5nm in the USA: TSMC's Board Approves $3.5B Fab in Arizona | Hacker News Reader

5nm in the USA: TSMC's Board Approves $3.5B Fab in Arizona

tomshardware.com

288 pointsmardiyah105 comments

I've always thought that TSMC's dominance in fab processes are a strategic asset in the defense of Taiwan against potential PRC encroachment. I would assume that as long as a large part of the global tech industry relies on TSMC, countries such as the USA would have a strong incentive to potentially protect Taiwan against threats. Does anyone have insight into the geopolitical impacts of Taiwan - USA relations in regards to TSMC fabs on US soil? Would Taiwan view this a strengthening of relations between the two, as I would assume this provides the USA with some "domestic" 5nm fab capacity, or is there some risk in that once this fab comes online the USA may see Taiwan as less critical? Anyone know some good sources discussing the above?

1. This is a very small Fab, comparatively speaking, and it is likely for specific use. ( I am guessing military )

2. 5nm in 2024 is at least 4 years / 2 Node behind leading edge. Given the complexity with the project I actually expect HVM happens in 2025. They will of course produce something in 2024 just to make headlines, but HVM always start later. ( The public wouldn't know or care anyway )

3. This means TSMC in Taiwan is still as important it was, and arguably more so in 2024. From Geopolitics prospective this Fab doesn't change a thing.

It seems that the arrangement is a win-win for both sides. By the time the 5nm fab is in operation in 2024, it will be old tech for TSMC, since they are planning to have 3nm in high-volume production and 2nm already in risk production. The fab may very well still be the most advanced in the US, depending on whether Intel gets their 5nm process ready as planned in 2023.

In a full scale war with China, I don't think TSMC's fab can be preserved. China can destroy it without landing on the island, easily.

Taiwan is too close to China, that is a reality. Plus, modern chipmaking is an incredible sophisticated business, for example, you need huge amount of electricity and raw materials imported overseas to sustain the fab. When the war breaks, all of that would end.

Their effort should be spent on to prevent the war from happening in the first place, that is the only way to 'win' the war.

China relies on the fabs in Taiwan just as much as everyone else though.

I agree that war should be avoided and "the only way to win is not to play"!

However, I think we should also be considering globally how we can act more like Holons [1] for greater resilience from all types of threats to global interdependence. We want loosely coupled systems, we should want loosely coupled government too!

[1] https://en.m.wikipedia.org/wiki/Holon_(philosophy)

Why would China destroy it when they can simply occupy Taiwan without much recourse?

I would bet in all their war-games, very few strats have the "TSMC destroyed" scenario checked.

Very likely TSMC would destroy it themselves...scorched earth works.

> without much recourse?

I think you underestimate how violently the US will respond to something like that. It honestly would the greatest existential threat to the US since WW2. Chips are a decent amount more important(at the moment) than oil and look at the crazy shit the US did for that.

When people say or imply that the US engages in "war for oil", it's virtually always (in my observation) left ambiguous whether the claim is that the US pursues genuine national interest, or that the foreign policy is subverted by some unspecified special interest not aligned with the US as a whole, such as private oil companies.

What are you really saying when you say the US did "crazy shit" [for oil]?

Your guess is pretty on the point. TSMC has been jokingly hailed as "The protector of the country" by the Taiwanese media. The chips have military significance for the USA, as it has been trying to avoid China made electronics in all (or most?) equipments.

At least, for now, I don't think the US will let Taiwan be eaten by China. The US did protect (and to some extent create) Taiwan long before Taiwan had any value.

Taiwan being invaded by China is a signal of the end of US hegemony around the world.

> Taiwan being invaded by China is a signal of the end of US hegemony around the world.

