Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
The American chip industry’s $1.5T meltdown (economist.com)
76 points by naves on Oct 25, 2022 | hide | past | favorite | 46 comments


Honestly? The world's chipmaking should be more decentralized simply from a national security perspective. One needs to only look at the effects of sanctions on russia, once one of the world's preeminent superpowers, now lacks access to chips to such a degree that they're repurposing hugely expensive surface to air (s3-400) missiles for ground attack and relying on iranian drones for precision attacks.

The fact that china could reduce 63% of the world's chip making capacity to rubble at a whim is a strategic threat to the global economy. This is one area where we could absolutely use some redundant supply, even at the cost of taxpayer dollars.


> even at the cost of taxpayer dollars.

And not even a major cost. 50bn sounds like a lot to a layperson (and makes for great clickbait). We spend 700bn on military every year.

It's more jobs, especially different jobs. It's fewer imports (possibly not more exports with this isolationist trend we're seeing globally). There's more trust in the supply chain.

Ths article/propaganda is very short-sighted, as per usual.


The Economist is not especially known for long-term thinking, despite the many self-aggrandizing ways economists describe their own field. The more cynical among us believe this to be a feature, not a bug, for those whose lives revolve around shuffling capital to and fro, as it makes for interesting and exploitable market conditions to enrich oneself.


>It's more jobs, especially different jobs.

It's jobs that Americans can't do (or at least not in sufficient volume). These are high-skill jobs; you can't just grab some uneducated person and expect them to do the work necessary to keep a place like TSMC running.

That's why these chips are made in Taiwan, and not in places like Cambodia, El Salvador, or Rwanda. America also doesn't have the skill necessary to replace TSMC. What chipmaking capacity it has is already tied up with Intel (which doesn't make chips for most things outside of PCs and servers). America doesn't have an excess of engineers that could be trained for this work; they're already busy doing other stuff (usually software), and there's already a shortage of them anyway.

This isn't a problem you can just throw money at to solve. There simply aren't enough skilled workers available to do what the leaders want here.


This is true, but the only way you're going to fix that problem is to start trying. Expertise can be built, even rebuilt, it's just difficult.

Plus, money DOES allow you to hire outside consultants and teachers.

Decentralising the chip foundries is a necessary step, arguably one that should've been made long ago.


Why should taxpayer money be used to fund private ventures though? It's obviously not economically viable to make chips in America, otherwise it would already be happening. Taiwan is not a low cost-of-labor country; it excels in chipmaking because of various factors like education and the industry there, just like America excels in some other industries for various reasons, such as software (e.g. Google and Apple).

Hiring consultants and teachers (from where?) isn't going to help with the complete lack of engineers. It takes many, many years to build a workforce of engineers needed for what's being discussed here, and even then it's not guaranteed because it relies on getting young people to go to engineering school and get degrees in relevant specialties, and then work in industry for a while to gain relevant experience. This isn't something you can just throw some money at to acquire or develop.


> It's obviously not economically viable to make chips in America, otherwise it would already be happening.

Path dependence is a thing. It might be perfectly economical once you get there, it was just simpler to establish in Taiwan for whatever historical reasons.


> There simply aren't enough skilled workers available to do what the leaders want here.

Given how long it takes foundries to get up and running, this seems like something you can solve in parallel.


It's notable that this is the primary argument for farm subsidies -- often pilloried on HN.

When it comes to food, energy, water and critical tech, it makes sense to control your supply, even if it comes at a premium.


> this is the primary argument for farm subsidies -- often pilloried on HN

I don't think there's a strong objection to farm subsidies on their own. The complaints mostly revolve around (a) the food that gets subsidized is food that makes us unhealthy, and (b) the politicians pushing farm subsidies are also against other forms of government assistance, and refuse to recognize the inconsistency in these attitudes.


Automatic car speed cameras are being raided here in Sweden to be sold on the black market to Russia for use in military components. That speaks to their level of need!


Previously discussed: https://news.ycombinator.com/item?id=33244767

2022 October 18 (7 days ago)

285+ points, 679+ comments


> Half a year later the dreams look nightmarish. Demand for silicon appears to be falling as quickly as it had risen during the pandemic.

The article is remarkably ignorant of history. Nightmarish is a absurd exaggeration. With the semiconductor industry, you must invest heavily during any downturn to properly reap the rewards during the inevitable good years. Rinse and repeat across decades.

> So far this year more than $1.5trn has been wiped from the combined market value of American-listed semiconductor companies

Which has only somewhat to do with the semiconductor industry and everything to do with the hyper overvaluation in nearly all asset classes thanks to the Fed's prior, prolonged, low rate program. Just ask the cloud stocks (smashed), or eg Microsoft (which all by itself is missing ~$720 billion off its market cap high). Tesla has been slashed in half, it's missing nearly $700 billion all by itself too. The examples of market cap destruction in the stock market are plentiful.

> In late September Micron, an Idaho-based maker of memory chips, reported a 20% year-on-year fall in quarterly sales.

Oh no, what ever will Micron do. It's a true nightmare. /s They have been around for 44 years and are doing quite well. Micron understands market swings, they have lived through many of them. So have AMD and Intel.

This article is one big click-bait, scaremongering nothingburger.

The only thing worth considering at this moment for the US semiconductor industry, is how much more should be invested than has already been allocated. The US Government should probably put tens of billions of dollars more into the pot for advancing semiconductor R&D, training labor, partnerships with universities, et al.

> It is enough to keep you awake in terror at night.

No.

You can be certain that this article is a particularly wonderful indicator of the exact opposite as to what is going to actually happen. ~6-10 years from now the US semiconductor industry will be stronger than ever and will still tower over China's badly flailing efforts at forcing a semi industry into existence.


