Hacker News new | past | comments | ask | show | jobs | submit login
Sony to join TSMC on new $7B chip plant in Japan (nikkei.com)
679 points by thedday 53 days ago | hide | past | favorite | 216 comments



"Bloomberg noted that India is currently studying possible locations with adequate land, water, and manpower resources. India reportedly said it would provide financial support by fronting half of the capital expenditure needed from 2023, along with tax breaks and other incentives."

India is going to put up half of the capex up front w/tax breaks and incentives. India govt. also going to scout out land for Taiwan.

TSMC will be in Taiwan, China, Japan, US, and India.

https://techhq.com/2021/10/heres-why-a-mega-chip-deal-betwee...


India's first forays into semiconductor fabrication in the 80s and 90s were likewise enthusiastically supported by the government (land, incentives, tax breaks and so on), but were ultimately hamstrung by more fundamental infrastructure issues that couldn't just be magicked away - water shortages and unstable power grids - each of which could grind manufacturing to a halt for months on end and delayed production cycles. (I think there was also a major fire in a leading SC plant that caused delays by years.) If anything these shortcomings could be exacerbated in 2021-22. The government would need a much more comprehensive infrastructural solution this time around.


Even so, it's not hard to see why India remains alluring for tech companies. Wages are low and there's a massive number of young workers with more on the way. In these respects, India is, now, what China was a few decades ago. Plus, it's right next door to existing supply chains and less encumbered by international politics.

If the problems can be solved, the returns will be great.


The last time that China's government was instituting economic policies that threatened the trade of the world's premier power (by hoarding silver and holding a monopoly in tea) it was solved by the British government and its agents diversifying tea supply via India and being far more militarily aggressive, which does seem to be mirrored somewhat by recent events.

If Facebook et al are truly addictive then I could make the stretch to that and the exportation of opium over foreign borders, but that's more of a worldwide blight.


I like that metaphor that Facebook could be this time's opium export. The US exported intellectual opium, but we suffer from it too. We did it to ourselves though in the us, instead of someone making us use it (or did russia push into this along with cambridge analytica - this crazy example almost pans out).


China is a lot more open this time around. Their silk/belt-road initiatives, span across continents.

What you are describing actually applies more to USA than China. Its USA which is a closed country and shutting out anyone it has a mild disagreement with(a.k.a sanctions).


Err... the remarkable thing about the US recently is that it is using sanctions at all. Not that it is "shutting out" people.

In comparison with China, the US is one of the most open countries in the world. China is horrifically closed-off to investment, imports, and so on.


> China is a lot more open this time around. Their silk/belt-road initiatives, span across continents.

Where to start ...

You realize that BRI doesn't have grants, just debt, right?

The purpose of BRI is to build ports for the CCP navy, but have somebody else pay for them.


Can USA corp like Facebook, YouTube, ebay gets into china. Can china wechat, …

Fact ah. USA is a closed country. Great 50 cents or what.


The clincher is whether it can be solved.

Infrastructure is so crucial to every society, every process, and every organization and yet it is also so often not given the attention that it needs. .


Plus, it's right next door to existing supply chains

Not as the boat flies. To be fair, I know nothing about shipping logistics... but it seems like quite a detour to me.


> Wages are low and there's a massive number of young workers with more on the way.

Please, check again. Indian engineers are by no means cheap.

They had decades long experience working with the West. They know American wage levels well, and bargain to the end.


> [China is] less encumbered by international politics

in what way(s)?


Agreed. They were absolutely hamstrung because their facilities burned down and were only restored to s shadow of their former selves. They are still generations behind everyone else, and not able to use their anywhere near their full capacity. Plus their competitors took the opportunity to hire the best people that India away from the state fabs. Shame really, if they could increase their output, this shortage would be the time to regain some footing.


>>water shortages and unstable power grids

In Bangalore the tech parks most of the time run on diesel generators. And the water is kind of de-facto supplied through tankers. There is a fair bit of rain water harvesting but they are not putting in enough to use that through borewells.

My guess is something similar will be worked out this time as well.

South Bangalore/Hosur seems like a nice place for something like this.


As much as I would love to agree (since I have a home in Bangalore on the route to Hosur), it's not really possible. Bangalore suffers from severe water shortages largely exacerbated by developers building over former water harvesting reservoirs. You'll need way too much power and water, that it would just be cheaper to shift manufacturing elsewhere (to, say, Pune).


Given how bad a power fault is on a fab, chemical processes that can't be interrupted and air filtration and pressure systems, I'd put a redundant Tesla battery as the UPS to a fab, anywhere.


I would install large diesel generators well before anything from tesla if I wanted a reliable power supply


Generators would have to kick on and over, with a battery backup it could be pass through. Tesla was used as a representative example.


Actually any decent backup system uses combination of both ...


Run on generator the whole time if you have an unreliable power system like India.


i was thinking how come India didn't take China's approach to prop up their own foundry like SMIC. since India seem to have a need for semis. from the article:

"India’s semiconductor demand is said to be valued at around US$24 billion and is expected to reach US$100 billion by 2025. The country’s semiconductor demand currently is entirely met through imports."


