Inside Taiwanese Chip Giant, a U.S. Expansion Stokes Tensions - Slashdot
source link: https://hardware.slashdot.org/story/23/02/25/052239/inside-taiwanese-chip-giant-a-us-expansion-stokes-tensions
Go to the source link to view the article. You can view the picture content, updated content and better typesetting reading experience. If the link is broken, please click the button below to view the snapshot at that time.
Inside Taiwanese Chip Giant, a U.S. Expansion Stokes Tensions
Follow Slashdot stories on Twitter
binspamdupenotthebestofftopicslownewsdaystalestupid freshfunnyinsightfulinterestingmaybe offtopicflamebaittrollredundantoverrated insightfulinterestinginformativefunnyunderrated descriptive typodupeerror
Sign up for the Slashdot newsletter! or check out the new Slashdot job board to browse remote jobs or jobs in your area.
Inside Taiwanese Chip Giant, a U.S. Expansion Stokes Tensions (nytimes.com) 54
Posted by EditorDavid
on Saturday February 25, 2023 @11:34AM from the east-meets-southwest dept.An anonymous reader shares this story from The New York Times. (Alternate URL for a shorter version here.)
Taiwan Semiconductor Manufacturing Company, the world's biggest maker of advanced computer chips, is upgrading and expanding a new factory in Arizona that promises to help move the United States toward a more self-reliant technological future. But to some at the company, the $40 billion project is something else: a bad business decision.
Internal doubts are mounting at the Taiwanese chip maker over its U.S. factory, according to interviews with 11 TSMC employees, who declined to be identified because they were not authorized to speak publicly. Many of the workers said the project could distract from the research and development focus that had long helped TSMC outmaneuver rivals. Some added that they were hesitant to move to the United States because of potential culture clashes.... Its factory expansion in the northern outskirts of Phoenix is meant to bring advanced microchip production closer to the United States and away from any potential standoff with China. Yet the effort has stoked internal apprehension, with high costs and managerial challenges showing how difficult it is to transplant one of the most complicated manufacturing processes known to man halfway across the world.
The pressure for the Arizona factory to succeed is immense. Failure would mean a setback for U.S. efforts to cultivate the advanced chip manufacturing that mostly moved to Asia decades ago. And TSMC would have spent billions on a plant that did not produce enough viable chips to make it worth the effort.
-
›
It will probably fail because it is forced against the rules of business, economics and capitalism that led to the transfer of the chip industry to Taiwan in the first place.-
What rules?
Why do you think nothing has changed since that happened?
-
I’m assuming they mean that it’s very difficult to succeed when you just use large amounts of money, subsidized or otherwise, to create a large plant wholesale compared to incremental growth. Getting everything dialed in is very difficult, whereas if it’s grown “organically” from the ground up it’s nearly a given this will have already been ironed out at lower costs in the past. It’s a similar concept to why you can’t just take farming conglomerates land and
-
It's hard the first time, but it's hard no matter how much money you spend, and no matter how quickly you try to work. If you're duplicating another plant, it should be relatively easy as you've hopefully recorded the solutions to the problems. You may face new problems, of course.
Except what actually happened is we took the land from tens or thousands of psuedorandom people, and handed it to farming conglomerates, and then they fucked it all up.
-
-
I don't usually waste my time on the NYT's bullshit, but I did read the other linked article. what it says is "the project could distract from the research and development focus that had long helped TSMC outmaneuver rivals. Some added that they were hesitant to move to the US because of potential culture clashes." But none of that speaks to the specific claims to which I replied.
-
-
Where are the people with expertise going to come from? there aren’t many that aren’t employed but have the skills in key areas of production and problems with splitting them is what the article talks about. Growing organically means you also grow expertise, something most management thinks is completely useless and irrelevant until the entire business model collapses. Those past mistakes would be made at a smaller scale, sure it’s hard to grow a plant but it’s actually harder to
-
If they haven't been creating them already, then they have fucked up already.
Yes, and if you don't deliberately transfer it to others then you're setting yourself up for failure even if you don't expand.
I'm sure it is, but it also offers architectural advantages as you can better plan for the full size from the outset.