It'll only loudly signal an end to US hegemony in Asia (which has already happened de facto). That is inevitable regardless of a Taiwan invasion and everyone knows it at this point. China has gained immense positioning during the pandemic while all other major nations have lost. There was never a scenario where the US could retain its former dominance in Asia as China arrived at being a full-fledged superpower (they'll be there militarily this decade). US hegemony, to the extent it exists, will remain intact elsewhere for decades after it lost that status in Asia. US hegemony isn't based on signals and perception, it's based on military force projection and whether there are equal counters to that in a given region. China is very close to powerful enough to counter the US military in the Asia region (not just near its borders) and that determines the hegemony regardless of whether China goes into Taiwan anytime soon. Anyone thinking the US still retains hegemony in Asia is most likely fooling themselves at this point.

George Friedman and Peter Zeihan say that it is not clear that China would win a war even just against Japan.

Both countries rely on imported oil. The most likely cause of a war between them is an oil shortage, giving nations an incentive to use their navies to capture oil tankers. The use of navies to capture cargoes of national importance was common for centuries till end of WWII. It ceased after WWII because of US naval dominance combined with the use of free trade as a part of US strategy in the cold war, but of course the cold war is over; so the US, which has been and remains the power least reliant on international trade, could choose not to spill the blood of US sailors to keep the sea lanes open to all.

Japan knows this and has been building a navy that can operate outside the South China sea and in the Indian ocean, so that it can escort oil tankers from the Persian Gulf to Japan. China's navy in contrast cannot operate outside the South China sea, he says.

Friedman says that aircraft carriers benefit a country only if it has admirals that know how to use them and points out that Japan's admirals have been trained by admirals trained by admirals that actually used aircraft carriers in a major war.

The main purpose of the People's Liberation Army is to put down a rebellion by Chinese citizens. The primary goal and concern of the rulers of China beyond their Party's remaining in power is to prevent the country from splitting apart. For most of the 2500-year history of China, it has been divided into multiple polities, and outside powers have played the pieces off against each other.

The US just lost a war against the Taliban, a war it spent many trillions of dollars on, so I doubt the US military worries about whether a loss in a future conflict would signal an end of US hegemony.

Can you provide sources for these claims? Thanks!

The congressional research service is a good place to start. Noting the source (government) of course.

While TMSC is no doubt ahead of competition, it is just by margin of 1-2 generations at max, which accounts for max 2-3 years for friendly country like Korea to catch up. So if some Taiwan politics comes into play, we will just lose the progress for some years of semiconductor fab tech, not a small thing but it is not that the chips couldn't be built if not for Taiwan. We would just lose something like 10-20% of performance in the chip.

>> While TMSC is no doubt ahead of competition, it is just by margin of 1-2 generations at max

No, they're ahead because they mastered EUV lithography. It's not just another shrink, it requires a lot of changes and they all have to work and play nice together. The fact that Intel can't seem to get it right, and global foundries voluntarily stopped short is very telling.

Having cleared the hurdles TSMC is rapidly going through a bunch of new nodes: 7,6,5,3,2.

Hasn't samsung already started production with EUV lithography?

Yes, Samsung's EUV process is in production since 2019.

IMHO the latest is 8nm, not ensuring true or half

EUV machinery was delivered to TMSC by a firm from The Netherlands.

And Intel has purchased such equipment from them as well. A machine is not a fab-process-in-a-box or everyone with money would be in production already.

Few years ago EUV was not considered production ready. Intel made a reasonable bet that given their lead in the traditional lithograph by continuing with it they would stay ahead of competition until EUV matures.

TMSC instead decided to try EUV. That was a risky bet. But then ASML from The Netherlands managed to solve the problem of EUV intensity and Intel had found that they just could not manage to push traditional lithography further.

the number please?

ie. how many intel and TSMC bought and is using now

ASML

You can't simply swap out the guidance and control systems of state of the art kit for a bunch of chips 5 years older. That's not how high performance embedded systems engineering works. Older chips need more power so bigger batteries, produce more heat, take up more space, have lower performance, all inside a missile, munition, satellite or fighter jet with extremely limited space and exacting design tolerances.

