All of this will be moot if the US does not figure out a way to keep the international EE and CS PhD students educated in elite US universities. The H1B visa is way overdue to make way to some kind of ranking/points-based system instead of the anachronistic lottery system from the 90s.
It will kill 2 birds in 1 stone; it will largely get rid of the fraud, abuse and low-wage job replacement issue raised by the system's critics and will also help retain the best and brightest in the country.
Easy peasy: pay them. There is no shortage of corporate lawyers or managers. Plenty of these corporate folk had strong STEM backgrounds, but moved to where the work was easier and the pay higher. (I know; I was one.) They used to say: "MIT men work for Harvard men." I suppose we an update this to "IIT-Kanpur men work for Harvard women."
If we insist on underpaying a certain class of workers, we have to rely on braceros. The mystical Asiatic math mind is as much a myth as a unique Mexican ability at farm work. And the US ed system is pretty good, except for PoC.
ASML (my employer) is basically the only company that can make commercial EUV machines. There's an outside chance that China may be able to leapfrog ASML by making advances based on some startup's tech somewhere, but it's not likely - it's like betting on the Russians to make a better stealth jet just because they bought the startup that owns that "invisibility cloak" fabric.
The core problem with EUV [ed: which I've personally worked on!] is that, in the most professional jargon I can summon, IT'S REALLY FUCKING HARD. Even ASML with all of our internal access to EUV IP can just barely keep the machines working. It's not because we're doing anything wrong (at least, not wronger than anyone else in any other industry), it's because the tech is so damned sensitive, it's a (professional jargon warning) FUCKING MIRACLE that it works at all.
Even if some spy magically handed China all of our IP tomorrow, it'd still take their best engineers and scientists at least a decade, if not two, to reverse-engineer an EUV machine. It's not that it's technically unfeasible for anyone else to develop EUV. It'd just cost ~$100-200B over a decade-plus, and no one has that kind of scratch sitting around.
[Ed: This isn't like the Russians reverse-engineering nuclear weapons, where you just need to throw enough money and minds at it to invent a bomb you can manufacture on the cheap at relatively low precision. Rather, it's the highest-precision industrial tooling effort _in the history of mankind_. Reverse-engineering EUV from scratch is about 300 separate Manhattan Project-level efforts, and any one of them can take down the entire enterprise.]
So yeah, as things stand today, the "soft blockade" should give the West plenty of time to build up semiconductor capacity, even under the most wildly pessimistic scenario.
>> ... we need to be promoting startups, so that we don’t rely on big old companies to carry the nation.
Noah, just to be clear... the "let's make this sector more dynamic with startups!" suggestion is just way out of whack with the facts on the ground. It's one of the most capital-intensive manufacturing efforts possible. You're basically making a machine that moves atoms into the right places by the _dozens_, without even a fancy supercollider or sci-fi nanomachine.
There really just aren't any plucky grad students toiling away in a university cleanroom somewhere who're just waiting to disrupt the entire industry. No one's going to pull a "Web-3" on this and distribute manufacturing to a hundred different mini-fabs - the cleanroom requirements alone would be a nightmare! [ed: No, seriously, if you merely _mentioned_ that task to half the project managers in our company, you'd give them pre-emptive PTSD.]
As of right now, most startups in the industry are essentially incubators - kind of like how Facebook and Google treated other startups (like Instagram and DeepMind) in their industry. The startups take concepts from the university labs and turn them into proofs-of-concept, at which point the big boys leverage their massive capital advantage to buy them out and scale up.
Any "startup" that actually disrupted the industry in the way you're asking -- IE that doesn't end up as the latest division acquired by the big boys -- has to look like Tesla, instead of trying to be Uber or Twitter and do everything on its own: Government picks a handful of promising "could-be-winners" and just hands them a buttload of cash so that they can make the capital investments necessary to compete with the big boys.
Finally... I just don't see any actually valid strategic reason why we would need to avoid "rely[ing] on big old companies to carry the nation". They carried us in WWII, they carried us in Cold War 1, and to paraphrase the old adage, you go to _cold_ war with the political economy you have, not the one you wish you had.