Yes, that's another way to put it. But the rules were set up in a way that made that possible. They didn't recognize the value of the way things were done before. Increasing centralization of
-
Yep, unchecked capitalism has been quite bad for everyone and the offshoring of all production to save a few cents short term because the value was more than a quarter away is coming back to bite us all. Well, those of us who aren’t in the 0.01% anyway because for some strange reason rules don’t seem to apply to them when it’s inconvenient.
-
-
-
-
The story is trying to tell the story of purported exceptionalism in the minds of the the TSMC Taiwanese that believe Americans are too lazy to be successful at running fabs. Perhaps there is some truth to that. I used to work with a Japanese fellow who wasn't that bright. However, he could run endless tedious experiments that no one else would have the patience to do. I think this idea of hard work is what the TSMC Taiwanese think is core to their exceptionalism. It's the moat that will prevent Intel from encroaching on their market.
There is a great irony here. Americans look down on Asian rote effort while Asians look down on America laziness, while each cherishes the supposed advantages of their own mindsets.
-
If I could up vote your comment I would.
-
-
-
It is done for political reasons.
-
It's being done for strategic reasons. Politics is part of how the strategy becomes action. The world cannot afford to be dependent on a single region which is under dispute for advanced semiconductors.
-
-
-
Isn't it just a manufacturing plant? The only real question is whether they can get their lines to work right with the quality of work they will encounter from US workers. Of course, for TSMC, trying to make it work may cause them to fall behind at home, which is the really dangerous failure.
-
Isn't a heart just a heart, a brain just a brain separate from everything else it needs to work? No.
-
-
Rules of business and capitalism such as state subventions [livemint.com] ?
Asian countries routinely engage in protectionism, and yet they sound surprised when the US do the same.
-
Nothing was transferred. Taiwan produced their own industry by simply being successful. TSMC wasn't "transferred" anywhere.
-
It was transferred in the sense that we stopped most production here and bought from them instead. The purchase of chips for American companies no longer came as much from chip makers in America.
-
The vast majority of TSMC's growth came from cellphone SoCs. Exactly which American chip fabs went out of business by being outdone by TSMC in that market?
-
-
-
The TSMC founder left the US for Taiwan because Americans kept hamstringing his career due to his accent and perception as a foreigner. That isnâ(TM)t the primary reason chip mfg is currently offshore of course, but it is an interesting instigator.
-
-
The sand quality in America won't be as good, which means that they'll either need a more extensive purification process OR risk a higher failure rate. You need the silicon to be ultra-pure and ideally mono-isotopic, to minimise defects in the silicon. That's going to be a potential problem for them.
They cite cultural differences as a potential issue. By implication, the company is going to be run using Asian philosophies towards work. But that might conflict with Federal health and safety standards (which, frankly, are too lax as it is). It will certainly conflict with American exceptionalism, which is a serious problem in Arizona.
There's also going to be the problem of both toxic wastewater and radioactive waste, which I believe were two of the factors that closed down a lot of American chip manufacture.
Finally, there may be problems with getting qualified employees. Universities are working on a profit motive in the US, and there's no profit in producing employees for an industry that barely exists. That means there will be a several year delay in getting any significant number of employees capable of working at the scale they're going to operate at. There will be differences in the tools and methodology in making chips at a scale of tens of nanometers, I'd have thought.
We have to accept that the company has thought all these issues through, but it seems clear that the employees of the company aren't confident the company has thought the issues through thoroughly. Which, when money is waved at them, is understandable.
-
The staff will largely come from Taiwanese refugees once China seizes control.
US will take the top refugees.
Without RKBA there's no way to repel an invasion and China is sick of US warships in the Taiwan Strait.
-
If China seizes control it would be too late for refugees with Fab expertise (or other strategically important knowledge) to leave. I think it is much more likely they would be forced to move to mainland China.
-
-
Radioactive waste?! LOL
-
Sand qualtiy? Radioactive waste? Unless those are supposed to be some joke I don't get (it's been known to happen), it sounds like you have no idea of what goes on in a wafer fab. As far as qualified employees go, many of the new employees will be technicians (2 year degree) and do highly regulated tasks they will be taught. There will be a good number of high level engineers, but they will need pretty much exactly the same skills as people doing similar jobs at Intel, Global Foundries, Micron and a bunch of other existing US fabs. Cultural issues - sure they could be a problem.