The best you could do is go back to making older systems for which the factories may not exist anymore, or design and build new systems using the old CPU tech and new everything else. Both options would impose a severe economic and military cost and take years to implement.

It doesn't matter that 5nm chips aren't used in military hardware today. You can bet military hardware is being designed now that will use 5nm chips. Therefore access to that production capacity is strategic.

I would argue the opposite, that it is a huge deal for defense.

Imagine if the situation is missile defense, both the missile and the defense systems are computerized and not only the milliseconds but also the microseconds are the difference between a hit and a stopped missile. Whoever had the newest fab node would have a notable advantage in being able to push their system harder. It could be better ML in missile to lock on and dodge, vs the defense system predicting paths as fast as possible with as low latency as possible to defend a ship or similar asset.

Its asymmetric in the example of course, but both sides would be doing both shooting and defending.

Your scenario is not founded in reality. Few problems that involve computers are limited by or can primarily be solved by computer technology. In your example, one of the main challenges is that every missile defense system can be overwhelmed by shear number of attack missiles.

Have a look at what type of microchips are used in “cutting edge” military equipment like fighter jets. They are often decade olds. More speed is simply not important enough.

I think what you say is correct, its so much easier to fire a missile than it is to stop one. Honestly, reliability and cost are bigger drivers than raw performance. The scenario is a bit bad I can admit, and did admit.

But if we were in desperate times, I think the limits would matter and fab tech would be an important strategic asset for what it enables. The nature of it is that its easier to attack, but we hopefully have something to defend.

Intrinsically larger process nodes are better for analog and have uses even as new fabs come along. In 7nm your analog signals are not treated very nicely, but these are important for sensors. Also power electronics are not really on the newest nodes either. The microchips are old because they get the job done and I don’t think we feel so threatened to push the limits.

Not only that but smaller cores are less radiation resistant. Meaning more likely to get interference from natural radiation sources. So less reliable.

Are defense systems on such cutting age tech? I would have assumed they are using 40-200nm as development cycle for such system is usually in decades.

Generally old chips are good, depends on the purpose. Some things like the f35 would like to use TSMC’s fabs. But not every chip is for only processing and larger process have their advantages for certain applications like power, analog, and em hardening

Also the fab for a 20year old process is probably easier/cheaper/faster to re-build in case the enemy destroys it, than the fab for 5nm process you've been building for 4 years.

Nevermind the nation probably has more than 1 of such old fabs, so it can't be technologically decapitated by destroying a single precious 5nm fab it has.

My guess would be there is a plan for the US to bomb the TSMC factories in the event of China winning an extremely costly war against Taiwan. I’m not sure the US would actually come to Taiwan’s aid. I mean surely Taiwan has enough arms to destroy a very expensive chunk of Beijing and Shanghai in response to any invasion? Terrifying stuff...

It’s almost certain that the US would be involved if the PRC chose to invade Taiwan. Losing Taiwan would be a geopolitical catastrophe: protecting Taiwan from an invasion with the Pacific fleet would be critical in preserving the Taiwanese Armed forces and thus preventing more American lives from being spent in the war.

Seems pretty far fetched considering today’s economic and political conditions though.

There is a reason the US maintains a nuclear deterrent.

have you considered the scenario that taiwan would peacefully reunite with prc? what would the us do? fight?

First of all, there is no ‘reunite’ because Taiwan has never been united with the PRC. It has been governed by the Republic of China, which predates PRC by many decades, since Japan’s defeat in WWII and subsequent surrender of its colonies. A unification scenario is less likely today than it has ever been before, and grows less likely every year. Millions of Taiwanese outright reject the idea that they are part of China. A diminishing minority of waishengren that evacuated the mainland to Taiwan in 1949 are about the only people who would favor unification.

Nothing wrong with considering that scenario, but as for a national referendum resulting in a peaceful unification? Chances are basically zero.

