"From 2001-2020, Intel blew $128 billion on buybacks (64% of net income) on top of paying out $68 billion as dividends (35% of net income),” notes Lazonick. That’s money that couldn’t go into innovation, retaining and training employees, R&D, and other critical areas.
Intel had a massive R&D budget and 110,000 employees circa 2020. (Vs 12k and 14k for AMD and NVIDIA, both of which have higher market caps than Intel now).
It wasn't a money problem, it was a badly run company on all fronts. When it comes to R&D the sheer magnitude of the money they squandered is legendary. They lost billions on Larrabee, billions on mobile chips, helped invent phase change memory and even made drives with it, and then dumped it for nothing, the list goes on. It's bad management all the way down, and impressively so, considering that many of these initiatives were the right products for the right time just with a horrible execution (massively parallel compute/GPUs launched Nvidia to a 4 trillion dollar valuation, mobile chips are a huge industry, etc). Arguably they needed to streamline and downsize well before they did. They were full on old school IBM level bloat.
The last few years before their implosion was indeed MBA style running the company into the ground (they had a particularly awful CEO during that period), but that was just the capstone on a large decline, during which Intel had massive budgets that they squandered.
>From 2001-2020, Intel blew $128 billion on buybacks (64% of net income)
Is that supposed to be a lot? Sure, hindsight is 20/20 and now we know they should have spent more on R&D, but what would be the correct amount? 50% 100%?
"blew" isn't accurate because stock buybacks aren't spending money. That's kind of like saying contributing to an investment account is spending money.
You can sell the stock again, or use it for employee compensation.
No, they are spending money, there is no substantial difference between dividends and buyback here since the companies issue their own stock, so they don't need to buy anything to sell it, they can try to get more money from investors in both cases.
Investment account analogy fails because you're not investing in your own ownership, so there is no "circular" relationship with yourself
Selling treasury stock has the same effect of diluting external shareholders, so the safety thing is the same - it depends on their assessment of the underlying reality.
It doesn’t matter if it will work or not, this is something that some large U.S. companies have been prepping for, the slow process of starting to move manufacturing to the U.S., or at least to play games to make it look like it, if tariffs can be avoided.
Personally, as much as I accept that the top chips were made in Taiwan, I don’t think it’s impossible in theory for top chips to be made in the U.S.- but it would require massive changes. Starting with the culture. We’d need to turn off the T.V., work longer hours, and homeschool kids for 15 hours per day with the world’s best educational resources. And even when other countries’ kids and parents do that, they still don’t succeed.
But, I think China will end up being self-sufficient, not needing the rest of the world. In a few years, they’ll be charging an arm and a leg for AI. The rest of the world won’t be able to compete with because we don’t have adequate energy production. They’ll also acquire the ability to make the chips they need.
We could just give it up, and I think that’s where most of us are- watching fascism slowly take hold, knowing we’re fucked and the next thing we’ll see is news of a full-scale global war.
In the meantime, maybe we’ll start making shitty products like Britain did in the mid-20th century.
TSMC already manufactures chips in the US, much of their value add comes from European companies like ASML and Zeiss, and the fundamental technology EUV was invented in the US.
> and homeschool kids for 15 hours per day with the world’s best educational resources
Taiwan doesn't have the world's best educational resources - what language do you think they're written in?
Btw, when they did try to do this, they ended up with a national epidemic of myopia and had to make the kids go outside.
> Taiwan doesn't have the world's best educational resources - what language do you think they're written in?
That’s not what they were saying. They were saying that in order to compete globally, the U.S. should ramp up education, which is valid, because Asians have dominated in higher education, which is needed to have first-class R&D, which is needed to lead in manufacturing and product development. The reason the U.S. excels is that they both outsource that to China and other countries as well as import top talent from China and other countries, but that has slowed and will slow.
> Btw, when they did try to do this, they ended up with a national epidemic of myopia and had to make the kids go outside.
You’re arguing for and against yourself. What point are you trying to make?
The US doesn't outsource product development to China. Final assembly isn't product development. And in Taiwan's case TSMC is basically their only highly-advanced company like this, so I don't know if I find it very convincing. Samsung and Intel are not really that far behind either.
> You’re arguing for and against yourself.
No such thing as for or against, just things to be aware of.
Didn't Biden sign a 280 billion dollar Intel subsidy for chip-making? I mean as far as I can tell we already do it, this isn't a fantasy, and there's no reason to think we can't absorb more of the market. I can't read the article but it seems more political than economic, judging from the first paragraph
From the CHIPS act. It allocated 280 billion, and was criticized for being an Intel bailout since many thought it was a response to their poor market performance. It was signed during the Biden admin, before trump entered office
But it wasn't $280 billion for Intel. Why are you conflating the entire amount with the amount put towards Intel?
Your statements are easily shown to be false. Intel didn't receive anywhere near that amount. It's a really poor way to try to engage with people unless your intent is to mislead.
The discussion was that this article calls Trump's push for chip making a "fantasy", while Biden made his own push years ago. The chips act was 280 billion, and it was widely viewed as a bill to help Intel, who is/was struggling. Nothing I said was misleading, and you're attempting to misconstrue what was said to win an internet argument with a person you'll never meet in a thread no one else will read.
And no, Biden didn't sign a "280 billion dollar Intel subsidy for chip-making." If you're talking about the CHIPS and Science Act, that number is overall authorized funding for many things. Intel has received ~10b until now.
Also:
"From 2001-2020, Intel blew $128 billion on buybacks (64% of net income) on top of paying out $68 billion as dividends (35% of net income),” notes Lazonick. That’s money that couldn’t go into innovation, retaining and training employees, R&D, and other critical areas.
https://www.ineteconomics.org/perspectives/blog/america-need...
Intel had a massive R&D budget and 110,000 employees circa 2020. (Vs 12k and 14k for AMD and NVIDIA, both of which have higher market caps than Intel now).
It wasn't a money problem, it was a badly run company on all fronts. When it comes to R&D the sheer magnitude of the money they squandered is legendary. They lost billions on Larrabee, billions on mobile chips, helped invent phase change memory and even made drives with it, and then dumped it for nothing, the list goes on. It's bad management all the way down, and impressively so, considering that many of these initiatives were the right products for the right time just with a horrible execution (massively parallel compute/GPUs launched Nvidia to a 4 trillion dollar valuation, mobile chips are a huge industry, etc). Arguably they needed to streamline and downsize well before they did. They were full on old school IBM level bloat.
The last few years before their implosion was indeed MBA style running the company into the ground (they had a particularly awful CEO during that period), but that was just the capstone on a large decline, during which Intel had massive budgets that they squandered.
>From 2001-2020, Intel blew $128 billion on buybacks (64% of net income)
Is that supposed to be a lot? Sure, hindsight is 20/20 and now we know they should have spent more on R&D, but what would be the correct amount? 50% 100%?
"blew" isn't accurate because stock buybacks aren't spending money. That's kind of like saying contributing to an investment account is spending money.
You can sell the stock again, or use it for employee compensation.
No, they are spending money, there is no substantial difference between dividends and buyback here since the companies issue their own stock, so they don't need to buy anything to sell it, they can try to get more money from investors in both cases.
Investment account analogy fails because you're not investing in your own ownership, so there is no "circular" relationship with yourself
> since the companies issue their own stock, so they don't need to buy anything to sell it,
Public companies diluting shareholders generally causes people to flee for safer investments.
A company is not a person. It doesn't always own 100% of itself.
Selling treasury stock has the same effect of diluting external shareholders, so the safety thing is the same - it depends on their assessment of the underlying reality.
And the person doesn't own himself, he is himself
It doesn’t matter if it will work or not, this is something that some large U.S. companies have been prepping for, the slow process of starting to move manufacturing to the U.S., or at least to play games to make it look like it, if tariffs can be avoided.
Personally, as much as I accept that the top chips were made in Taiwan, I don’t think it’s impossible in theory for top chips to be made in the U.S.- but it would require massive changes. Starting with the culture. We’d need to turn off the T.V., work longer hours, and homeschool kids for 15 hours per day with the world’s best educational resources. And even when other countries’ kids and parents do that, they still don’t succeed.
But, I think China will end up being self-sufficient, not needing the rest of the world. In a few years, they’ll be charging an arm and a leg for AI. The rest of the world won’t be able to compete with because we don’t have adequate energy production. They’ll also acquire the ability to make the chips they need.
We could just give it up, and I think that’s where most of us are- watching fascism slowly take hold, knowing we’re fucked and the next thing we’ll see is news of a full-scale global war.
In the meantime, maybe we’ll start making shitty products like Britain did in the mid-20th century.
I agree that the cultural difference is the biggest differentiator. Asians put more emphasis on education and trading off work life balance.
>In a few years, they’ll be charging an arm and a leg for AI.
In my humble opinion, AI is only useful as a mechanism of oppression, so the fewer countries able to use it against their people, the better.
TSMC already manufactures chips in the US, much of their value add comes from European companies like ASML and Zeiss, and the fundamental technology EUV was invented in the US.
> and homeschool kids for 15 hours per day with the world’s best educational resources
Taiwan doesn't have the world's best educational resources - what language do you think they're written in?
Btw, when they did try to do this, they ended up with a national epidemic of myopia and had to make the kids go outside.
https://www.theguardian.com/world/2025/mar/01/shortsighted-t...
> Taiwan doesn't have the world's best educational resources - what language do you think they're written in?
That’s not what they were saying. They were saying that in order to compete globally, the U.S. should ramp up education, which is valid, because Asians have dominated in higher education, which is needed to have first-class R&D, which is needed to lead in manufacturing and product development. The reason the U.S. excels is that they both outsource that to China and other countries as well as import top talent from China and other countries, but that has slowed and will slow.
> Btw, when they did try to do this, they ended up with a national epidemic of myopia and had to make the kids go outside.
You’re arguing for and against yourself. What point are you trying to make?
The US doesn't outsource product development to China. Final assembly isn't product development. And in Taiwan's case TSMC is basically their only highly-advanced company like this, so I don't know if I find it very convincing. Samsung and Intel are not really that far behind either.
> You’re arguing for and against yourself.
No such thing as for or against, just things to be aware of.
Didn't Biden sign a 280 billion dollar Intel subsidy for chip-making? I mean as far as I can tell we already do it, this isn't a fantasy, and there's no reason to think we can't absorb more of the market. I can't read the article but it seems more political than economic, judging from the first paragraph
> 280 billion dollar Intel subsidy for chip-making?
Where are you pulling this number from?
From the CHIPS act. It allocated 280 billion, and was criticized for being an Intel bailout since many thought it was a response to their poor market performance. It was signed during the Biden admin, before trump entered office
But it wasn't $280 billion for Intel. Why are you conflating the entire amount with the amount put towards Intel?
Your statements are easily shown to be false. Intel didn't receive anywhere near that amount. It's a really poor way to try to engage with people unless your intent is to mislead.
The discussion was that this article calls Trump's push for chip making a "fantasy", while Biden made his own push years ago. The chips act was 280 billion, and it was widely viewed as a bill to help Intel, who is/was struggling. Nothing I said was misleading, and you're attempting to misconstrue what was said to win an internet argument with a person you'll never meet in a thread no one else will read.
You should consider reading the article: https://archive.ph/h1RrC
And no, Biden didn't sign a "280 billion dollar Intel subsidy for chip-making." If you're talking about the CHIPS and Science Act, that number is overall authorized funding for many things. Intel has received ~10b until now.
[dead]