Great essay. I find myself increasingly annoyed by these CEOs who lack vision and choose short term financial wins over long term strategic wins.
It’s a little reassuring that the government seems to have properly diagnosed the problem. Hopefully they can steer the ship in the right direction with enough speed to save us from catastrophe.
Are there any books you’d recommend on this topic?
«annoyed by these CEOs who lack vision and choose short term financial wins over long term strategic wins.»
In the USA the American Dream means "winners do whatever it takes" and "fuck you! I got mine" and those CEOs have the same mindset as other real americans.
CEO tenure is usually too short to reap the benefits of “long term strategic wins” and there is always the threat of takeovers, so they do what is best in those circumstances, and shareholder give even give them strong incentives to do that.
Good to see someone appreciating how impressive modern microchips really are.
Not sure why you used photos of circuitboards up there. Granted they're more colorful. Because the surface features of CPUs and GPUs can't be seen with the naked eye, or even an optical microscope. Electron microscopes are the only way to get there, and even then the features are indistinct. It's really hard to overstate how tiny the individual features are, or how easy it would be (and is) to fuck them up.
But the whole chain of technologies that enables this is even more mind boggling. "I, Pencil" is closer to banging rocks together than it is to modern microchip design and manufacture.
For one thing it wouldn't even be possible had we not bootstrapped the world, and history's, largest machine out of previous iterations of the same technology. By which I mean the internet, which is not some ephemeral thing that exists in "the cloud", but one giant interconnected machine that spans the entire planet and parts of low earth orbit. One estimate has the total weight of this machine at around 100 million tons, about the weight of all the buildings in Manhattan. Made of countless billions of parts and requiring millions of people to operate (not counting the billions who interoperate with it).
Anyway, that is all to say, the scale goes up as much as it goes down. Each of those tiny transistors is part of a microchip, over a *trillion* of them per chip, and most of those microchips are one small component among billions in this gigantic machine. And every single part of it has to more-or-less work and everybody working on it has to do his job sufficiently or modern civilization ends and a lot of us die as everything dependent on it breaks down and we go back to sending mail by foot (because btw we don't have enough horses to reboot the pony express and you're not getting any gas if it goes down).
Which makes disrupting all that because Taiwan got oopsie blowed up in a fight between the US and China equivalent to a full nuclear exchange in slow motion. Because if we can't replace those chips as they wear out the rot sets in fast, and just to catch up we'd have to rebuild *more* capacity than we had before, and somehow do that while things were falling apart. Which is not going to happen. Mean time between failure in commercial/industrial use for microchips is about 2-4 years. Guess how long it would take to rebuild Taiwan's production capacity.
Either the musical performance is too short, or I read complex analyses like this too slowly...
The headline about the gov purchasing a 10% stake in Intel immediately brought to mind your history of M&A tactics - obviously a solid base to start influencing the behavior of a publicly owned company. A stunning article in many respects, from the awe of the engineering miracle, inconceivable to me, to the shock of such a vulnerable supply chain with each node governed apparently by a single manufacturer, to the cry of my writer's voice, 'it's not just the national security at stake here, it's my digital pen!' This is not the blockchain, this is a chain that could be far too easily breached at any of several critical joints. Normally, I would cringe at the thought of the gov trying to advise a "business" that doesn't have the freakin common sense to know it should court its customers with the passion of a lover...but this isn't "normally", by any means, and the current admin does have a good cohort of business advisors under its wing. If the gamble works, then it might pull the rug out from under the war machine screaming for blood from the other end of the South China Sea. And then DJT might get a jewel for that peacemaker crown of glory that he says he wants. Proof is in the pudding. Well, it's got my blessing, but I was kinda hoping for a little more elaboration on the respective game theory perspectives of the market vs geopolitics, but maybe that's some other article...
I'm a slow reader too on topics I really want to understand; I think a better way to frame it is an "intentional" reader. imprinting information into our minds requires digesting it and skimming quickly just doesn't cut it for retention.
I've been much more impressed with Trump 2.0; the tactics around trade (he seems to have advisors that concretely understand balance of payments dynamics) and strategic interventions in things like Intel are positive signs for addressing structural issues that cannot be swept under the rug forever. you can't take half-measures on these problems, gotta rip the band-aid off.
I did a breakdown on the semiconductor supply chain and ‘commoditizing the complement’ on twitter in 2023; it has aged extremely well. I’ve held these positions ever since. I’ve been a semi respecter for quite a while (per the esoteric diligence in the quoted essay).
takeaway: listen to your good buddy Dmitry when he’s rummaging about in odd areas with quirky analysis.