> First of all, there is no ‘reunite’ because Taiwan has never been united with the PRC.

There is no reintegration of Taiwan into the PRC, since it has indeed never been part of it.

But the term 'Chinese reunification' is the correct one, because China has indeed split over "unfinished" business in 1949.

In any case, IMHO this narrative is moot when discussing peaceful reunification because I can only see this having a chance of happening after the mainland has transformed into a democracy, which would probably mean that it would not be the 'PRC' anymore.

Taiwan hasn’t been integrated with mainland China since the Qing Dynasty. It goes further than unfinished business in 1949. Korea and Vietnam also used to be integrated with ‘China’. The Taiwanese have had plenty of time to forge their own independent culture. Democracy on the mainland or not, the people of Taiwan would benefit very little from a union with a country that has half the per capita GDP that they have.

I am trying to stick to facts and to refrain from pushing an agenda here.

Taiwan was seized from China in 1895, and given back in 1945. It has remained 'China' ever since, just not part of the PRC. The people are ethnic Chinese and in fact quite a few people fled the communist forces on the mainland by moving to Taiwan, including the government itself. One example being the founder of TSMC. Another example, Foxconn's founder was born in Taiwan right after his parents fled from the mainland.

I find the comparison with Korea and Vietnam factually incorrect and rather disingenuous, irrespective of what the people of Taiwan might think of reunification.

you don't have to look at the history of Taiwan. all you have to do is look at all the recent election in Taiwan. Taiwanese reject reunification.

Seems unlikely considering everything that's happening in HK.

Of course they would be involved. But they would not attack China directly.

It would be insanity for the US to declare war on China (and they know that), and I am always puzzled (and slightly worried) by the number of American commenters who seem to think that the US can 'just' attack China.

It's amazing how we're making lofty long-term projections, including defense strategy, Taiwan's sovreignty and global state of affairs from something that wasn't true just 5 years ago. TSMC has lead in development right now. Intel isn't dead and it can catch up, I am hoping just as TSMC caught upto Intel. There is no guarantee that TSMC might get stuck at a node in future for Intel to catch up. Neither is there any guarantee for Intel to find a technology breakthrough and make progress in leaps and bounds.

Remember, there was no tick-tock cycle before Intel invented the 2 year cadence back in ~2008. It was a business strategy to push alternate cycles of architecture changes and process node changes. It was conceived and invented by Intel. The point is that the cadence the entire industry judges itself was coined by the leader of the industry at the time. Now, the leader's podium position isn't at the top.

>Remember, there was no tick-tock cycle before Intel invented the 2 year cadence back in ~2008.

There was no Tick Tock then, but there was Moore's law. Intel unable to improve their node and manufacturing has nothing to do with Tick Tock cycles.

>There is no guarantee that TSMC might get stuck at a node in future for Intel to catch up.

They did. 3nm is no longer using GAAFET. And they have a plan B with 3nm using FinFET.

We judge a company by its past, and present experience. TSMC has always been an open, transparent and conservative company. Much like ARM, they are not there to take the spotlight. They were suppose to be in the background helping companies, it was only in recent 5 years Main Stream Media decided to hype it. Mostly because of Apple.

1st place is only part of the equation. They are a very important player, have been for some time and many big companies depend on them and will still depend on them even if Intel gets its stuff together.

But global dynamics seem very interesting currently. Big tech influence on political issues and wars have been big for some time but is becoming very visible now. Taiwan would be a very different country without TSMC. Heavily optimized for cost and speed and less on reliability supply chains and mining industries.

It seems hopeful to me that future wars will mostly be cyber and economical, it could mean less people suffering but it's not granted.

Every major bloc now has significant investments in reshoring indigenous semi capabilities, policy discussions dance around the fact that diversifying would be Taiwan's doom. There is limited timeline where TSMC will serve as security shield for the island.

