Intel Corp
> to create a new lobbying group to press for government chip manufacturing subsidies
In terms of fairness to these companies, yeah. They're rich and it sucks to give them our tax dollars.
But in terms of maintaining the competitiveness and security of the US, seems like they do.
As a rule, I don't care if giving someone money is fair. I care about whether it causes a general increase in the welfare of our society.
Even as a non-recipient of these subsidies, I might be better off if they existed if they reduce risk to the supply chain I'm downstream of as a consumer. If it's unfair to me but makes me better off, I'm not going to cry over the unfairness.
(Maybe tariffs on imported chips are a better solution, but I suspect that's more politically fraught, and solutions that work are better than solutions that don't.)
> As a rule, I don't care if giving someone money is fair.
Sure
> I care about whether it causes a general increase in the welfare of our society.
But by what metric do you measure that? There has been a tech boom in the last couple of decades but the inequality divide has been growing day by day.
But economics works on incentives. Problem is these incentives are discontinuous and can change on a dime. You can accelerate a change by aligning incentives.
Chip manufacturing is important for products, but its also a security risk. The government should put up support to incentivize these companies to redirect capital towards US manufacturing. Otherwise the incentives for these companies are not going to align on a timeframe that is acceptable to employees and shareholders.
Now we can make all kinds of normative claims about how this should play out, but I honestly don’t have the energy to entertain that kind of conversation anymore.
> A push back against globalisation and more domestic production in each region of the world.
There are many countries that have always wanted to be self reliant but the powers be can't allow it. Coz it's sets a bad example that countries can exist without certain giants. I don't think they are about to allow that
For damned good reason. "Self reliant" countries tend to behave like complete assholes internationally even by the standards of Realpolitik. North Korea is one of the most
Autarkous nations on earth.
The "victimized by a conspiracy" anthromorphization is also deeply wrong. They cannot be oppressed - they do not have rights.
Governments are fundamentally made of coercion. Without it they are an empty pile of words, a rules card for poker when the deck is being used for blackjack or building houses of cards. The coercion may be used to support rights like "not being murdered" but granting it rights to further human rights is like trying to increase a number by multiplying a negative number by a grearer than one positive number.
To sum it up if you apply the "corporation as a sociopath" conception governments are basically so alien to be cthulu. Corporations are at least made of people essentially, at least until the first AI run no human owner corporation is founded.
The same discourse was made before WW1, and we all know how that turned out. As far as I could tell the world was even more globalised back then, at least for the elites.
> As far as I could tell the world was even more globalised back then
It was not. How could it be? Cheap and fast long range transportation, the Internet, mass consumerism,interconnected financial systems...
What could possibly make you think the world was more globalised in the 1910s than 2020s?
And yes, many people said a war is impossible due to the globalisation. I guess what they missed was telling it to the people in charge ( many of whom,like the Kaiser and Tsar didn't want a war, but felt pressured and like it was inevitable).
European economies were fairly integrated, though. Not as much as today, but enough that the war caused a massive disruption to everyone involved.
Europe of 1914 was actually much more liberal than people tend to think. With the exception of Russia and Turkey, which required passports from visitors, you could travel passport-free around the entire Europe, settle down and work wherever you wanted.
There were many railway lines crossing the borders and cargo flows among countries were fairly heavy. The social elite was used to sending their kids abroad to study languages and spending vacations in fashionable foreign resorts.
If there was enough of a concentration of power then those things don't matter as much, but even then they'd be considered deterrents to war. Given we mostly have democratically elected governments now war will be much harder to sell.
> What could possibly make you think the world was more globalised in the 1910s than 2020s?
For example borders were a lot more porous in the late 1800s - early 1900s compared to the 2020s, Ellis Island could have never happened today, when you have children from El Salvador or Guatemala kept in cages for basically doing the same thing that children from Calabria or Sicily were doing back then.
Globalization is not what happened during the last few decades. Investing your entire supply chain in China is not globalization. Corona is causing a push back against sinicization.
If the world actually did have globalized production, the pandemic, which originated in China, would probably have hurt relatively less.
> Intel Corp to create a new lobbying group to press for government chip manufacturing subsidies.
Assholes. They don't need subsidies to produce chips competitively in America (since they already do) and Intel is one of the reasons the industry is so reliant on TSMC.
I'd much rather see this money go to founding a domestic TSMC competitor than to Intel. Or even to TSMC to have them build plants in North America.
Fabless design is here to stay, so it behooves the country (and world) to have enough designless fabricators to ensure sufficient manufacturing capacity.
Yes, let's subsidize a domestic TSMC competitor and pay for it by increasing corporate tax rates / closing loopholes. This is not a snarky comment btw.