Micron is the only domestic US DRAM manufacturer. They are a critical defense asset and will never be allowed to fail.


It's even better than that.

In the long term, the chip ban will greatly reduce China's AI/HPC capabilities. And those technologies are critical for leading in most areas - science, r&d, robotics and manufacturing and others.

This would have the effect of letting the west, to become the economic leader again. I'm sure it's a good future for the Semi companies.


I am less optimistic. Maybe some short term gains but if we go into WWIII over this I doubt our standard of living will survive that.


True. If that happens it would be really bad.


Wasn't this foretold during the chip shortage by most articles written about it? That there'd be a down turn and everyone would have over invested in building new plants? Even worse because the fed is terrified of inflation / higher labour costs the US government is basically causing the melt down.


Yes, Toyota actually saw it coming and stockpiled a ton of semiconductors, which allowed them to manufacture normally way past others.


No, this is the other side of that hill.

Toyota saw the shortage coming. Now we're seeing a glut that will take a decade to work itself out and result in another shortage.


Since the semiconductor industry is known to be boom/bust, I don't know why the US government doesn't have an official, automatic countercyclical spending policy... Micron down 20% YOY? Surely some supercomputer centers could use RAM...


[dupe]

Plenty of discussion over here 8 days ago: https://news.ycombinator.com/item?id=33244767


Is flagging appropriate for (recent) reposts like this? The submission guidelines don't seem to explicitly say so, but there's no "downvote" for posts.


nah, just highlighting to readers that there's a whole recent thread with discussion and this isn't fresh news is all we can do really.

Mods will clean up things if it's closer to original posting date but a week out I dunno.



Aren't these comparing different product segments?

Okay, consumer processor chips are down from their work-from-home/bitcoin boom. But I was under the impression that these new factories were focusing on industrial inputs - microcontrollers and such.


The article mentions that carmakers and the like stockpiled a bunch of chips during the Covid shortages, so there is reduced demand from them now while they work through their stockpiles.


They've been claiming microchip shortage for the last two years. Didn't sound like a stockpile. Ford and Tesla both had production slowdowns or product regressions specifically due to difficulties getting certain chips, and surely others. What's your source?


This article seems pretty questionable because they combined very different situations together. There's no doubt that car companies have big backlogs from shortages of parts needed for final assembly of vehicles, in large part needing semiconductors. Tesla has mentioned this, rivian is another but legacy auto also sees this. But the article conflates possible oversupply in different market segments like cloud, GPUs seem to be in oversupply too says Nvidia even though that won't help car makers. It's just a terrible article.


I don't understand how that can be true. I put a deposit on a new Hyundai last month and Hyundai Canada is telling me it won't even be built until February 2023, specifically citing chip shortages.


If you're buying one of Hyundai's new EVs, those have been so well received I wouldn't be surprised if there were delays on them independent of any chip shortage.


Nope, unfortunately an EV is not in my budget. It's an Elantra with an ICE.


In the Mid Atlantic region you have to pay 20% over MSRP to buy an ordinary Toyota Corolla off the lot right now, it's not limited to new technology or high demand units.


yes, i've been wondering this as well. I mean, i understood in 2021 but it's been over a year now already.


This is hard to believe given that Ford is still selling new trucks with missing features because they don't have chips for them.

The reporter here could still be mixing categories. Ford has thinky computer chips for their infotainment systems, but not enough dumb chips that control the heated seats.


>The reporter here could still be mixing categories. Ford has thinky computer chips for their infotainment systems, but not enough dumb chips that control the heated seats.

Automotive EE. It’s nearly entirely this.

No one that sees how vehicles are made would ever follow the recipe if they were concerned about a chip shortage.

Driver seats, with 8 modules in them. That’s at least 8 micros, at least 8 switching power supplies, at least 8 bootloaders, applications, verifications, 8 different (ish) suppliers, etc. The reality is most of those are more than 8.

There is a movement going on right now that next gen goes back to larger computers like Tesla does. Which is how Tesla was able to say they were so nimble during shortage. They aren’t; they just had a lot less to do.

Current system is bad. But don’t worry, we’ll screw up the next one too!


Super interesting!

So is it a fair simplification that a Tesla really just has everything running through a giant tablet as a cost cutting measure? Wouldn't that affect vehicle longevity? I have 10 year old tablets that can barely boot up any more, but 20 year old power windows that still work perfectly.

I was also hearing that one of the responses to this might be more base models with manual features (cranks, push buttons, etc). Any truth to that?


The screen you see isn’t (shouldn’t) really be deciding anything.

I don’t work for Tesla, but as I understand it their architecture is largely one (effectively a) PC per quadrant.

There are sub modules of course; but as I understand it they are kept pretty dumb relative to the larger controllers.


I think some Audis also still have a "Semiconductor Shortage package", which deletes blind spot/cross traffic alerts, rear collision detection, adaptive cruise control, integrated tolls and wireless charging.


This is partially true. More complex MCUs like STM32F7 and STM32H7 were stockpiled by military as they can be used in missiles and drones. They are still basically unobtanium unless you pay 10-100x to scalpers.


Digikey says it's in stock?

https://www.digikey.com/en/products/detail/stmicroelectronic...

Unless you are saying they are the scalpers.


That's a development board; if you look at the chip that's on that board (STM32F722ZE), there's no stock.


Semico stocks are basically on discount now


For real, even if they don't go back to previous highs they're vastly underpriced right now.


More like game over. When you are competing with the best of the best, with an org overloaded with mediocre to useless people there is no hope of reaching previous highs.


Yet we still can't find raspberry pis anywhere. I call BS on demand being down.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: