Well, this is practically nail in the coffin news:
Meanwhile Intel is laying off 15,000 workers, slightly over 15% of its workforce and announced yield issues with its new “Meteor Lake” semiconductors. Its last two generations of chips have hardware problems which cause and crashes and have no software fix. (Buy AMD for your next computer.)
The CHIPS act was supposed to cause a rebirth of US chip manufacturing, but so far this seems unlikely. The US no longer has the necessary skills. Meanwhile, China is roaring on chips, taking over the legacy end of the market and pushing hard on the rest of it.
The end result of semiconductor sanctions is going to be that China will become the dominant player in semiconductors, a field which American allies dominate.
This wasn’t necessary. Chinese companies preferred foreign semiconductors before the sanctions: they sneered at domestic made ones and felt the quality was inferior, which it was. But what was done to Huawei told every Chinese company and China’s leadership that they couldn’t rely on foreign suppliers under Washington’s thumb.
A little history, because sanctions and CHIPS were aimed just at China.
Back in 1986 the US dollar was very high due to Volckers high interest rates. The great powers of the time got together and agreed to intervene to depreciate the dollar. The great loser was Japan, and the great winner in Semiconductors were Intel and Micron: Japan had previously been clearly in the lead. The Plaza accord later contributed to Japan’s crash, and the real end of its post war golden age. Doing so may have helped America, but it weakened the American alliance system’s overall power significantly.
CHIPS and sanctions can be seen as an attempt to return the lead from Taiwan and South Korea to America, but this is failing and it will also lead to American allies losing the lead.
As Kissinger once said, It may be dangerous to be America’s enemy, but to be America’s friend is fatal. This is especially true as American hegemony ends. America is cannibalizing or trying to cannibalize its allies: much of German industry crushed by anti-Russia sanctions, for example, has moved to America.
But America no longer has the real engineers (not software engineers) and factory know-how to scoop up cutting edge semiconductor production.
We’ll see if China, which is now more powerful than the US in every way that matters except deployed military, is better to its allies.
Smart nations will cut deals with the Chinese and re-align towards it if they can get anything close to a good deal. The West is going down and the decline will not be halted any time soon.
Ian Welsh
I think it’s more a case of “just because the consensus says” rather than it is always wrong.
Though you can make a case that every model is wrong or at least leaves things out.
As an aside, in the 19th century most experts also thought that heavier than air travel was impossible.
Ian Welsh
I think it’s more a case of “just because the consensus says” rather than it is always wrong.
Though you can make a case that every model is wrong or at least leaves things out.
As an aside, in the 19th century most experts also thought that heavier than air travel was impossible.
Daniel Lynch
Ian said “The US no longer has the necessary skills … no longer has the real engineers.”
Well, if you lay off 15,000 engineers and they can’t find another job in the semiconductor industry so they drive Uber or sell real estate and lose their technical skills, what do you expect? It’s not that American workers are dumb, but we have to adapt to the job market, which is less manufacturing for the masses and more providing services for the rich.
Back in the day I was a real engineer for a semiconductor fab, and loved the work. But my Japanese owned fab was sold in a leveraged buy out to a ruthless Silicon Valley private equity firm, who promptly laid off 2000+, including me, and outsourced much of its manufacturing to TSMC. So yeah, American can’t make anything anymore, but don’t blame the workers, blame our Ayn Rand flavor of capitalism.
Today I would not recommend semiconductor engineering as a career simply because it is not stable. Even in the good old days it was a cyclical industry with career-wrecking downturns and layoffs. Better to become a civil engineer and work for the government — the work is not as technically interesting, but at least it is stable.
Oakchair
“The last will be first, and the first last.”
How this is playing out currently is that the first cease to really innovate because “we’re the best”. They instead spend their time/resources establishing monopolies, Ponzi schemes, marketing, and looting.
They’ve become arrogant and corrupt. Those espousing new ideas or questioning the orthodoxy are canceled. As a result the populace of the “first nations” are becoming more and more heavily poisoned. Anyone questioning the poison is a “crazy anti-science conspiracy theorist” so it goes on and on. IQ rates are falling, dementia autism rates and illness rates are skyrocketing diminishing the cognitive functional ability to even do things like build an airplane, or read.
China is already ahead and will continue to get further ahead, but they can easily (and in part already have) fall prey to same traps the West jumped into.
I suspect the west understands this, and that is why they’ve spent the last 7+ decades systematically destroying countries and the health of poorer nations. They’re strangling any future competitors before they can arrive. Similar to the dark forest explanation for the for the Fermi paradox.
Ian Welsh
Daniel,
yes, layoffs and so on are very destructive to keeping skill sets, and this idiocy has been going on for a couple generations at least.
OakChair:
they will fall prey to various traps. They already have in some significant ways. But for now they’re the upcoming primary great power, and they’ll retain that place for as long as I’m alive and at least for the lives of anyone over, say thirty, barring particularly catastrophic and specific climate change results.
Oakchair
But for now they’re the upcoming primary great power, and they’ll retain that place for as long as I’m alive
—–
Yep, and looking back the contrast of China compared to the rest of the undeveloped world is revealing.
In 1950 China was as poor as Africa.
Go around asking people how they think the Chinese have done economically under the Communist Party.
Then witness the level of emotional, and intellectual maturity when you point out that in 1950 China was as poor as Africa.
Soredemos
I’ll buy AMD if the performance is better for my purposes.