I see a lot of discussion here that US would glass TSMC fabs if China ever attempted takeover, when realistically China would be the ones to disrupt TSMC proactively and prematurely if PRC indigenous semi industry reaches useful levels. Catching up to EUV is much harder than undermining it. Chinese military modernization = the first island chain strategy is dead. Taiwan never got security guarantee precisely because other strategy to contain China existed (malacca / sloc chokepoint), TSMC dominance is an aberration that no one apart from Taiwan is interested in. Hence I wonder how smooth the Arizona plant will go, possibly foxconn Wisconsin redux, but more pressing.

Of course TSMC makes Taiwan important for the US, so in that sense it protects Taiwan.

But it is also a big game of chicken. If the mainland decided to invade (and they are not in a position to do so) the US would not start WWIII by going into a direct war against them.

There is always another target after if you don't defends your allies.

That's not how it works.

First: US make a lot of promises to a lot of countries. If they are not willing to go to war over Taiwan, that makes most of those promises (like NATO pact) basically worthless. So US has to defend Taiwan.

Second: Military confrontation between US and Chinese forces does not mean WWIII - they won't start shooting nukes, it will be more conventional war. We had such examples during previous Cold War: Korea, Vietnam, Afghanistan.

All three of these examples (Korea, Vietnam, Afghanistan) were at the time of their "conflicts" impoverished, agricultural economies. Korea, Vietnam and Afghanistan were more important to US allies than to the US. Taiwan, on the other hand, is a modern economy, wealthy and highly integrated with the US economy. Fortunately, all three sides of the Taiwan situation understand this, and because of that, nothing is going to happen (hopefully).

I'm not even sure it would be a conventional war. More a digital and commercial war I would bet (which is already happening I guess). Something like the US and their allies stopping to purchase anything from China, if that's even possible. Look at what happened in Ukraine, Russia got a few sanctions from EU, but as far as I know, they are still trading with them, purchasing their gaz.

> If they are not willing to go to war over Taiwan

They are not willing to do to war over Taiwan. The opposite would be insane. Equally China is not willing (and able) either.

This is a balance of threats with both sides knowing where to stop.

The US are obviously not stupid and they never committed to intervene against China. They have just committed to 'help' and make sure that Taiwan can defend itself, which is very different.

> Military confrontation between US and Chinese forces does not mean WWIII

It does when the US directly attack the Chinese over Chinese territory.

I.e. the US would be declaring war on China. I would not let anyone who does not see the insanity of that prospect have a job in government.

> I've always thought that TSMC's dominance in fab processes are a strategic asset in the defense of Taiwan against potential PRC encroachment.

You’ve always thought what gets mentioned every single time tsmc is discussed. Interesting.

TIL: TMSC partly comes from a guy not getting a promotion at Texas Instruments and Taiwan wanting to getting in the semiconductor industry at the time with nothing but a plan.

>In 1983, a 52-year-old senior executive at Texas Instruments was passed over for the company's top job.

>He would go on to found and build the most strategically important company in the world.

https://twitter.com/SahilBloom/status/1353369463190560773

Ironically, that's what Mong Song Liang who was passed over for the top position at TSMC and now heads SMIC in China, has been doing.

He's also the one that got samsung where they are with semi fab.

He was responsible for Samsung's EUV development?

Odd..I live in AZ and local news is reporting the Fab at $12B, and the total project at $38B over 20 years (including 2K new jobs(1)(2)

also, they want to have it production ready by 2024, which seems incredibly tight timeline. My understanding is even the air handlers have to run for like 6 months before production can start.

Anyway, I'm hoping this will give the local economy a nice post-covid boost.

1: https://azbigmedia.com/real-estate/taiwan-semiconductor-laun...

2: https://www.taiwannews.com.tw/en/news/4089176

edit: formating

Well, the problem with these factory $ value figures (and I wish reporters would dig a little deeper) is that it doesn't really correctly depict the impact to the local economy very well. Either in the construction phase or ongoing operations phase.

In some ways the construction is little different than building an office park or warehouse, modulo some very special installed equipment and ventilation / power systems. Most of that $xB capital invested is going to the companies that make the extremely sophisticated line equipment, and that is not local.

As for the operations phase, it might employ 1,000 people, optimistically?

Not saying it's not a benefit to the Arizona economy. And certainly the economic calculations for something like this are far better than the bullshit $ thrown around for a sport stadium or team, for example. Just saying that the $12-$38B figure is not what Arizonans will feel.

Long term, however, it'll attract high-yield jobs. These jobs (and the people working on them) will have a high impact on employment. People who have high salaries can afford eating out regularly, Uber/Uber Eats and expensive custom made stuff. This helps the lower strata of society.

They also demand better schools and have kids who largely end up being educated and are anchored to the region.

yeah - there will definitely be a huge equipment budget. But the construction work will be substantial as well. I have a friend that's doing pipe fitting at the Intel plant, and he's looking at being booked solid for like the next 5 years. Not to mention all the road improvements, etc. the local municipalities have committed to.

I believe they're predicting more like 2K jobs for the operations phase.

Though, the 2019 phoenix economic output was about 260 billion. So even the full 1 billion / year is just a half percent increase.

Still I'm optimistic this can help PHX recover from the covid slump, and start getting the local economy back on track.

From the article:

A $12 Billion Plan

"Approved an investment to establish a wholly-owned subsidiary in Arizona, United States of America, with a paid-in capital of US$3.5 billion (approximately NT$99.75 billion)," the statement by TSMC reads.

From the first article I linked above:

"a new North Phoenix manufacturing facility and up to 1,900 new local jobs over a five-year period. With an estimated 20-year economic impact of $38.2 billion, TSMC represents Arizona’s largest-ever foreign direct investment."

From the second article I linked above:

"ahead of its planned US$12 billion wafer fab north of the city. "

So there seems to be some confusion...

Edit: ahh the OP seems to be referring to just the first payments. The Toms Hardware article goes on to say:

"In total, TSMC plans to spend $12 billion on its fab in Arizona from 2021 to 2029. The factory will employ about 1,600 people directly and thousands more indirectly."

Edit: add quote from first link

I've asked this question before but didn't get a good answer. Suppose I had $100M to blow. Could I build a fab in the USA that produced chips at, say, 65nm? I understand that a cutting-edge node like 7nm would be out of reach at that price point, but I'm curious what exactly is in reach.

Well I reckon for 100M you could fabricate a chip at nearly any process size you want but probably not at a speed at which you could sell them for a profit. For example many universities have the capabilities to fabricate their own chips for research and development.

It also depends on what type of thing you're trying to fabricate. For example in a phone: cpu, ram, power, radio and sensors might want their own unique fabrication. It gets even more crazy when you want to go into things like automotive or industrial power.

I'll over generalize things and say that you can probably make a fab that goes down to 90nm without too much issue. It was the 65nm node when the lithography started to need something different and new chemistries were required for the fabrication process due to leakage becoming and issue that could no longer be ignored.

At least for me it was the jump to 65nm when I remember things getting hard and staying hard. The cell deigns had to change, the layout took twice as long, the models for simulations were always off and tapeing out to different fabs has vastly different results. It seemed like very node after that the documents that arrived from the fab about the new process were at least twice as thick as the last.

SkyWater bought a primarily 200nm, but 65nm capable fab in Minnesota for 30M in 2017.

https://www.eenewsanalog.com/news/ceo-interview-sonderman-sk...

Granted, it's tooled for 200mm wafers (300mm are the standard right now), but it's still significantly cheaper than expected.

Prices are likely to ramp up really quickly when you get to the 65->45nm jump - that's when Intel started using multiple patterning.

And that was an amazing deal given the glut at 200mm being apparent even back then.