And to that point... no offense, but it seems like you're injecting your wishes here. I even agree with them! Sure, we all wish we had a more dynamic economy in general, more strongly driven by startups, and a democracy less corrupted by the polluting effects of overly-concentrated corporate power. And "now more than ever" is a cliché for a good reason - it's bad to waste rare opportunities for good reforms.
But if we'd spent all of our mental energy in WWII trying to pump up tech startups so we wouldn't have to rely on "big old companies" to deliver all the tanks and planes, we might have squandered the strategic advantage those companies represented. We might have made the same mistake as Germany and ended up with a vaporware wunderwaffen.
As you just argued in "The Financialization of Tech", that's a mistake we're already at risk of doing. So, in conclusion, I'd say startups can DEFINITELY be a PIECE of the puzzle - we need them as engines of innovation, there's no doubt about that. But beyond that, we need fabs, and lots of them. If a startup pops its head out and is able to show that it can be a Tesla, then great for them, and we should definitely have the government financing on hand to make it happen. But let's not cut the "big old companies" completely out of the picture.
It wasn't all that long ago that Wall St, the City of London et al insisted that manufacturing be offshored to China & call centres to India. Not just because of lower production costs & bigger shareholder returns, but also the now-discredited belief that when China opened up economically, it would open up politically.
That ship has long sailed, first with Tiananmen Square, and now with Xi's dollarshop Maoism displacing Hu's Dengism. The COVID pandemic also exposed the limitations of the "offshore everything" approach. And when the gains from fiscal globalisation end up hogged by the few & don't trickle down to the many...
The last part is simply not true. The last forty years of globalization have seen substantial improvements in the quality of life across the developing world in almost every parameter you can look at - absolute poverty, hunger, child mortality, etc. And while Western countries may have lost manufacturing jobs, which should not be disregarded, consumers have also enjoyed much lower prices of goods and services. As for the Covid-19 disruptions, they are an argument for more balanced and diverse supply chains, not a return to the protectionist past. I am actually surprised that Noah, who I very much respect, would embrace this type of populist approach towards free trade and globalization.
It wasn't Wall St/City of London. It's just basic international economics 101. It was making US companies richer and China appeared to be less of a belligerent actor. I'd say the turning point was arguably Trump deciding to launch a trade war. China throwing its weight in its region didn't bother anyone in the West too significantly. The pandemic led to China's emergence as a global actor through vaccine diplomacy while the world watch Trump try to steal the election in the bastion of democracy.
Since the Ukraine tensions the idea of independence has come to the fore in key areas. In Europe it is energy. In USA key military components. Globalisation is coming to an end. And probably just in time sourcing. How widespread and what sectors other than energy and the military isnt known. It won't be easy to do as there is much to unwind.
Typo: "there’s a general shift away from thinking that economists’ nostrums about free trade and government non-intervention contain are worth listening to" Suspect you started to write "contain anything worth", and then tried to pare back to just "are worth"?
I need to get the disclosure out of the way first; I've owned Intel stock for about 20 yeas now. I would add to Noah's post that in addition to Semtech, there is the Semiconductor Research Corporation (SRC) that has been around since 1982 and supports basic research and is funded by industry members of the consortium (I visited them way back in 1983 when they were just getting started).
The US funds a huge amount of basic research in STEM and there is no reason that it should not be involved in applied research funding and manufacturing as well. Policy makers cannot close their eyes to what other nations have and are doing. However, a lot of what is done by nation states has had mixed results. I well remember back in the 1980s when Japanese industrial policy was greatly feared but those fears ended up misplaced. China may be spending huge amounts of money but one only need look at their failure in the Covid vaccine area to understand that just spending money is not enough.
I am hopeful that the US can move forward with an industrial policy that both makes sense and achieves concrete results. I agree with Noah that this is just the beginning.
Computer chip production is important, but we should not lose focus: with typical chips quantity is far more important than quality. Even 'weapon systems of the future' can do nicely with off-the-shelf components, e.g. this recent demonstration of a smart drone swarm which has nothing special as far as chips go:
Had things been the other way around and it were China with the new node tech and larger manufacturing capability, the West would have still been fine so long as it kept some decent capability.