How much of that is replaceable though? Most of those are Intel/AMD/etc chips that you can't just drop in another foundry to replace.


>India's first forays into semiconductor fabrication in the 80s and 90s were likewise enthusiastically supported by the government

https://www.youtube.com/watch?v=isBYV6QWDIo


On this topic I recently came across this very good video on this https://www.youtube.com/watch?v=isBYV6QWDIo


Good way to de-risk the possibility of a CCP invasion of Taiwan destroying world wide chip production.


Only leading-edge chip production. There are tons of chips made by other companies.


Leading edge chips go into the highest margin products.


Wait, China uses TSMC? I'm surprised PRC govt allows that.


China's official position is that Taiwan is a province in rebellion, but economics mean they can't treat it exactly like an ongoing civil war so... It's complicated


More like confederate vs union each permanently settled a chunk of land but with no ceasefire nor surrenders


China and Taiwan are very very closely economically and culturally linked. The PRC and ROC governments do work together, it's no longer the 1970's.


> it's no longer the 1970's.

Oh? https://www.bbc.com/news/world-asia-china-58854081


To be fair, the RoC agrees completely with Xi Jinping.

They just disagree on who should run the reunified nation.


That's not accurate. My understanding is the KMT agrees, and DDP doesn't (specifically: 1992 consensus), and the fact that the population opinion has been shifting towards independence is the primary tension currently.

At any rate, summing it up as "the RoC agrees" would seem to simplify a cultural argument along the lines of "the US agrees completely that guns are good".

For example: https://www.scmp.com/comment/opinion/article/3151551/tsai-in...


> In response to Xi’s speech, Taiwan’s presidential office said most Taiwanese reject that model and argued that developments in Hong Kong show how “one country, two systems” can turn on a dime.

> On Sunday, at an event marking the same revolution, Tsai called on the Taiwanese people to renew their commitment “that the Republic of China and the People’s Republic of China should not be subordinate to each other.

https://www.washingtonpost.com/world/2021/10/10/taiwan-china...

Tsai has not made this one country position part of her administration, and her speech a couple of days ago made that even more explicit.


That’s only their official position and they can’t change it because that’s one of the PRC red lines. They consider Taiwan dropping their claim on the mainland as a change in the status quo and a casus belli. Everyone know the RoC doesn’t aspire to a reunited China anymore.


So, officially, they agree with Xi Jinping.


They say what the PRC forces them to say but de facto don’t believe in it. It’s a complete misrepresentation to argue that Taiwan agree with China regarding reunification but with reversed roles.

Taiwan is a democracy and a majority of Taiwanese want independence but argue for keeping the status quo in order to avoid a war (both the KMT and the DDP - their disagreement is more technical than that). Some wanted Taiwan to declare it in the 90s to force the hand of the USA and win a decisive war but this position seems more precarious nowadays.


No, its not. If I point a gun to your head, and tell you that I am going to shoot you, if you do not "agree" that the moon is made of cheese, you are not actually "agreeing".

We both know, that you do not agree that the moon is made of cheese, in this situation. I have just made words come out of your mouth, to something that you do not agree with.

To say that they "agree" is just a silly word game, that does not reflect the truth of the matter, and instead is playing into propaganda that denies the reality that Taiwan is already a country, and is already independent of china, and that Taiwan is not interesting in being taken over, or taking over china.


Nice goalpost move from "completely" to "officially" (under duress).


What?! One country two system is definitely not something ROC agrees with.


It’s no longer the 70’s as it appears we are heading to the 50’s or worse.


The PRC is not a monad like most hit piece journalism and armchair geopoliticians would like to make you believe to sell you their easy to implant ideas. The ROC does behave more like a monad if that's a thing, but on the PRC side the military and the banks don't even share the same plan for dealing with Taiwan.

I agree on the cultural and economical aspects both governments work together but I am not sure that on other aspects tensions are not as high as in the 1970s or worse. I wish they were not, I just don't know.


I thought I knew what a monad was and the only problem was all the tutorials and other people clearly struggling to understand it. But now I’m not so sure.


That made me chuckle. But I think the GP is using the term "monad" more in the sense of Leibniz's philosophy than that of functional programming or category theory – "monad" as meaning ultimately simple and indivisible, as "atoms" were in the original ancient Greek atomic theory (as opposed to modern atomic theory in which the so-called "atoms" turned out to not actually be atomic after all). Of course, even in that sense the GP is using the term figuratively – nobody literally believes that China is a single indivisible entity, a hive-mind or Borg, but the GP is claiming that Chinese society (and even the Chinese government) contains more divisions of opinion and interest and attitude than many outside observers assume. And I'm sure there is some truth in that – but, I think the GP is wrong in suggesting Taiwanese society is different – that is just as true of Taiwanese society, and Taiwan being a democracy puts these differences more out in the open (DPP vs KMT etc), China's more closed system means many of these differences exist behind closed doors; and even if sometimes people inside China get away with speaking of some of them openly, they have to be careful what they say and how they say it, to a much greater extent than people in Taiwan have to


Thanks to XJ, tensions are high again. All of Deng Xiaoping's hard work and planning have been thrown away by Xi.