The reason you aren't getting an answer is because you're asking about something that isn't mass produced. So every single build is going to be a contracted cost with many variables. Saying "I want a 65nm fab" doesn't mean anything. What type of EUV scanners do you want, how large of wafers do you want to handle, how many wafers at once, size of the clean room, etc. There is no answer, it's whatever you could get someone to do the job for + materials cost. We only have estimates for cutting edge facilities because we know what public companies have recently spent to produce those fabs.

It's like asking "if I had $100m, how large of a secret underground meth production facility could I have built"....who knows, it's bespoke.

Now I'm just wondering what would $3.5B meth lab look like? it will be breaking all records for sure

Umm..., a pharmaceutical company?

There is a list of fabs, including plant cost, here:

https://en.wikipedia.org/wiki/List_of_semiconductor_fabricat...

The few I can see within your range seem to be very specialized (e.g. Murata) or educational (SUNY). Maybe have a closer look at

https://en.wikipedia.org/wiki/SUNY_Poly_CNSE

Some light research shows that it would cost closer to 3-4 billion dollars to build a fab from 2014[1] This makes a good bit of sense since the clean rooms required are massive and require a good bit of work to get right. Also from my experience with auditing PCB manufacturing in the united states. 100M might be able to get you a old TI Fab to make some TTL Parts at.

[1] https://en.wikipedia.org/wiki/Semiconductor_fabrication_plan...

Do you happen to have any resources (checklists?) On what auditing PCB manu. entails ?

Thought I would point out that The Economist has two articles on this topic in the latest issue. The geostrategic nature of the TSMC fab is only going to increase in the coming years. The articles explain the context to people like me who did not think much about the recent developments.

1. https://www.economist.com/leaders/2021/01/23/the-struggle-ov... 2. https://www.economist.com/business/2021/01/23/chipmaking-is-...

This factory is going to specialize in EUV and is "financed" indirectly by Apple.

You could say its financed by Apple, or your could say its financed by every user of a smartphone or GPU.

TSMC will manufacture relative low volumes of chips (at presumably high prices) chips for industries which require domestic sourcing like the military so its not going into an iPhone.

If Apple didn't buy up the 5nm capacity, the next highest bidder would have for slightly less than Apple was willing to pay.

>>> the next highest bidder would have for slightly less than Apple was willing to pay.

What makes you think Apple pays top dollars? Apple is notorious for squeezing their suppliers to save their margin and they often use their sales volume as the main leverage to gain discounts. It seems the demand for TSMC is however so high that now TSMC doesn't have to offer such incentive for large customers:

https://www.tomshardware.com/news/tmsc-is-reportedly-termina...

Yup, if it means Apple needed TSMC more than in the other way around

It’s more complicated. Apple has the branding to have a higher probability of extracting a higher profit margins at higher volumes compared to other brands, so a different buyer is probably not able to guarantee the level of payments Apple can to TSMC.

I don’t think it’s useful to think of it in terms of Apple needs TSMC more or the other way around. They’re both benefiting quite nicely from the cooperation.

But not 3nm technology. That doesn't leave Taiwan.

Only if ASML in The Netherlands allows it; they control the most critical bits of Taiwan's fab tech.

No, only if the US allows it. ASML has no say in the actual export restrictions.

And when the US asks NL to jump, we ask how high?

There's a story about how in the mid 20th Century Russia asked to buy one of Britain's cool new jet engines and Britain obliged. Then Europe found itself worrying about MIG fighter jets with essentially that British jet engine inside.

Once bitten, twice shy?

Last time I checked the process is very water intensive. Where does one get enough water in Arizona?

> Last time I checked the process is very water intensive. Where does one get enough water in Arizona?

As I understand it, modern fabs are supposed to reclaim and reuse the water, very little should end up being poured down the drain.

Trying to imitate Intel tech? https://ktar.com/story/76977/intel-plant-replenishes-water-s...