I’d be quite surprised if they don’t resolve the hardware issues with the next gen (which is Arrow Lake, not Meteor Lake), since it’s a whole new architecture on a new, smaller process node anyway. But even if they somehow don’t, in practice many, many people have been using the previous generation chips with no problems in daily use.
It’ll likely come down to which company’s CPU has better single-core performance. That will probably be AMD, with Intel having better threaded performance. So AMD for gaming, Intel for video editing and the like. I’ll wait for the benchmarks.
On the GPU side the best is going to remain Nvidia, with AMD simply continuing to gets its ass handed to it. Intel is carving out a very interesting niche there where they’re giving potentially (driver support remains spotty, but is constantly improving) ludicrous price to performance, laying claim to price segments that the other two abandoned years ago.
marku52
China has spent some time looking at where the west went wrong, specifically the worship of capital and financialization. They’ve cut billionaires down to size, given the death penalty to big time fraudsters. They are not going to let their version of Wall Street destroy what they have achieved, which is really nothing short of miraculous.
IN 1950 they were Africa. Now their cities are Cities of the Future. Africa should be going to China and asking “Show us how…..”
GlassHammer
If you are smart enough for STEM your probably smart enough for a bunch of non-STEM careers as well.
People follow the money and… THE LEAST AMOUNT OF LABOR WITH THE BEST WORK LIFE BALAMCE.
People forget that the money was all that made the absurd amount of hours and awful work life balance of the STEM fields tolerable. So when the money wasn’t enough they just switched careers.
Jefferson Hamilton
>real engineers (not software engineers)
lol
Godfree Roberts
In terms of deployed military, China is as far ahead as it is in research. China’s navy is much bigger, much newer, and much more powerfully armed. Ditto its army, which has beaten the US (badly) once already. And we have no aircraft to match their J-20, nor the missiles it carries/
The, too, China’s capacity to manufacture war materiel is 300%-500% greater than ours, its economy is much more powerful and its morale is far higher than ours.
StewartM
The CHIPS act was supposed to cause a rebirth of US chip manufacturing, but so far this seems unlikely. The US no longer has the necessary skills.
Is it that, or is it as much at least a case of “you can’t do anything if every initiative you make is clawed back by every dip in the stock price”, resulting in staffing cuts, cancellation or deferrals in capital purchases, and whatnot?
I’m thinking that we have people who know how to get this done, but they were pushed how the door years ago to be replaced by cheaper kids without the know-how and experience. That’s how,, by my experience, our firms lose institutional knowledge.
The fundamental problem created by Milton Friedman “only stockholders matter” capitalism is that you can’t do anything other than creative paper-shuffling if the stockholders run the company.
Altandmain
Ian, it’s even worse than you think – note the last column, the change in adjusted share of research year over year – the 3 Western institutions are in decline relative to China.
The US still has some semiconductor engineers, but the “fabless” model is more common (that’s when the West designs its chips in the US and then pays someone else, usually TSMC or Samsung) to actually make the chips. Intel has lost foundry leadership to TSMC, something that it historically led, due to its issues on 10nm nodes (poor management at the company and decades of stock buybacks).
The main bottleneck for China is now building their own domestic EUV technology, something I suspect will take a lot less time than the West hopes. The US has asked Dutch maker ASML, which makes the EUV machines, to sanction China. So far, all of the sanctions on China and Russia have badly backfired for the West.
The other issues that would have to occur in the US as follows, the education system would have to change to train more foundry engineers (that takes years) and chips would have to be more attractive as a career, certainly compared to management consulting, finance, and software development. Oh and there’s also the issue that China is way ahead in fields like math and science in schools. That’s not going to be easy to resolve.
It was inevitable that China was going to enter into the semiconductor space, but the US actions have accelerated the pace of China’s progress by giving the Chinese a sense of urgency that previously did not exist.
Willy
I met a real live Mayan descendant after the Apocalypto movie came out. Like many, he said the sets, costumes, acting, and action were pretty good, but the historic accuracy sucked. He was also a student of world history. He said the evidence was mounting that his ancestral civilization didn’t collapse because of resource depletion or environmental degradation or even Mel Gibson arriving with the Spanish priests. But it was SOSDD.
Civilizations grow when the economic and cultural ground is fertile for the many. Civilizations shrink when the economic and cultural ground become fertile for only the few.
What the hell does that have to do with China? The current Chinese leadership is obviously more into trying to grow their civilization on behalf of their civilization and basing efforts on proven results, than they are into growing their own peer group’s sucking off that civilization and rationalizing with ideologies, religion, and mantras.
It’ll be interesting showing that chart to neoliberal trickle downers. If they don’t have some kind of denial fit, will they blame…. Satan?!
Carborundum
I suspect what will happen is that no one will dominate in semiconductors and that we’ll end up seeing a more distributed market, which strikes me as a good thing given the importance of the tech. I’ll be very interested to see what happens to fab investments when interest rates stabilize at more normal levels.
As to measures of relative research throw weight, we really need to come up with something better than volumes of articles published and patents issued. Given the now significant distortions in their respective “markets” (the explosion of lower quality academic publications and patent lawfare) neither of these are the indicators they used to be.
Poul
Remember the all-importance of stock buy-backs in US industry. Cash in the profits and forget the long term consequences. More simply put it’s “Take the money and run!”-manegement
https://www.nakedcapitalism.com/2024/08/hoisted-from-comments-is-intel-going-down-the-boeing-path.html