-
Their current plan to deal with the cultural differences is to do all training in Taiwan. I imagine itâ(TM)ll take 10 years for the cultural differences to be circumvented.
-
-
TSMC can get double the yields as Samsung out of their advanced EUV nodes, some of that is going to be process, but I wouldn't be surprised if they have some secret modification for the stepper (more specifically, the pellicle). How well can secrecy be maintained outside of Taiwan?
PS. ASML employees will generally complain when I suggest a lowly customer might tamper within their holy machine beyond adding an industry standard pellicle, but when a single company makes up most of the revenue, they can keep secrets even from a subset of the ASML employees. Service engineers, management and lawyers would know if there are some to them black box components added inside the enclosure, but it can still be need to know.
-
Can EUV reticles even have a pellicle (reflective optics in a vacuum)? Wouldn't it absorb all 13 nm light? TSMC might well have some secret sauce, but if it is related to changes to their EUV tools ASML will know all about it - there are several full time engineers camped out with each NXE tool. However I bet both TSMC and ASML have signed plenty of NDAs with each other.
-
After a little googling - I stand corrected. There is a fancy EUV reticle devloped by ASML. You could well be right.
-
pellicle - not reticle!
-
TSMC had good yields even before the pellicle was out, that's why I suspect they had some homegrown solution.
Maybe sweeping the space in front of the mask with an electron beam so a charged mask could repel the tin droplets? Who knows. They've got as many rocket scientists walking around as ASML and as much money too.
-
-
-
"Some added that they were hesitant to move to the United States because of potential culture clashes"
They're afraid of Magatards.
-
You are literally getting subsidized by the global hegemon to make a new plant, in a state with a high knowledgebase and resources for chip plants and you are worried about it being distracting? When you also get the benefit of one of your fabs being in both a massive geopolitcally and geographically stable region which does a great deal to spread your risk profile out.
Also a lot of this seems like culture tension between management and working styles and honestly, a little bit of good ol' fashioned racism from the Taiwanese towards Americans.
Also employees of TSMC working in Taiwan, your jobs have a way higher likelyhood of not being around without the global military umbrealla America provides, especially to Taiwan. Suck it up and figure it out, American citizens subsidize your national defense.
-
Of course, the Taiwanese businessmen have to consider both advantages and disadvantages of moving to an American "right to work" state.
On the up side, they won't have to worry about a lot of environmental regulations, high salaries or worker safety nonsense. On the other side, though, they'll have to contend with poorly educated workers, casual racism and outright white supremacy, mass shootings, and the necessity of paying toward health care for employees, because it isn't "universal" as is the case in ci
-
-
So you think being functionally turned into indentured servants for the Chinese state a compelling alternative?
-
-
Well if TSMC is a well functioning organization then its corporate culture is well melded to the culture of the people it employs.
If you open a plant in the US, that culture is going to be different. An obligation an employee would think nothing of in Taiwan might be objectionable to an American worker, and vice versa.
And yes, a lot of your most senior people are going to involved in setting up this new plant which will distract them.
It's still a good idea long term... but it's not without risk.
Well that's
-
Sure, cultures are different and it's a challenge but this is hardly new territory for any business nor even semiconductor businesses.
Every Japanese car manufacturer successfully operates vehichle plants in the US, many of them in multiple states (My Toyota truck was assembled in California for example). Japanese working culture is vastly different than the US but they all have figured it out to great success.
Another way to look at whats being said here is these employees are more expressing here is a lack
-
Microchips are arguably a more complicated process, and more importantly, auto manufacturing has a long history of being done all over the world.
I'd say it is quite a bit harder for TSMC, who has never done anything like this, to suddenly open a new plant in the US.
Yes, for any of those countries it would be the same argument.
It's not really a lack of confidence in the leadership as an acknowledgement that they're talking about a very difficult undertaking.
The US has long seen the benefit in using its milit
-
It might be "harder" but it is not unprecedented by any stretch and one could make an argument that operating an auto plant which requires thousands of blue collar workers in middle America where they tend to be is harder for a foreight company to adapt to culture than a highly automated technical production line in a region where there are already several of these and a well established population of trained workers in the field. If TSMC is struggling with this it's a leadership problem, not a problem with the concept of this fab plant.
It's not really a lack of confidence in the leadership as an acknowledgement that they're talking about a very difficult undertaking.
Yes it is absolutely a lack of confidence in leadership because this by all other accounts is a good business decision that will be profitable. It's a huge expansion directly in the largest consumer of it's products (in the signle largest consumer economy on the planet) in a stable region that they are even getting subsidized for. The only downside to it is that yes, things cost more to build in the US, which is exactly why we are giving them (and several other companies) billions of dollars to do it! And they accepted it!
The Trumpian part is trading in that good will for transactional benefits (we protect you, now you owe us a microchip plant).
That is a far more pointed and serious accusation than the scenario or opinion I laid out and an extrordinary claim so you need to post some evidence for that.
Also I am not here to defend Trump or his business decisions but the idea that this type of program or the incentives that make a company like TSMC want to enjoin it are not exclusive to Trump or particulary unique in diplomatic relations. The idea of framing this as "Trumpian" is giving the man far more credit than I ever would.
The same thing happens in diplomacy. If you're always looking to squeeze your partners they get the sense that you're playing a zero-sum game. And if that's the case then you're no longer an ally and they look for better partners.
Except this is not a squeeze by any stretch, this is perfectly aligned incentives on the part of the US, TSMC and the nation of Taiwan. Everybody is getting something they want out of it.
-
Is it? Your original quote:
Also employees of TSMC working in Taiwan, your jobs have a way higher likelyhood of not being around without the global military umbrealla America provides, especially to Taiwan. Suck it up and figure it out, American citizens subsidize your national defense.That's not a particularly subtle threat (we walk away and China conquers you), combined with a presentation of a bill (Americans subsidize your defense), and a suggestion of how to pay (build a microchip plant here).
It's perha
-
Sorry that was my personal opinion in regards to the employees leveraging complaints here, I read that as your idea of official US policy so maybe beams were crossed. To me it's a matter of have a little perspective of how the world is operating around you. You can call me Trumpian if you want but I was not making it a "transaction" but you more of a "you don't exist in a bubble"
Here an analogy I think works; people who complain about the US taking measures to fight climate change becuase "other countrie
-
-
-
-
-
-
That's not very likely. If anything, Taiwanese favor Americans too much.
-
-
If the semiconductor crown jewels are duplicated outside of Taiwan, there is the option to use Taiwan in a proxy war. Without that semiconductor redundancy, that option is off the table.
-
I doubt that it will be the most leading edge stuff being built outside Taiwan. And I don't hear anything about the R&D shifting out.
If what's being build in US is a "glorified factory", unless they keep upgrading the nodes in that factory / build new buildings to keep on running the latest nodes, it will not be leading edge within a few years anyway.
As long as Taiwan is the only place with the newest and greatest leading edge nodes available, Taiwan will have the US safety net. Anyway it will look bad
-
-
shifts away from Taiwan...one of two things will happen. China will invade, no one will care. China won't invade, because there isn't anything worth stealing.
-
doing many different things, rather than one with everything on the line. Going too "high concept" with economics does not typically work. Always aim to be a flood rather than just a landslide.
-
They are worried that after global semi conductor capacity leaves Taiwan there will be less of an interest to defend it. TMSC will only export 7% of capacity to the US, that would hardly ride us over in the event of a war, and be enough to keep defense going
-
It's not like AZ doesn't have fabs. Now Ohio? That seems like culture clash, but it's Intel there.
Also...you build the culture with your actions. Who you hire. Where they live. Where they spend their money. Locals like new money.
-
This needs to go forward.
Because, the entire country is one bad day away from being destroyed and nationalized by China (which will further destroy it).And as much as Intel and other manufacturers can take up SOME load, TSMC's fall could cripple the industry for years.
And if you think CHINA will respect patents and the like when they can just steal everything steal-able...
-
Mark my electrons, by the time the facility is operating to specification, TSMC will be seaching the couch for cash. If you shares or future are tied up with the company, you should be worried.
Come back here in five years and tell me if I'm wrong.
Recommend
About Joyk
Aggregate valuable and interesting links.
Joyk means Joy of geeK