"Each day, 2 million gallons of industrial wastewater are piped from Intel plants in Chandler, Arizona, to a facility a mile away where it’s treated, then returned to an underground aquifer. Intel, the city’s largest employer, recycles about 60 percent of its water" So, 40% is not recycled, which seems high? given that "the leftover effluent[s] or concentrated dissolved salts that result from making chips" are claimed to be filtered with reverse osmosis before being returned to the aquifer. It is also disappointing that it seems to be more cost effective to have the RO instead of an evaporation + optional vapor collection setup to get 100% recycled or at least to collect the salt waste given the hot dry ambient desert location would seem like the most cost effective place to do so.

you sure you want heavy metal salts just laying around in a huge lake of brine? not sure if this is a workable proposition.

Personally, I am not in a position to be cognizant of all the solution options and their considerations. I did not intend to imply that current solution was inappropriate, but only express my feelings towards the current in the face of likely alternatives. In regards to an evaporation pool, it too has risks such as the heavy metal salts being turned into dust and blown across the land if not contained and managed.

As a resident of Arizona, this is especially good news. It is also good news for the US in general. Advanced chip manufacturing is a key strategic concern.

There's tons of advanced chip manufacturing in the USA. There are fabs all over the USA for companies like Intel, Infineon, Texas Instruments, ON, NXP (Motorola). It's the country's fourth largest manufacturing export.

TSMC leveraged a huge gap in the market early on. Since most fabrication plants were owned by companies that also designed and sold chips, it was hard to get contracts for production volumes. This made it basically impossible to be a fab-less design company, like nVidia, because Intel/TI/etc might not produce your chips if they didn't have spare capacity.

With a foundry, that's not a concern.

This is from Nov 2020

Even if this is a small fab, it's a win for the USA in a way I don't see mentioned: it creates more fab jobs and experience in the USA. This in turn will improve the employee pool here for this particular skill set.

It also means that if a nightmare scenario happened we'd have a TSMC 5nm fab to reverse engineer China-style.

With modern fabs, how much of the operating cost is labor?

Its mostly equipment and design. The fab has to be extraordinarily well-protected against seismic disturbance, and the equipment is highly, highly automated. This will produce jobs, of course, and high paying ones at that, but it won't be a "game changer" for the state. That city, possibly.

EDIT: Article says 1,600 for the fab when its at full capacity and "thousands indirectly", which of course means restaurants and services nearby.

Looks like its going to be built north of Peoria, which puts it near Glendale, Phoenix, Tempe, Scottsdale, etc. This'll probably add jobs worth about $100,000,000 - $150,000,000 a year. Somewhere in that neighborhood. The support-based jobs probably $150,000,000 to $200,000,000. Adds another $8.5 to $28 million a year in tax revenues for the state and cities.

This actually surprised me, as I would have guessed it would have been closer to Intel, to make use of infrastructure already in place..

Also, not sure how the jobs will break down, but when I was at Intel, techs actually working in the clean rooms were not making anywhere near 6 figures. They commonly told me, they were just there for the stock.

I'm always surprised more American cities don't pander to host high-end factories. I don't know what just chips would do for the city, but I imagine hosting something like a Samsung phone factory in exchange for a massive local discount on the product would be a pretty nice perk for the citizens.

Things like smartphone assembly aren't going to happen in the US any time soon. Not only is the labor more expensive, but essentially the entire supply chain is in eastern asia.

There was the Foxconn Wisconsin thing which seems to not really turned to much as of yet.

So far it seems to have been a bit of a con: https://www.theguardian.com/us-news/2020/dec/08/wisconsin-fo...

>I'm always surprised more American cities.....

Electricity, infrastructure, Additional cost in Tax, Environment protection etc. All of these add up very quickly.

It is not impossible. But it doesn't seems Government cares that much other than talking about it. There is also the close proximity with other supply chain. Where there are currently close to Zero in US.

Reply on news.ycombinator.com