1. If it's a remote switch, wouldn't Chinese hackers also be able to potentially block this kind of kill signal? (Unless it's a dead man's switch I suppose).
2. What would be the strategic implications of China getting access to these machines though? I mean if they didn't, they could just make 5 or 8 or 10 nm chips, right?
What is the strategic advantage of the better manufacturing that would make it worth the US gov saying it would bomb those foundries, or installing kill-switches?
> 1. If it's a remote switch, wouldn't Chinese hackers also be able to potentially block this kind of kill signal? (Unless it's a dead man's switch I suppose).
Not necessarily. It could be triggered by, say, a shortwave radio embedded in the machine that listens for a code on secret frequencies.
> 2. What would be the strategic implications of China getting access to these machines though? I mean if they didn't, they could just make 5 or 8 or 10 nm chips, right?
I don't know. I'd say it would probably give them something to reverse engineer. But I'd say the bigger strategic implication of an invasion would be getting access to the TMSC engineers. They could fly them all up to SMIC, and some would probably cooperate (because they'd have to move on to the new reality after a defeat).
Isn't this just installing a really big "Push me!" button for our adversaries? Why do they need to take control of the fabs if they can just blow them up remotely Without taking responsibility? "Oh I guess there was a bug in the self-destruct system..."
That would be a vulnerability of a naive, always on implementation. But I'd think that risk would be significantly mitigated if the self-destruct system has to be armed first.
e.g. China invades Taiwan, someone then immediately presses the "self-destruct enable" button, which cannot be disabled for a long period of time. If the invasion succeeds, someone transmits the self-destruct code. If it fails they don't, the machine stays intact, and someone gets to work disarming the self-destruct.
I think it's plausible such a device could be built. Just make it battery powered and tamper-proof fail-deadly except when the battery runs down. Size the battery for two or three times the time Taiwan could be expected to hold out against an invasion.
Afaik those laser mirror assemblies are fine tuned to the nano meter range and still they need to finetune for a few months on site to account for the environment where the machine will be operating.
All it takes is a bolt or some hardwired mechanism to render a 300 million dollar machine useless.
I don't think people realize just how complicated these machines are. A remote off switch just works faster than the alternative. A steady stream of parts and experts from various locations around the world is required to keep these things operating smoothly.
Without the ASML service contract most would be unusable after a few months. The amount of technology brought to bear here is really only possible because of worldwide cooperation among thousands of vendors and tens of thousands of engineers. These machines are an extension of some pretty impressive institutions (ASML corp, for example), not just metal and glass.
If China were to invade and if Taiwan or the US fought back, we would all be in a world of hurt, economically. Taiwan makes most of the high end logic chips the world uses, and China makes many of the lower tier, but still necessary chips for things like power management, etc. With sources for these devices offline, a lot of economic activity would begin to halt. We really must avoid war at all costs.
China is not going to be able to invade Taiwan, it's too old and too broke to do so. Too old as in its demographics will be 30% over 60 years old in 10 years. Too broke as in the local governments are bankrupt, and the central government is barely hanging on, so much so that it had to issue ultra long bonds in 2 of the last 3 years. Keep in mind that Chinese government differs from other governments in that the other governments source of main revenue is taxes. Whereas the Chinese government has the revenue from all the state owned enterprises which are monopolies in major industries such as banking and oil. so if the Chinese government still can't get enough revenue from those state owned enterprises to cover existing debt payments, it is pretty fucked.
It's also hard to attack an opponent, when the opponent has been preparing for a number of years, with its allies. US, Japan, Australia, India.
I hope you're right but think about it like this: with a high youth unemployment and supposedly more men than women, they can throw away a couple million people.
Also them being a manufacturing powerhouse, I think they may have enough, even if they're at low point in the economic cycle.
It wouldn't be smart and even the invasion were successful, not sure what they'd gain, besides more hate from all of Southeast Asia and maybe Western sanctions.
I have family in Taiwan and I'm told their military isn't very well funded or equipped.
China can't afford to throw away only sons, and have them sink into the bottom of the ocean. Families would revolt pretty quickly. As we saw from the millions of protestors in Hong Kong in 2019, and the violent protests in Shanghai in lockdown 2023, Chinese citizens are capable of grouping together in mass numbers.