I mean you only need to get one agency head, or one head of state / cabinet member, or the majority of congress who votes in unison right now to agree, once
If you want the Bezos's of the world to pay more tax, then lobby for that directly. Using the corporate tax as a proxy for "screw rich people" is distortionary and ineffective.
Actually, they do. You pay half of the payroll tax and the company pays the other half. This is payroll tax and not income tax, that is Social Security and Medicaid. This is why if you're self employed, there is a self employment tax.
That said, I think that the parent is making a different point. I think the parent's point is that if the owners and employees of a company are paying taxes, then taxing the profit of a company taxes twice, once for the profit and then again for the income that comes from the dividend.
Because they are freed from the liability of the business's actions by way of the state allowing liability limiting corporate entities to exist. When the business assumes liability from the owners, it also assumes tax responsibility. If you don't want to pay corporate taxes, don't form a corporation.
Apple and the likes need more chips to make more money and they have big coffers. Instead of public subsidizing those fat cats let them take care of that pesky money issue. They can afford it.
Wealthiest entities on Earth lobbies government for handouts.
I'm not even being snarky, this is just a lobby group trying to socialize costs to avoid spending their own money. Explicitly: “Government should refrain from intervening as industry works to correct the current supply-demand imbalance causing the shortage,”. The industry is not stripped for cash at all, so if the government is not expected to steer, why should it put up the people's money?
The worst part is that they might get such handouts. I don't understand how, given their mighty fortunes, they haven't achieved 99.9% automation to keep costs as low as in cheap labour countries. If anything I think the EU, US and UK should subsidise research in achieving near full automation. Sucks for low skilled labour but they’ve already been fucked when jobs were moved overseas.
If I understand the situation correctly, the current president is already willing to enact such a law. The lobbying seems to be about pressuring congress into actually passing it. Fair enough, everyone is entitled to lobby for their interests. What I take umbrage with is that they want to turn subsidies into handouts.
It's beyond ironic to call for subsidies but then label oversight as "intervening". Subsidies are interventions. Without a government plan and oversight they are just legalized corruption.
This is just the PR spiel we're seeing. What they're really saying is the usual: Local cost is too high to be competitive, subsidize or lose jobs. It's a naked power play (or pragmatic statement of fact, depending on your viewpoint). They're just framing it in a way they think will fit the current, protectionist mood.
What are the geopolitical consequence of a de-coupled chip supply chain? Recall TSMC is headquartered in Taiwan, which has been a point of contention in US-China relationship since the 1950s. Please discuss civilly.
Same reason Foxconn is a proponent for manufacturing in America. If the shift is happening and you can't control it, may as well hedge by being part of it.
On one hand, this could cause other countries to pay more attention to the situation, since -they- will still be dependent on TSMC and may not have the US doing all the posturing.
But that's assuming we wound up truly 'de-coupled'. TSMC tends to be the leader in fab tech, and I am fairly certain any near term plans will involve providing incentive for TSMC to build more fabs in the US.
I think the main question is how much of the US support for Taiwan comes from the chip supply chain, vs ensuring that china needs american permission to trade
In game theoretic terms - If you feel like the only leverage you have over your opponent is about to be obviated by some other approach (i.e. domestic chip fabrication becomes dominant within the decade), then there could be some increase in tensions.
My implication here being that China is aware that they are capable of making an unstoppable strike against all TSMC's leading edge fabrication capacity, and that this action would be far more harmful to the strategic positioning of the US than to any other nation on earth. The ability to execute this option is time-limited by the US wishing to grow domestic capacity.
Whether this implication holds in reality is something I am very much open to suggestion on. I am fully aware that the US military's response to such an action is probably the leading reason it is not something we have seen occur already. I am also unsure if the impact of all 7nm node loss is substantially greater for the US vs any other country.
To all those saying that those companies can afford it themselves, remember that there are reasons they don't. The economics don't add up. Companies tend to do what is in their economic interest to do.
Something like this can change the economics, which could be a good thing.
It's only a monopoly if the government says you are using market position anti-competitively and/or against the common good. If the US government says that chip production capability is now a national focus, companies can expect government cooperation for a joint venture.
Can you please elaborate? I struggle to discern the ultimate effect of private capital versus government funds in this situation. Ideally, a business’ problems are handled by itself, not via the federal government using taxpayer monies.
In a vacuum, sure. But these companies are competing with foreign governments using taxpayer monies to make a product.
Cheaper just to buy from the foreign government? sure. But if a product is considered core to survival it may be good to have domestic production. Whether it's food, weapons, computer chips, water supply, etc.
The basic argument is that domestic production of integrated circuits isn't economical for American businesses. Having a more flexible supply chain is beneficial to these companies (e.g. when a global pandemic hits), but the required investment outweighs these benefits.
On the other hand, the government potentially has a lot to lose if domestic companies are completely unable to produce hardware needed for critical infrastructure. The government also benefits from these companies expanding their supply chain to include domestic production, so the companies are asking the government to provide enough capital/tax breaks to make domestic production viable, and they're trying to convince the government that this provides enough value to the USA for it to be worth it.
But this is being pushed by lobbyists paid for by giant corporations, so it's fair to be skeptical. In reality, the $50B number probably sits somewhere between "the bare minimum needed to make this viable" and "a handout pocketed by greedy corporations".
Obviously there is a difference in that private capital invests in what will help that particular company, while government funds should go toward what benefits the nation as a whole. A company might say "we won't bother investing, let our competition invest, and we'll just get the components from whoever will sell them to us cheapest." That can make sense individually (and especially in the short term), but as a whole (whether talking about industry as a whole, or nation as a whole), they are worse off.
Essentially, a prisoner's dilemma.
There is also the fact that American companies are competing against Chinese companies, where the government can subsidize various Chinese companies and initiatives. China is communist, of course, which does sort of favor this sort of thing. If the US was to close itself off to China, then it's a lot easier to be 100% capitalist as in your ideal scenario.
>> To all those saying that those companies can afford it themselves, remember that there are reasons they don't.
Automotive chips for example tend to be produced on older nodes, not fancy 14nm and below at TSMC. I'm not talking about SoC used in infotainment but all the micro controllers all over the car.
There is really zero incentive for a chip maker to expand capacity of old nodes since they are obsolete in some sense. It doesn't pay to upgrade to a new node without a product that demands something from the new node, and even then it's still an obsolete node. Think 130nm, 90nm, or even 45nm, the world could use more capacity at all of those but would you really want to invest heavily in any of those?
I suspect the end game for most of the industry is commodity priced chips at every "last node" where that term means the last node that doesn't require X, and X is a technology change like immersion lithography, multi-patterning, material change, or EUV. With that kind of environment nobody can afford to increase capacity much.
I'm not arguing for government funding here, just pointing to what I suspect is one of the problems.
I used to work for OnSemi, so take the following for what you will.
On sells a variety of components -- some are completely commodity and are produced in super cheap fabs in Malaysia. Some are mid-tier and are (or were) made at fabs in Phoenix or Pocatello. And some are high-end and come out of the Gresham fabs that they bought when LSI went fabless.
There's value in producing the commodity stuff. Obviously they're not going to enter a bidding war with TSMC for cutting edge manufacturing equipment to make commodity voltage regulators. But they still invest in fab technology at the lower end.
I don't think ONN is particularly special in this space. They had plenty of competitors and they worked to distinguish themselves on cost, depth of their product catalog, and ability to execute.
By having in-country fabs, have you just pushed the problem to some sort of precursor materials or parts?
For that matter, I wonder how a government-subsidized fab could compete with an old overseas fab making lower end parts that has been long-ago paid for?
It doesn't matter what the economics adds up to when the government is printing trillions of dollars and spending it on 'what it thinks needs to happen', whereupon trillionaire companies can lobby for some of that money.
Perhaps these companies should create a fund to start a few foundries etc. because they can definitely afford it.
> Companies tend to do what is in their economic interest to do
It will be in their economic interest when the PLA will show up at TSMC's door. Of course, that won't probably happen in the next few quarters, so on that you are correct, it's not in their short economic interest.
What these companies are basically doing is asking the US government to subsidise the costs incurred by geo-politics.
"To all those saying that those companies can afford it themselves, remember that there are reasons they don't."
This sounds a lot like "God moves in mysterious ways and has a plan for everyone, everything happens for a reason". No actual analysis and reasoning required.
This free market religion is what gave us 2008 and will keep giving.
Ah yes, the free market utopia every libertarian dreamed of existed in 2008. There were only just over 100,000 pages of federal law and regulations. Not to mention state and local laws.
I see, if we had 0 pages of regulations, then banks would be ethical, then they would not create CDOs and then synthetic CDOs and destroy globaln economy.
Rather than advocate for a specific conclusion, I'd like to propose a way of looking at this problem:
What matters is a society's capacity to produce. Not whether something is "fair", not whether an industry deserves this or that. Not whether some cherished economic theory is being honored.
Specifically, what matters is the creation and support of productive ecosystems, which are networks of credit, knowledge acquisition and transfer, skilled workers, skilled product managers, and supplier/vendor/retailer relationships.