What does matter? First, keeping capability onshore. Second, new tech like Quantum Computing which promises to revolutionize encryption. Third, being careful of hardware backdoors.
This act is useful, but there are better arguments for industrial policy - and way better reasons to defend Taiwan than TSMC, like say the fate of its 24 million people.
I'd still call myself a neoliberal and a free trade advocate, but there's no question that national security is a blind spot for the more simplistic and Pollyannaish form of that ideology. It's simply not a good idea to be utterly dependent on trade partners for crucial resources unless they're both friendly and right on your border.
I also think that the failure of opening up trade with China to produce political liberalization in that country is an argument for restricting future free trade deals to other democratic countries, or at least making entry into them for authoritarian regimes contingent on political liberalization.
What are the other industries that Congress should focus on in the same way, with strategic value? Energy is the obvious one, we're pretty close to autarkic on fossil fuels but the supply chain for lithium batteries and solar panels is very much not domestic. Rare earths, as a predecessor step to several of these industries.
Education would be a more fraught one, but it's also a sustainable competitive advantage, so probably worth considering. What's on the list of a would-be industrial policymaker?
All of this will be moot if the US does not figure out a way to keep the international EE and CS PhD students educated in elite US universities. The H1B visa is way overdue to make way to some kind of ranking/points-based system instead of the anachronistic lottery system from the 90s.
It will kill 2 birds in 1 stone; it will largely get rid of the fraud, abuse and low-wage job replacement issue raised by the system's critics and will also help retain the best and brightest in the country.
Yep.
Easy peasy: pay them. There is no shortage of corporate lawyers or managers. Plenty of these corporate folk had strong STEM backgrounds, but moved to where the work was easier and the pay higher. (I know; I was one.) They used to say: "MIT men work for Harvard men." I suppose we an update this to "IIT-Kanpur men work for Harvard women."
If we insist on underpaying a certain class of workers, we have to rely on braceros. The mystical Asiatic math mind is as much a myth as a unique Mexican ability at farm work. And the US ed system is pretty good, except for PoC.
I don't think you are aware of reality.
There are plenty of Chinese PhDs who are as good as Americans.
Try looking from another side, ask Professor Arokyaswamy Paulraj.
Noah - Just want to add some context.
ASML (my employer) is basically the only company that can make commercial EUV machines. There's an outside chance that China may be able to leapfrog ASML by making advances based on some startup's tech somewhere, but it's not likely - it's like betting on the Russians to make a better stealth jet just because they bought the startup that owns that "invisibility cloak" fabric.
The core problem with EUV [ed: which I've personally worked on!] is that, in the most professional jargon I can summon, IT'S REALLY FUCKING HARD. Even ASML with all of our internal access to EUV IP can just barely keep the machines working. It's not because we're doing anything wrong (at least, not wronger than anyone else in any other industry), it's because the tech is so damned sensitive, it's a (professional jargon warning) FUCKING MIRACLE that it works at all.
Even if some spy magically handed China all of our IP tomorrow, it'd still take their best engineers and scientists at least a decade, if not two, to reverse-engineer an EUV machine. It's not that it's technically unfeasible for anyone else to develop EUV. It'd just cost ~$100-200B over a decade-plus, and no one has that kind of scratch sitting around.
[Ed: This isn't like the Russians reverse-engineering nuclear weapons, where you just need to throw enough money and minds at it to invent a bomb you can manufacture on the cheap at relatively low precision. Rather, it's the highest-precision industrial tooling effort _in the history of mankind_. Reverse-engineering EUV from scratch is about 300 separate Manhattan Project-level efforts, and any one of them can take down the entire enterprise.]
So yeah, as things stand today, the "soft blockade" should give the West plenty of time to build up semiconductor capacity, even under the most wildly pessimistic scenario.
Addendum: Responding to this quote:
>> ... we need to be promoting startups, so that we don’t rely on big old companies to carry the nation.