China may have the factories for manufacturing, but they need materials and money to keep a war going. That's hard when they are broke, and also supporting other dictatorships, such as Russia's war effort, Iran, North Korea.
Putin can't invade it doesn't make economic and demographic sense to my western values universal sensibilities.
Meanwhile they have regularly brushed millions of deaths under the rug,to stay in power, to save face and just propagandized that away.
Something would be won if a lesson was learned from Ukraine. The west is not universal and a deterring reaction must be tailored to the deterred, not to the deterring.
I'm not so sure I would bet on it. I think China would rather "reunite" without force, but Xi has put his personal imprimateur on "reunification". Nothing gets people to forget about economic anxiety like a good old-fashioned patriotic war. At this point, I think a lot of it is bluffing- better to have the Americans give up on defending Taiwan without a shot being fired. Better to make the Taiwanese think that "reunification" is inevitable and forget about notions of freedom.
If what you say is true then that increases the chance of war soon, it doesn't decrease it.
AFAICT, the Chinese administration believes it must reunify Taiwan; they've backed themselves into a corner with their rhetoric. They haven't because they believe it's better to wait because the chances of it happening militarily or peacefully are greater in the future than they are now.
But if it's going to be harder in the future, they must invade now.
Luckily I don't believe your argument holds.
Taiwan's birth rate is roughly similar to China's, so it doesn't really affect their relative positions.
China's debt is 85% sovereign so it's not a forcing factor either.
As a peer mentioned, China has 12 million unemployed men. That's more than in all of NATO.
> China imports vast quantities of goods, so yes, debt matters.
Yes, external debt matters. However, they don't have much external debt.
And China's imports are decreasing faster than their exports are. And that trend will accelerate because the two biggest are electronics which we are sanctioning and oil which they are rapidly weaning themselves off of.
Meatgrinder hasn't exactly worked for Russia yet. Especially if the Chinese soldiers are all sitting en mass on large, slow moving boats going across a 100 mile strait.
Unless you think the China's plan is to fill the Taiwan straits with so many dead bodies/boats so quickly, that it will create a corpse bridge for tanks to roll on.
> Meatgrinder hasn't exactly worked for Russia yet.
The point is that delaying the war would have made things even worse than they presently are. 2022 wasn't a good time for a war, but it was a better time than any subsequent year.
Its more a glue a state falling apart together with external opposition kitsunagi approach. Taiwan wirhout tsmc is less valuable, same as hong kong is less valuable once owned by the ccp.
Finally there is a level of people you integrate who know how flimsy dictatorships are that a dictatorship cant survive.
> But if it's going to be harder in the future, they must invade now.
That depends on the timeframe outlook. My impression has been they keep the rhetoric up so as not to abandon their claim over Taiwan, but are willing to wait a few lifetimes to get it.
China could invade Taiwan anytime they wanted, and given the current instability in the world, it would probably be the most opportune time. Additionally, Taiwan's defences would easily be annihilated with several strategic hypersonic missile salvos. Taiwan wouldn't know what hit them. And since they're an island, China would prevent all imports until they surrender. As for your argument regarding its allies, no one is going to mess with China and its arsenal of hypersonic nuclear weapons.
Ukraine is still holding its own despite Russia's hypersonic missile prowess.
If you are talking about a blockade, as we've seen from Russia's invasion, Russia's world's second best navy is already powerless against Ukraine's drone attacks and missiles. And US has already warned China on a potential blockade as being very high cost.
That seems like a massive exaggeration. The Russian navy wasn't that great even before the war. I'm not sure they really have anything going for them besides their submarines, even their single aircraft carrier is permanently under repair and won't reenter service any time soon. Even the Royal Navy is probably significantly strong even on its own.
There were a lot of good reasons why Putin shouldn't have invaded Ukraine either. Yet Putin had been making shows of aggression and threats to invade, while securing his political position in Russia for a decade. You can see that Xi had been pursuing similar moves. I'm not sure autocrats see things as reasonably as you do.
Putin thought the war was only going to be for a few days - according to research by the think tank RUSI, when Vladimir Putin began his invasion, he expected to take control of Ukraine within 10 days https://www.iwm.org.uk/history/how-Putin-lost-in-10-days. He couldn't pull out of the war subsequently in fear of losing control and face. Xi is an idiot, but even he can see his potential downfall, given Russia's failure.