A key part of these productive ecosystems is jobs. If you want to learn how to build bridges, you do not set up schools to teach engineering, you fund the building of bridges, and people will figure out how to build them on the job, and the demand for civil engineers will pull students to study this and then universities will open up programs to cater to that demand. The jobs come first. Only 5% of China's population attends university, yet they are able to build all the infrastructure they need because they have a laser like attention to creating jobs. They will even build bridges for America, as long as Chinese workers get to make them. They will build a port for anyone who wants it, as long as Chinese workers are the ones building the port. The entire Belt and Road initiative is an attempt to import infrastructure jobs by sending workers all over the world to build infrastructure, as long as China gets to build it. They know that half these projects will default and not make any money, but what they get out of that is a skilled workforce, and with a skilled workforce they can do anything. They will let American companies set up factories in China as long as there is a knowledge transfer as part of the deal. That all their state owned enterprises are losing money is not important, the acquisition of skills and the creation of productive ecosystems is what matters. I wonder when the US will realize this.
Western economic thinking is focused on P&L, rule of law, etc, and assumes as an article of faith that in such an environment, productive ecosystems will just arise all on their own, like rats being created from piles of trash. It will just happen, because in the past it just happened. So we are focused on abstract principles, but what the last 30 years has shown us is productive ecosystems being destroyed right and left as production migrates over to Asia.
Apparently these abstract principles don't work in all cases, so people are going to need to abandon their faith in free markets just as much as they need to abandon their concerns over some fat cat getting a subsidy -- unless they want their nation to completely deindustrialize.
However abandoning our faith in what worked before still leaves as muddied the concerns of what will work now, because it's not so clear that Asia's economic model can be successfully copied. Latin America tried to adopt this model in the 1970s -- a policy of import substitution, production subsidies and taxes on consumption. But it didn't work. Similar policies were tried in Africa in the 1980s and they also didn't work. So apparently you need more than just a policy of subsidizing production. What is that X-factor?
One proposal is that the X-Factor is human capital. Whether cultural or genetic or literary, for some reason China has human capital for which these policies work well but in Latin America/Africa they don't. Another proposal is that the X-Factor is exports - it's not enough to just subsidize production you need to subsidize exports as well. A third proposal is that the X-Factor is state organizational capacity (a kind way of saying "less democracy, more centralization"). A fourth is nationalism -- being willing to sacrifice some of your own wealth for the benefit of the nation, rather than say boosting your bonus by outsourcing. And last, there is the notion of practicality. The West is often quite fanatical in holding to beliefs that no l...
I don't know how you can justifiably call Africa a has-been. Africa is by far the youngest continent in the world, with 60% of the population below 25. The median age is 19. The fertility rate is incredibly high. Most likely, Africa is the continent of the future.
I wonder what the "strings attached" will be. Perhaps the back doors the FBI has always wanted? Sounds like a double fleecing of the taxpayer. We get to give them subsidies and they get to give law enforcement an easier vector to spy on us. What a deal!
This crosses into nationalistic flamewar. We don't want that on HN, regardless of $nation. It leads to predictable, nasty discussion that kills the curious conversation which this site is supposed to be for.
The demand to always be curious and never angry is itself a limitation on the natural unfolding of curiosity. Sometimes the most interesting questions can only be asked by people who are aware of their own anger and the mutual anger of others around them.
I might question that a bit, but it's an empirical question that one could talk about. I don't think we've ever asked people not to get angry, though - that would be not to be human. The request is to not allow it to drive one into internet comments in a predictable way. That's also a lot to ask, but at least it's doable.
Can anger lead one to explore and learn things that wouldn't have happened without the anger? Sure it can, and comments coming from that place are more likely to be on topic here. But that's already a later stage of a pretty complex process. I'm talking about the split-second moment that we all experience when a flash of anger first arises, where we're reacting adversely to something we've heard before and are propelled to respond immediately with something we've said before. A lot of internet comments, including HN comments, are generated in that state, and they aren't distinguished by curiosity.
> Some of the world’s biggest chip buyers, including Apple Inc, Microsoft Corp and Alphabet Inc’s Google, are joining top chip-makers such as Intel Corp to create a new lobbying group to press for government chip manufacturing subsidies.
> “Government should refrain from intervening as industry works to correct the current supply-demand imbalance causing the shortage,” the group said.
I feel like in about 4 years time there's going to be a massive bust in chip prices from massive oversupply based on the plans for chip manufacture around the world?
I agree, there's going to be a massive oversupply of chips. If the US, EU, China, and also the current chip producers all increase output that's way too many chips.
I recently watched a youtube video predicting a semiconductor bust that https://www.youtube.com/watch?v=Z7QkIECEkVc, though I don't quite understand if the differing say <7nm versus mid range 14~28 nm chips will have differing oversupply in the future.
I think this is a healthy perspective. Certainly, the F22 becomes quite a challenging thing to build if all chip supplies dry up. Priorities finally seem to be adjusting accordingly.