You are not wrong in this analysis but I would add (and I could be wrong) that the real answer is to make the usa more attractive for these businesses to choose freely as their preferred location. Didn’t the founder of TSMC want his chips to be built here and chose Taiwan instead due to factors such as labor costs, regulatory burden, tax incentives, etc? IIRC he worked for texas Instruments at the time. Maybe we can save Intel with partial nationalization, IF the government can resist playing central planning meddlers, which it rarely can resist, and IF Intel gets their shit together, and IF they can stop bowing to quarterly earnings (which still matter to 90% of their shareholders) but what we really need is for all cutting edge technologies to WANT to be in the USA over the long run, which goes a long way toward truly solving the problem, and is the original reason INTC was in the USA to begin with. And this move doesn’t fix that problem. I do agree with your analysis regarding the extreme importance of cutting edge chip manufacturing to a nation’s sovereignty.
I once heard that the 4yr election cycle with increasing divided parties makes long term investment impossible. A solar plant built by a Dem could be torn down by a Rep and an industrial plant commissioned under a Rep could be canceled by a Dem. The two parties no longer honor long term agreements like that and it's getting increasingly schizophrenic to outside dollars.
«the real answer is to make the usa more attractive for these businesses to choose freely as their preferred location.»
Why would any real american investor want to immobilize their capital in low profitability manufacturing then trading property, stocks, debt is not only much more profitable but also in practice fully backstopped by the Treasury and the Federal Reserve? Private USA investors are putting their money in property and algorithmic trading rather than in chip manufacturing because of that.
«Only five semiconductor manufacturers operate EUV in high-volume production: TSMC, Samsung, Intel, SK hynix, and Micron.»
They are mostly asian and that is because in "the west" it is far more profitable to trade in property and stocks and debt than to make chips, and anyhow manufacturing is reckoned to be much more easily infected by labor unions; asian investors are less reluctant to accept lower returns. However the economy of China-Taiwan has been dominated by trading in property for some decades which has been far more profitable so TSMC is an exception that may not last that long.
Great essay. I find myself increasingly annoyed by these CEOs who lack vision and choose short term financial wins over long term strategic wins.
It’s a little reassuring that the government seems to have properly diagnosed the problem. Hopefully they can steer the ship in the right direction with enough speed to save us from catastrophe.
Are there any books you’d recommend on this topic?
«annoyed by these CEOs who lack vision and choose short term financial wins over long term strategic wins.»
In the USA the American Dream means "winners do whatever it takes" and "fuck you! I got mine" and those CEOs have the same mindset as other real americans.
CEO tenure is usually too short to reap the benefits of “long term strategic wins” and there is always the threat of takeovers, so they do what is best in those circumstances, and shareholder give even give them strong incentives to do that.
Solid analysis of this vital sector 😎
Tyvm 🤝
Good to see someone appreciating how impressive modern microchips really are.
Not sure why you used photos of circuitboards up there. Granted they're more colorful. Because the surface features of CPUs and GPUs can't be seen with the naked eye, or even an optical microscope. Electron microscopes are the only way to get there, and even then the features are indistinct. It's really hard to overstate how tiny the individual features are, or how easy it would be (and is) to fuck them up.
But the whole chain of technologies that enables this is even more mind boggling. "I, Pencil" is closer to banging rocks together than it is to modern microchip design and manufacture.
For one thing it wouldn't even be possible had we not bootstrapped the world, and history's, largest machine out of previous iterations of the same technology. By which I mean the internet, which is not some ephemeral thing that exists in "the cloud", but one giant interconnected machine that spans the entire planet and parts of low earth orbit. One estimate has the total weight of this machine at around 100 million tons, about the weight of all the buildings in Manhattan. Made of countless billions of parts and requiring millions of people to operate (not counting the billions who interoperate with it).
Anyway, that is all to say, the scale goes up as much as it goes down. Each of those tiny transistors is part of a microchip, over a *trillion* of them per chip, and most of those microchips are one small component among billions in this gigantic machine. And every single part of it has to more-or-less work and everybody working on it has to do his job sufficiently or modern civilization ends and a lot of us die as everything dependent on it breaks down and we go back to sending mail by foot (because btw we don't have enough horses to reboot the pony express and you're not getting any gas if it goes down).
Which makes disrupting all that because Taiwan got oopsie blowed up in a fight between the US and China equivalent to a full nuclear exchange in slow motion. Because if we can't replace those chips as they wear out the rot sets in fast, and just to catch up we'd have to rebuild *more* capacity than we had before, and somehow do that while things were falling apart. Which is not going to happen. Mean time between failure in commercial/industrial use for microchips is about 2-4 years. Guess how long it would take to rebuild Taiwan's production capacity.
Donald Trumpf lol
The $10 and $2.7 trillion numbers are eye watering.