I had the same expectations about Russia in February 2022. Too old, too decrepit, not enough soldiers on the border to take Ukraine (200 000 is far from enough, country of that size needs 10x as much occupiers to be safely subdued into submission), not worth to burn all the relations with the West over.
Oh boy, was I wrong and the doomsayers were right. It is tricky to emulate mind of a 70-y.o. wannabe emperor who dismantled all the previous oligarchy in favor of one-man rule.
And that sound like China, too, with the difference being that the sea between China and Taiwan is a very formidable obstacle that just might raise the overall stakes too high.
You would be mostly correct if their leadership was logical and well informed. I am not convinced that has been the case for at least the last 3 years. When your outcome depends on one dude been told something he isn't going to like by the mountain of yes men left under him after purging disagreeable people for years, you end up wrong way more often than is ideal.
I mean, most likely how you called it above will come to pass but the chance of something deranged isn't close enough to zero anymore to ignore it. I would put the change of something deranged at more like 10-15 % but that's just my gut feeling.
Besides, who really cares? China isn't going to invade taiwan for silly companies like TSMC. If it invaded, it would be to take back their territory. What's with the constant artificial media blitz regarding tsmc? China's aim for reunification dates back decades before tsmc existed. Also, it would be nice if 'news' companies stated who ( public or private ) is sponsoring their articles.
> and will likely never be a part of the People's Republic of China.
What's your definition of 'likely'? That taiwan will be a part of the mainland is pretty much a certainty. It's demography, geography and economy. One has 1.5 billion people and the other is 20 million people.
I feel like just chucking a big ol' jug of sand or salt water into the machine would probably be enough to pretty much permanently disable it, let alone something flammable.
It's not like TSMC engineers and technicians would be raptured the first moment, leaving the fabs just standing there deserted and intact.
This really seems like an answer someone gave to a clueless politician asking a stupid question. "Yes senator, there's a 'self-destruct' (eye roll)". Just leave out that the self-destruct is one dude with the fire bucket from the smoker's shelter in the car park.
43 comments
[ 3.3 ms ] story [ 114 ms ] thread2. What would be the strategic implications of China getting access to these machines though? I mean if they didn't, they could just make 5 or 8 or 10 nm chips, right?
What is the strategic advantage of the better manufacturing that would make it worth the US gov saying it would bomb those foundries, or installing kill-switches?
Not necessarily. It could be triggered by, say, a shortwave radio embedded in the machine that listens for a code on secret frequencies.
> 2. What would be the strategic implications of China getting access to these machines though? I mean if they didn't, they could just make 5 or 8 or 10 nm chips, right?
I don't know. I'd say it would probably give them something to reverse engineer. But I'd say the bigger strategic implication of an invasion would be getting access to the TMSC engineers. They could fly them all up to SMIC, and some would probably cooperate (because they'd have to move on to the new reality after a defeat).
I have to imagine a lot of the expertise with the EUV stuff comes from ASML which is a Dutch company, thankfully.
I have to give props to the US gov for getting the CHIPS act going and we're already seeing Fabs built in the US as a result.
This might end up being a Y2K thing, where good preparation reduced the impact.
e.g. China invades Taiwan, someone then immediately presses the "self-destruct enable" button, which cannot be disabled for a long period of time. If the invasion succeeds, someone transmits the self-destruct code. If it fails they don't, the machine stays intact, and someone gets to work disarming the self-destruct.
I think it's plausible such a device could be built. Just make it battery powered and tamper-proof fail-deadly except when the battery runs down. Size the battery for two or three times the time Taiwan could be expected to hold out against an invasion.
All it takes is a bolt or some hardwired mechanism to render a 300 million dollar machine useless.
Without the ASML service contract most would be unusable after a few months. The amount of technology brought to bear here is really only possible because of worldwide cooperation among thousands of vendors and tens of thousands of engineers. These machines are an extension of some pretty impressive institutions (ASML corp, for example), not just metal and glass.
If China were to invade and if Taiwan or the US fought back, we would all be in a world of hurt, economically. Taiwan makes most of the high end logic chips the world uses, and China makes many of the lower tier, but still necessary chips for things like power management, etc. With sources for these devices offline, a lot of economic activity would begin to halt. We really must avoid war at all costs.