Is this for US companies only? I mean if all they want is Fabs in US surely the funding should be available for TSMC as well?
Otherwise why should TSMC invest billions in US while having US government handing out money to American companies competing against TSMC with their own money?
I don't want to seem politically ignorant but isn't the entire existence of Taiwan is simply dependent on the US wanting it to exist. If the US didn't provide billions in defense -- TMSC would be a Chinese FAB pretty fast. I see your point but I'd imagine the extra layer of geopolitics is propping up this odd situation.
TSMC would blow their fabs up if China invaded. China might not mind though since some the us government has forced TSMC to not sell to some Chinese companies. Probably would be good for their semi industry.
All: this thread got a flash flood of angry-predictable-obvious comments. Please don't post those! They're really bad for this site. (Edit: if you want to see the comments I'm talking about, they're collapsed at the bottom of the page now. Before that, they were the page.)
The issue with such points isn't that they're wrong, it's that they're not interesting. Nothing predictable and obvious is. We want interesting conversation here, and that comes from curiosity. If you're feeling agitation rather than curiosity, please wait until that polarity flips.
The interesting things in a story like this are any diffs from what one would have expected.
It's a significant move, so there must be interesting diffs. But sometimes one has to hunt for those, or at least wait for them to occur to one. That sort of waiting is key to getting good HN discussion, which is reflective rather than reflexive.
This is what bringing back supply chain looks like. I am not against it. What is happening is not great, but you know the whole necessary evil thing.. this is it.
This will start with Tech, but will extend to other domains.
The gas pipeline shut down is currently showing this to be an issue for refined fuel. The hurricane that squatted on top of Houston a couple of years ago also shows that to be true as well. We also saw that in toilet paper as well.
297 comments
[ 3.8 ms ] story [ 255 ms ] thread[1] https://www.google.com/search?q=rare+earth+us+national+secur...
But yeah, if the government is serious about securing supply chains, then it's not enough to have semiconductor manufacturing here, one also needs to be able to source the inputs closer to home (quartz, etc).
And I'm not even sure it requires government to directly fund the manufacturing/mining - it would be enough if government took a "not-hostile" attitude towards the environmental and engineering approvals, and then incentivized domestic sourcing.
The relevant part is kind of buried but tl;dr: new mines do take a while (but are in the works). Luckily, there are old mines that can be productively restarted:
> “Our mission as a company is to restore the full supply chain,” said James Litinsky, CEO of MP Materials, the new owner of the U.S.’s only rare-earth mine, located in Mountain Pass, in California’s Mojave Desert.
> “We are probably further ahead than people realize.”
> MP Materials restarted the mine in 2017; the previous owner, Molycorp had declared bankruptcy a few years prior.
> “It was mismanaged for some time,” Litinsky said. Mountain Pass’ reserves are particularly rich in rare earths and now supply, in unpurified form, 15% of the rare earths consumed globally each year.
If the government gives money then it better talk to these companies about where their items are produced.
Intel has been building fab capacity pretty much always, it's just that their 10nm process doesn't work as intended, and they haven't really been able to fix it, so their development is stalled and their production numbers aren't great and they haven't been able to stop making processors at 14nm to do other things with those fabs. So far, I think we've been hearing of delays on their 7nm node as well, so no good news there.
That said, if there's a market for a 14nm or more fab in the US, Intel has shown they can build that, but they'd probably prefer to spend their money getting 7nm to work than building an old tech fab; spending other people's money on a new 14nm (or whatever) fab in the US is still good for Intel though, so it's no wonder they'd put their hand out to do that.
~Montgomery Burns...probably
Because they are for-profit entities with shareholders interest as top priority, and it is much more profitable to use TSMC/Samsung for them. With subsidies, the situation may change.
I agree that they should be the ones to invest in these fabs, but it also makes sense for the government to make sure that is an internationally competitive move for them to make, rather than forcing them to pour water uphill.
Which policies? I believe you, genuinely curious
Obviously any barriers to the success of US fabs should be removed, but it's not clear if there's a market for more expensive US chips - if there were, you'd think someone would build it without the need for a subsidy
Often trade deals do include agreements about subsidies.
> it's not clear if there's a market for more expensive US chips - if there were, you'd think someone would build it without the need for a subsidy
They wouldn’t be more expensive with the subsidy.
Yes but it's complicated when you're dealing with China where the govt is heavily involved in many "private" cos
Too many outsiders tend to err on giving the benefit of the doubt, thinking that there's some sort of Chesterton's fence (you can see this in the comments here), but it's just not true. So much of what causes delays that I've seen firsthand comes down to unbelievably inefficient business processes plus the less-than-stellar pool of candidates that the industry overwhelmingly prefers to hire from.