Noah, just to be clear... the "let's make this sector more dynamic with startups!" suggestion is just way out of whack with the facts on the ground. It's one of the most capital-intensive manufacturing efforts possible. You're basically making a machine that moves atoms into the right places by the _dozens_, without even a fancy supercollider or sci-fi nanomachine.
There really just aren't any plucky grad students toiling away in a university cleanroom somewhere who're just waiting to disrupt the entire industry. No one's going to pull a "Web-3" on this and distribute manufacturing to a hundred different mini-fabs - the cleanroom requirements alone would be a nightmare! [ed: No, seriously, if you merely _mentioned_ that task to half the project managers in our company, you'd give them pre-emptive PTSD.]
As of right now, most startups in the industry are essentially incubators - kind of like how Facebook and Google treated other startups (like Instagram and DeepMind) in their industry. The startups take concepts from the university labs and turn them into proofs-of-concept, at which point the big boys leverage their massive capital advantage to buy them out and scale up.
Any "startup" that actually disrupted the industry in the way you're asking -- IE that doesn't end up as the latest division acquired by the big boys -- has to look like Tesla, instead of trying to be Uber or Twitter and do everything on its own: Government picks a handful of promising "could-be-winners" and just hands them a buttload of cash so that they can make the capital investments necessary to compete with the big boys.
Finally... I just don't see any actually valid strategic reason why we would need to avoid "rely[ing] on big old companies to carry the nation". They carried us in WWII, they carried us in Cold War 1, and to paraphrase the old adage, you go to _cold_ war with the political economy you have, not the one you wish you had.
And to that point... no offense, but it seems like you're injecting your wishes here. I even agree with them! Sure, we all wish we had a more dynamic economy in general, more strongly driven by startups, and a democracy less corrupted by the polluting effects of overly-concentrated corporate power. And "now more than ever" is a cliché for a good reason - it's bad to waste rare opportunities for good reforms.
But if we'd spent all of our mental energy in WWII trying to pump up tech startups so we wouldn't have to rely on "big old companies" to deliver all the tanks and planes, we might have squandered the strategic advantage those companies represented. We might have made the same mistake as Germany and ended up with a vaporware wunderwaffen.
As you just argued in "The Financialization of Tech", that's a mistake we're already at risk of doing. So, in conclusion, I'd say startups can DEFINITELY be a PIECE of the puzzle - we need them as engines of innovation, there's no doubt about that. But beyond that, we need fabs, and lots of them. If a startup pops its head out and is able to show that it can be a Tesla, then great for them, and we should definitely have the government financing on hand to make it happen. But let's not cut the "big old companies" completely out of the picture.
It wasn't all that long ago that Wall St, the City of London et al insisted that manufacturing be offshored to China & call centres to India. Not just because of lower production costs & bigger shareholder returns, but also the now-discredited belief that when China opened up economically, it would open up politically.
That ship has long sailed, first with Tiananmen Square, and now with Xi's dollarshop Maoism displacing Hu's Dengism. The COVID pandemic also exposed the limitations of the "offshore everything" approach. And when the gains from fiscal globalisation end up hogged by the few & don't trickle down to the many...
The last part is simply not true. The last forty years of globalization have seen substantial improvements in the quality of life across the developing world in almost every parameter you can look at - absolute poverty, hunger, child mortality, etc. And while Western countries may have lost manufacturing jobs, which should not be disregarded, consumers have also enjoyed much lower prices of goods and services. As for the Covid-19 disruptions, they are an argument for more balanced and diverse supply chains, not a return to the protectionist past. I am actually surprised that Noah, who I very much respect, would embrace this type of populist approach towards free trade and globalization.
Globalisation in itself isn't the problem, rather it's the uneven distribution of the gains that's the problem. Here's a graph that sums up the issue:
https://twitter.com/Nicolas_Colin/status/1030150226923974658
It wasn't Wall St/City of London. It's just basic international economics 101. It was making US companies richer and China appeared to be less of a belligerent actor. I'd say the turning point was arguably Trump deciding to launch a trade war. China throwing its weight in its region didn't bother anyone in the West too significantly. The pandemic led to China's emergence as a global actor through vaccine diplomacy while the world watch Trump try to steal the election in the bastion of democracy.