There is no world in which the US gov and military would be allowed to fight a conflict/blockade for longer than about 2 days.
Private corporations and lobbies would be inflicting extreme political costs on Congress and the pentagon etc etc.
The INDOPACOM commander would be flipping burgers after a week.
Either the musical performance is too short, or I read complex analyses like this too slowly...
The headline about the gov purchasing a 10% stake in Intel immediately brought to mind your history of M&A tactics - obviously a solid base to start influencing the behavior of a publicly owned company. A stunning article in many respects, from the awe of the engineering miracle, inconceivable to me, to the shock of such a vulnerable supply chain with each node governed apparently by a single manufacturer, to the cry of my writer's voice, 'it's not just the national security at stake here, it's my digital pen!' This is not the blockchain, this is a chain that could be far too easily breached at any of several critical joints. Normally, I would cringe at the thought of the gov trying to advise a "business" that doesn't have the freakin common sense to know it should court its customers with the passion of a lover...but this isn't "normally", by any means, and the current admin does have a good cohort of business advisors under its wing. If the gamble works, then it might pull the rug out from under the war machine screaming for blood from the other end of the South China Sea. And then DJT might get a jewel for that peacemaker crown of glory that he says he wants. Proof is in the pudding. Well, it's got my blessing, but I was kinda hoping for a little more elaboration on the respective game theory perspectives of the market vs geopolitics, but maybe that's some other article...
I'm a slow reader too on topics I really want to understand; I think a better way to frame it is an "intentional" reader. imprinting information into our minds requires digesting it and skimming quickly just doesn't cut it for retention.
I've been much more impressed with Trump 2.0; the tactics around trade (he seems to have advisors that concretely understand balance of payments dynamics) and strategic interventions in things like Intel are positive signs for addressing structural issues that cannot be swept under the rug forever. you can't take half-measures on these problems, gotta rip the band-aid off.
for those interested, here's more insight on BOP accounting and addressing trade dislocations in a meaningful way: https://thedosagemakesitso.substack.com/p/the-bug-strategy-tariffs-labor-and
terrific breakdown; thank you for this
you are very welcome
I did a breakdown on the semiconductor supply chain and ‘commoditizing the complement’ on twitter in 2023; it has aged extremely well. I’ve held these positions ever since. I’ve been a semi respecter for quite a while (per the esoteric diligence in the quoted essay).
takeaway: listen to your good buddy Dmitry when he’s rummaging about in odd areas with quirky analysis.
2023 DD:
https://x.com/BackTheBunny/status/1670701442032877571
You are not wrong in this analysis but I would add (and I could be wrong) that the real answer is to make the usa more attractive for these businesses to choose freely as their preferred location. Didn’t the founder of TSMC want his chips to be built here and chose Taiwan instead due to factors such as labor costs, regulatory burden, tax incentives, etc? IIRC he worked for texas Instruments at the time. Maybe we can save Intel with partial nationalization, IF the government can resist playing central planning meddlers, which it rarely can resist, and IF Intel gets their shit together, and IF they can stop bowing to quarterly earnings (which still matter to 90% of their shareholders) but what we really need is for all cutting edge technologies to WANT to be in the USA over the long run, which goes a long way toward truly solving the problem, and is the original reason INTC was in the USA to begin with. And this move doesn’t fix that problem. I do agree with your analysis regarding the extreme importance of cutting edge chip manufacturing to a nation’s sovereignty.
I once heard that the 4yr election cycle with increasing divided parties makes long term investment impossible. A solar plant built by a Dem could be torn down by a Rep and an industrial plant commissioned under a Rep could be canceled by a Dem. The two parties no longer honor long term agreements like that and it's getting increasingly schizophrenic to outside dollars.
«the real answer is to make the usa more attractive for these businesses to choose freely as their preferred location.»
Why would any real american investor want to immobilize their capital in low profitability manufacturing then trading property, stocks, debt is not only much more profitable but also in practice fully backstopped by the Treasury and the Federal Reserve? Private USA investors are putting their money in property and algorithmic trading rather than in chip manufacturing because of that.
I wrote something similar!
https://kainesianmacro.substack.com/p/chipmaking-capitalism-with-american
«Only five semiconductor manufacturers operate EUV in high-volume production: TSMC, Samsung, Intel, SK hynix, and Micron.»
They are mostly asian and that is because in "the west" it is far more profitable to trade in property and stocks and debt than to make chips, and anyhow manufacturing is reckoned to be much more easily infected by labor unions; asian investors are less reluctant to accept lower returns. However the economy of China-Taiwan has been dominated by trading in property for some decades which has been far more profitable so TSMC is an exception that may not last that long.