Hah - one can hope. But really, is there a chance that could happen? I'm not sure if there's any precedent.
https://news.ycombinator.com/item?id=27116601
2. Lobby government for subsidies
3. Laugh all the way to the off shore bank
I said onshore banks open accounts for many entity types that are not limited to just humans or just businesses from the 50 states.
The prior poster assumed a bank in another country was necessary, which can be very difficult to open and maintain for a US citizen owner of an offshore company.
> to create a new lobbying group to press for government chip manufacturing subsidies
None of those companies need subsidies.
Commoditize your complement. To some approximation and with few exceptions, hardware is everybody's complement.
But in terms of maintaining the competitiveness and security of the US, seems like they do.
As a rule, I don't care if giving someone money is fair. I care about whether it causes a general increase in the welfare of our society.
Even as a non-recipient of these subsidies, I might be better off if they existed if they reduce risk to the supply chain I'm downstream of as a consumer. If it's unfair to me but makes me better off, I'm not going to cry over the unfairness.
(Maybe tariffs on imported chips are a better solution, but I suspect that's more politically fraught, and solutions that work are better than solutions that don't.)
Sure
> I care about whether it causes a general increase in the welfare of our society.
But by what metric do you measure that? There has been a tech boom in the last couple of decades but the inequality divide has been growing day by day.
But economics works on incentives. Problem is these incentives are discontinuous and can change on a dime. You can accelerate a change by aligning incentives.
Chip manufacturing is important for products, but its also a security risk. The government should put up support to incentivize these companies to redirect capital towards US manufacturing. Otherwise the incentives for these companies are not going to align on a timeframe that is acceptable to employees and shareholders.
Now we can make all kinds of normative claims about how this should play out, but I honestly don’t have the energy to entertain that kind of conversation anymore.
They acknowledge the importance of the issue, but are not willing to take risks in a difficult an foreign (to them) business.
So, mostly worthless words.
There are many countries that have always wanted to be self reliant but the powers be can't allow it. Coz it's sets a bad example that countries can exist without certain giants. I don't think they are about to allow that
The "victimized by a conspiracy" anthromorphization is also deeply wrong. They cannot be oppressed - they do not have rights.
Governments are fundamentally made of coercion. Without it they are an empty pile of words, a rules card for poker when the deck is being used for blackjack or building houses of cards. The coercion may be used to support rights like "not being murdered" but granting it rights to further human rights is like trying to increase a number by multiplying a negative number by a grearer than one positive number.
To sum it up if you apply the "corporation as a sociopath" conception governments are basically so alien to be cthulu. Corporations are at least made of people essentially, at least until the first AI run no human owner corporation is founded.
It was not. How could it be? Cheap and fast long range transportation, the Internet, mass consumerism,interconnected financial systems...
What could possibly make you think the world was more globalised in the 1910s than 2020s?
And yes, many people said a war is impossible due to the globalisation. I guess what they missed was telling it to the people in charge ( many of whom,like the Kaiser and Tsar didn't want a war, but felt pressured and like it was inevitable).
Europe of 1914 was actually much more liberal than people tend to think. With the exception of Russia and Turkey, which required passports from visitors, you could travel passport-free around the entire Europe, settle down and work wherever you wanted.
There were many railway lines crossing the borders and cargo flows among countries were fairly heavy. The social elite was used to sending their kids abroad to study languages and spending vacations in fashionable foreign resorts.
For example borders were a lot more porous in the late 1800s - early 1900s compared to the 2020s, Ellis Island could have never happened today, when you have children from El Salvador or Guatemala kept in cages for basically doing the same thing that children from Calabria or Sicily were doing back then.
I don't see why becoming some sort of Juche state would help anyone.
If the world actually did have globalized production, the pandemic, which originated in China, would probably have hurt relatively less.
Assholes. They don't need subsidies to produce chips competitively in America (since they already do) and Intel is one of the reasons the industry is so reliant on TSMC.
I'd much rather see this money go to founding a domestic TSMC competitor than to Intel. Or even to TSMC to have them build plants in North America.
Fabless design is here to stay, so it behooves the country (and world) to have enough designless fabricators to ensure sufficient manufacturing capacity.
7 figures max
That said, I think that the parent is making a different point. I think the parent's point is that if the owners and employees of a company are paying taxes, then taxing the profit of a company taxes twice, once for the profit and then again for the income that comes from the dividend.
I'm not even being snarky, this is just a lobby group trying to socialize costs to avoid spending their own money. Explicitly: “Government should refrain from intervening as industry works to correct the current supply-demand imbalance causing the shortage,”. The industry is not stripped for cash at all, so if the government is not expected to steer, why should it put up the people's money?
It's beyond ironic to call for subsidies but then label oversight as "intervening". Subsidies are interventions. Without a government plan and oversight they are just legalized corruption.