Since the Ukraine tensions the idea of independence has come to the fore in key areas. In Europe it is energy. In USA key military components. Globalisation is coming to an end. And probably just in time sourcing. How widespread and what sectors other than energy and the military isnt known. It won't be easy to do as there is much to unwind.
Hey Noah - can you do a deep dive into Immigration reform in a follow up post?
This relates to CHIPS, sure. But also in general.
Highly skilled, STEM based immigration reform needs, current status and next steps.
Feels like the next relevant topic here - to properly staff the Fabs (among other things).
Part 2? - https://www.politico.com/news/2022/07/31/microchip-immigration-tech-00048242
Always enjoy your articles
Thanks!!
Typo: "there’s a general shift away from thinking that economists’ nostrums about free trade and government non-intervention contain are worth listening to" Suspect you started to write "contain anything worth", and then tried to pare back to just "are worth"?
Thanks! The opposite, actually...
I need to get the disclosure out of the way first; I've owned Intel stock for about 20 yeas now. I would add to Noah's post that in addition to Semtech, there is the Semiconductor Research Corporation (SRC) that has been around since 1982 and supports basic research and is funded by industry members of the consortium (I visited them way back in 1983 when they were just getting started).
The US funds a huge amount of basic research in STEM and there is no reason that it should not be involved in applied research funding and manufacturing as well. Policy makers cannot close their eyes to what other nations have and are doing. However, a lot of what is done by nation states has had mixed results. I well remember back in the 1980s when Japanese industrial policy was greatly feared but those fears ended up misplaced. China may be spending huge amounts of money but one only need look at their failure in the Covid vaccine area to understand that just spending money is not enough.
I am hopeful that the US can move forward with an industrial policy that both makes sense and achieves concrete results. I agree with Noah that this is just the beginning.
Its striking that only 17 Republicans supported this -- meaning that 33 Republicans opposed something that is critical to national interest.
Pretty sure we know who the ultimate winners will be: semiconductor executives.
You must be one of those people who refuses to own a cellphone?!?
Industrial policy = special interests entrenched in government. https://www.google.com/search?q=mohair+subsidy
Computer chip production is important, but we should not lose focus: with typical chips quantity is far more important than quality. Even 'weapon systems of the future' can do nicely with off-the-shelf components, e.g. this recent demonstration of a smart drone swarm which has nothing special as far as chips go:
https://www.msn.com/en-us/news/technology/watch-a-swarm-of-drones-navigate-a-forest-without-crashing/ar-AAWVmxH
Had things been the other way around and it were China with the new node tech and larger manufacturing capability, the West would have still been fine so long as it kept some decent capability.
What does matter? First, keeping capability onshore. Second, new tech like Quantum Computing which promises to revolutionize encryption. Third, being careful of hardware backdoors.
This act is useful, but there are better arguments for industrial policy - and way better reasons to defend Taiwan than TSMC, like say the fate of its 24 million people.
I'd still call myself a neoliberal and a free trade advocate, but there's no question that national security is a blind spot for the more simplistic and Pollyannaish form of that ideology. It's simply not a good idea to be utterly dependent on trade partners for crucial resources unless they're both friendly and right on your border.
I also think that the failure of opening up trade with China to produce political liberalization in that country is an argument for restricting future free trade deals to other democratic countries, or at least making entry into them for authoritarian regimes contingent on political liberalization.
For a history of EUVL, contact me.
What are the other industries that Congress should focus on in the same way, with strategic value? Energy is the obvious one, we're pretty close to autarkic on fossil fuels but the supply chain for lithium batteries and solar panels is very much not domestic. Rare earths, as a predecessor step to several of these industries.
Education would be a more fraught one, but it's also a sustainable competitive advantage, so probably worth considering. What's on the list of a would-be industrial policymaker?
Curing the cost disease of infrastructure projects.
Thought back to this post recently when I saw this at Politico: https://www.politico.com/news/2023/02/14/taiwan-tech-king-pelosi-powerhouse-microchip-industry-00082646
Can't tell, though, whether Morris Chang was trying to persuade the U.S. to do more or to not try at all.