It's also hard to attack an opponent, when the opponent has been preparing for a number of years, with its allies. US, Japan, Australia, India.
Also them being a manufacturing powerhouse, I think they may have enough, even if they're at low point in the economic cycle.
It wouldn't be smart and even the invasion were successful, not sure what they'd gain, besides more hate from all of Southeast Asia and maybe Western sanctions.
I have family in Taiwan and I'm told their military isn't very well funded or equipped.
China may have the factories for manufacturing, but they need materials and money to keep a war going. That's hard when they are broke, and also supporting other dictatorships, such as Russia's war effort, Iran, North Korea.
Meanwhile they have regularly brushed millions of deaths under the rug,to stay in power, to save face and just propagandized that away.
Something would be won if a lesson was learned from Ukraine. The west is not universal and a deterring reaction must be tailored to the deterred, not to the deterring.
AFAICT, the Chinese administration believes it must reunify Taiwan; they've backed themselves into a corner with their rhetoric. They haven't because they believe it's better to wait because the chances of it happening militarily or peacefully are greater in the future than they are now.
But if it's going to be harder in the future, they must invade now.
Luckily I don't believe your argument holds.
Taiwan's birth rate is roughly similar to China's, so it doesn't really affect their relative positions.
China's debt is 85% sovereign so it's not a forcing factor either.
And it's not China vs Taiwan straight up. It's China vs Taiwan, US, Japan, Australia, parts of nato.
China imports vast quantities of goods, so yes, debt matters.
> China imports vast quantities of goods, so yes, debt matters.
Yes, external debt matters. However, they don't have much external debt.
And China's imports are decreasing faster than their exports are. And that trend will accelerate because the two biggest are electronics which we are sanctioning and oil which they are rapidly weaning themselves off of.
Unless you think the China's plan is to fill the Taiwan straits with so many dead bodies/boats so quickly, that it will create a corpse bridge for tanks to roll on.
The point is that delaying the war would have made things even worse than they presently are. 2022 wasn't a good time for a war, but it was a better time than any subsequent year.
They could easily turn around and censor any mention of taiwan until everyone forgets about it
Finally there is a level of people you integrate who know how flimsy dictatorships are that a dictatorship cant survive.
That depends on the timeframe outlook. My impression has been they keep the rhetoric up so as not to abandon their claim over Taiwan, but are willing to wait a few lifetimes to get it.
If you are talking about a blockade, as we've seen from Russia's invasion, Russia's world's second best navy is already powerless against Ukraine's drone attacks and missiles. And US has already warned China on a potential blockade as being very high cost.
That seems like a massive exaggeration. The Russian navy wasn't that great even before the war. I'm not sure they really have anything going for them besides their submarines, even their single aircraft carrier is permanently under repair and won't reenter service any time soon. Even the Royal Navy is probably significantly strong even on its own.
Oh boy, was I wrong and the doomsayers were right. It is tricky to emulate mind of a 70-y.o. wannabe emperor who dismantled all the previous oligarchy in favor of one-man rule.
And that sound like China, too, with the difference being that the sea between China and Taiwan is a very formidable obstacle that just might raise the overall stakes too high.
I mean, most likely how you called it above will come to pass but the chance of something deranged isn't close enough to zero anymore to ignore it. I would put the change of something deranged at more like 10-15 % but that's just my gut feeling.
Discussion: https://news.ycombinator.com/item?id=40426843
'ASML and TSMC Can Disable Chip Machines If China Invades Taiwan'
https://news.ycombinator.com/item?id=40426843
Besides, who really cares? China isn't going to invade taiwan for silly companies like TSMC. If it invaded, it would be to take back their territory. What's with the constant artificial media blitz regarding tsmc? China's aim for reunification dates back decades before tsmc existed. Also, it would be nice if 'news' companies stated who ( public or private ) is sponsoring their articles.
What's your definition of 'likely'? That taiwan will be a part of the mainland is pretty much a certainty. It's demography, geography and economy. One has 1.5 billion people and the other is 20 million people.
It's not like TSMC engineers and technicians would be raptured the first moment, leaving the fabs just standing there deserted and intact.
This really seems like an answer someone gave to a clueless politician asking a stupid question. "Yes senator, there's a 'self-destruct' (eye roll)". Just leave out that the self-destruct is one dude with the fire bucket from the smoker's shelter in the car park.