On one hand, this could cause other countries to pay more attention to the situation, since -they- will still be dependent on TSMC and may not have the US doing all the posturing.
But that's assuming we wound up truly 'de-coupled'. TSMC tends to be the leader in fab tech, and I am fairly certain any near term plans will involve providing incentive for TSMC to build more fabs in the US.
My implication here being that China is aware that they are capable of making an unstoppable strike against all TSMC's leading edge fabrication capacity, and that this action would be far more harmful to the strategic positioning of the US than to any other nation on earth. The ability to execute this option is time-limited by the US wishing to grow domestic capacity.
Whether this implication holds in reality is something I am very much open to suggestion on. I am fully aware that the US military's response to such an action is probably the leading reason it is not something we have seen occur already. I am also unsure if the impact of all 7nm node loss is substantially greater for the US vs any other country.
Something like this can change the economics, which could be a good thing.
No, its a monopoly if you have market power.
It’s only illegal if you are abusing market power.
The fact they are holding it as cash is an indication they are not bullish on the fundamentals.
Cheaper just to buy from the foreign government? sure. But if a product is considered core to survival it may be good to have domestic production. Whether it's food, weapons, computer chips, water supply, etc.
On the other hand, the government potentially has a lot to lose if domestic companies are completely unable to produce hardware needed for critical infrastructure. The government also benefits from these companies expanding their supply chain to include domestic production, so the companies are asking the government to provide enough capital/tax breaks to make domestic production viable, and they're trying to convince the government that this provides enough value to the USA for it to be worth it.
But this is being pushed by lobbyists paid for by giant corporations, so it's fair to be skeptical. In reality, the $50B number probably sits somewhere between "the bare minimum needed to make this viable" and "a handout pocketed by greedy corporations".
Essentially, a prisoner's dilemma.
There is also the fact that American companies are competing against Chinese companies, where the government can subsidize various Chinese companies and initiatives. China is communist, of course, which does sort of favor this sort of thing. If the US was to close itself off to China, then it's a lot easier to be 100% capitalist as in your ideal scenario.
Automotive chips for example tend to be produced on older nodes, not fancy 14nm and below at TSMC. I'm not talking about SoC used in infotainment but all the micro controllers all over the car.
There is really zero incentive for a chip maker to expand capacity of old nodes since they are obsolete in some sense. It doesn't pay to upgrade to a new node without a product that demands something from the new node, and even then it's still an obsolete node. Think 130nm, 90nm, or even 45nm, the world could use more capacity at all of those but would you really want to invest heavily in any of those?
I suspect the end game for most of the industry is commodity priced chips at every "last node" where that term means the last node that doesn't require X, and X is a technology change like immersion lithography, multi-patterning, material change, or EUV. With that kind of environment nobody can afford to increase capacity much.
I'm not arguing for government funding here, just pointing to what I suspect is one of the problems.
On sells a variety of components -- some are completely commodity and are produced in super cheap fabs in Malaysia. Some are mid-tier and are (or were) made at fabs in Phoenix or Pocatello. And some are high-end and come out of the Gresham fabs that they bought when LSI went fabless.
There's value in producing the commodity stuff. Obviously they're not going to enter a bidding war with TSMC for cutting edge manufacturing equipment to make commodity voltage regulators. But they still invest in fab technology at the lower end.
I don't think ONN is particularly special in this space. They had plenty of competitors and they worked to distinguish themselves on cost, depth of their product catalog, and ability to execute.
For that matter, I wonder how a government-subsidized fab could compete with an old overseas fab making lower end parts that has been long-ago paid for?
Perhaps these companies should create a fund to start a few foundries etc. because they can definitely afford it.
It will be in their economic interest when the PLA will show up at TSMC's door. Of course, that won't probably happen in the next few quarters, so on that you are correct, it's not in their short economic interest.
What these companies are basically doing is asking the US government to subsidise the costs incurred by geo-politics.
Not necessarily, maybe those companies can just make a deal with the CCP to keep access to TSMC's fabs.
This sounds a lot like "God moves in mysterious ways and has a plan for everyone, everything happens for a reason". No actual analysis and reasoning required.
This free market religion is what gave us 2008 and will keep giving.
What matters is a society's capacity to produce. Not whether something is "fair", not whether an industry deserves this or that. Not whether some cherished economic theory is being honored.
Specifically, what matters is the creation and support of productive ecosystems, which are networks of credit, knowledge acquisition and transfer, skilled workers, skilled product managers, and supplier/vendor/retailer relationships.
A key part of these productive ecosystems is jobs. If you want to learn how to build bridges, you do not set up schools to teach engineering, you fund the building of bridges, and people will figure out how to build them on the job, and the demand for civil engineers will pull students to study this and then universities will open up programs to cater to that demand. The jobs come first. Only 5% of China's population attends university, yet they are able to build all the infrastructure they need because they have a laser like attention to creating jobs. They will even build bridges for America, as long as Chinese workers get to make them. They will build a port for anyone who wants it, as long as Chinese workers are the ones building the port. The entire Belt and Road initiative is an attempt to import infrastructure jobs by sending workers all over the world to build infrastructure, as long as China gets to build it. They know that half these projects will default and not make any money, but what they get out of that is a skilled workforce, and with a skilled workforce they can do anything. They will let American companies set up factories in China as long as there is a knowledge transfer as part of the deal. That all their state owned enterprises are losing money is not important, the acquisition of skills and the creation of productive ecosystems is what matters. I wonder when the US will realize this.
Western economic thinking is focused on P&L, rule of law, etc, and assumes as an article of faith that in such an environment, productive ecosystems will just arise all on their own, like rats being created from piles of trash. It will just happen, because in the past it just happened. So we are focused on abstract principles, but what the last 30 years has shown us is productive ecosystems being destroyed right and left as production migrates over to Asia.
Apparently these abstract principles don't work in all cases, so people are going to need to abandon their faith in free markets just as much as they need to abandon their concerns over some fat cat getting a subsidy -- unless they want their nation to completely deindustrialize.
However abandoning our faith in what worked before still leaves as muddied the concerns of what will work now, because it's not so clear that Asia's economic model can be successfully copied. Latin America tried to adopt this model in the 1970s -- a policy of import substitution, production subsidies and taxes on consumption. But it didn't work. Similar policies were tried in Africa in the 1980s and they also didn't work. So apparently you need more than just a policy of subsidizing production. What is that X-factor?
One proposal is that the X-Factor is human capital. Whether cultural or genetic or literary, for some reason China has human capital for which these policies work well but in Latin America/Africa they don't. Another proposal is that the X-Factor is exports - it's not enough to just subsidize production you need to subsidize exports as well. A third proposal is that the X-Factor is state organizational capacity (a kind way of saying "less democracy, more centralization"). A fourth is nationalism -- being willing to sacrifice some of your own wealth for the benefit of the nation, rather than say boosting your bonus by outsourcing. And last, there is the notion of practicality. The West is often quite fanatical in holding to beliefs that no l...
Please read comments you reply to.
China maintains economic well being by donating the chip market and also the supply chains.
The Quad has emerged to dislocate Chinese tyranny in the Indo-Pacific, and US chip takeover will be one of the last strokes in China’s downfall.
https://news.ycombinator.com/newsguidelines.html
Can anger lead one to explore and learn things that wouldn't have happened without the anger? Sure it can, and comments coming from that place are more likely to be on topic here. But that's already a later stage of a pretty complex process. I'm talking about the split-second moment that we all experience when a flash of anger first arises, where we're reacting adversely to something we've heard before and are propelled to respond immediately with something we've said before. A lot of internet comments, including HN comments, are generated in that state, and they aren't distinguished by curiosity.
> “Government should refrain from intervening as industry works to correct the current supply-demand imbalance causing the shortage,” the group said.
What do they think "intervening" means?
I recently watched a youtube video predicting a semiconductor bust that https://www.youtube.com/watch?v=Z7QkIECEkVc, though I don't quite understand if the differing say <7nm versus mid range 14~28 nm chips will have differing oversupply in the future.
Otherwise why should TSMC invest billions in US while having US government handing out money to American companies competing against TSMC with their own money?
I don't want to seem politically ignorant but isn't the entire existence of Taiwan is simply dependent on the US wanting it to exist. If the US didn't provide billions in defense -- TMSC would be a Chinese FAB pretty fast. I see your point but I'd imagine the extra layer of geopolitics is propping up this odd situation.
The issue with such points isn't that they're wrong, it's that they're not interesting. Nothing predictable and obvious is. We want interesting conversation here, and that comes from curiosity. If you're feeling agitation rather than curiosity, please wait until that polarity flips.
The interesting things in a story like this are any diffs from what one would have expected.
https://hn.algolia.com/?dateRange=all&page=0&prefix=false&so...
It's a significant move, so there must be interesting diffs. But sometimes one has to hunt for those, or at least wait for them to occur to one. That sort of waiting is key to getting good HN discussion, which is reflective rather than reflexive.
https://hn.algolia.com/?dateRange=all&page=0&prefix=true&sor...
This will start with Tech, but will extend to other domains.
The gas pipeline shut down is currently showing this to be an issue for refined fuel. The hurricane that squatted on top of Houston a couple of years ago also shows that to be true as well. We also saw that in toilet paper as well.