That's a very unnuanced, MSM kind of take. As a Chinese, I am very disappointed at the absolutely one-dimensional analyses of western China watchers. Their takes are rarely accurate, often full of ideological bias, and they don't help me understand China better at all.


And your comment isn't "one dimensional" and "unnuanced"?

As a fellow Chinese person, it's pretty accurate. Deng Xiaoping deserves the credit for modern China's economy. China was slowly but surely opening up and slowing becoming decentralized for efficiency, and it was a moderate rule by committee with term limits vs a one man dictatorship for life that's veering into a centralized economic planning disaster reminiscent of Mao. He is destroying China, one industry at a time.

I can understand your "patriotism" if you're commenting from the mainland. Not so much if you're writing from elsewhere.


> it's pretty accurate.

Not remotely.

Xi was selected to fix Deng's unequal growth phase strategy. Hence all the regulations and crack down because Deng's model ran it's course. And Deng's policies only worked due to the base Mao built. Cultural revolution secularized and mobilized human capital, even great leap forward was massively successful in building industrial base - PRC in the 70s was less urbanized but considerably more industrialized than other low income peers. Which facilitated Deng's transition, which wouldn't have worked in India trapped in multiculturalism, historic power structures and rural economy. It only worked because Mao made sure everyone spoke mandarin, society had no strong attachments to history etc.

>He is destroying China, one industry at a time.

He's regulating previously underregulated industries. He's doing what the west talks about doing in policy papers for years but systemically can't. Sino-US ideological competition is battle of political systems, and so far PRC's has demonstrated to be more nimble and focsing on correct priorities. It's why US and others are trying to copy PRC's industrial policy, which PRC once copied from west. Also the fact that Xi managed to build SCS islands and modernize military sufficiently that US now discusses greater power competition in miiltary terms handly justifies his leadership. BoXiLai wasn't up to that task. The amount of comprehensive national power Xi built is insane. He's PRC's 3rd and maybe 4th term FDR in time of remarkable opportunity and crisis. He may sound and look dopey as hell, but his results have been remarkable.


> Xi was selected to fix Deng's unequal growth phase strategy.

Xi outmaneuvered his rivals. I highly doubt that the others wanted to forfeit their own power for his. Deng's policy was also already proven to work. Expanding it and not destroying it would have been more logical and responsible.

> And Deng's policies only worked due to the base Mao built. Cultural revolution secularized and mobilized human capital, even great leap forward was massively successful in building industrial base

That sounds delusional. Mao's policies were a complete disaster that killed millions of Chinese people. It didn't prepare China for the future. The only things it accomplished were massive suffering and international embarrassment. It was a complete mess. That isn't to say that Mao didn't achieve anything good, but the Great Leap forward and the Cultural Revolution are not good examples. If anything, that greatly hindered both Zhoe Enlai and Deng's reforms, and that's putting it lightly.

> He's regulating previously underregulated industries.

He is centralizing control and power. In his new order, there's no room anyone exceptional in private industry, or really anyone else. There's only room for his cult of personality. He is slowly regressing China into a Mao style of rule which completely throws out most of the lessons from the past: decentralization is more efficient and yields more production vs a centrally planned economy (relics of 20th century Communism)

> PRC's has demonstrated to be more nimble and focsing on correct priorities.

How can you be nimble when you're re-centralizing everything?

> It's why US and others are trying to copy PRC's industrial policy, which PRC once copied from west.

I don't know what you're trying to convey here.

> Also the fact that Xi managed to build SCS islands and modernize military sufficiently that US now discusses greater power competition in miiltary terms handly justifies his leadership

In a world with global ballistic nuclear missiles, it looks like Xi just copied a bad US habit, which is a source of massive corruption and waste.

> He may sound and look dopey as hell, but his results have been remarkable.

If it walks like a duck, and sounds like a duck; it's a duck.

> his results have been remarkable

This is still up in the air and highly debatable. For example, Xi is extremely impatient. One China Two Systems was working until he decided to completely trash it. We can assign the mess in HK and Taiwan's new calls for independence on him. If he didn't interfere in Deng's policy, China would have quietly completely swallowed both in a decade or two.


>Xi outmaneuvered his rivals.

He outmaneuvered them by being compromised candidate that no one had any objection to. And no one objected because Xi was the kind of person LKI compared to Nelson Mendella calibre of person. CIA dossier concluding Xi's incorruptible by money suggests Xi had the reputation to and disposition to negotiate the gilded age left by Deng's policies that Hu couldn't stem. Xi had the qualities to at least direct PRC in right direction vs Bo.

>that sounds delusional

PRC can trade lives for progress. It's harnessing the capital part of human capital. Bluntly frontloading a few million deaths to quickly set up the right conditions for future growth is the correct decision. Of course there were missteps, but overall Mao did the right things considering external factors. Again, look to India for a comparable alternative. It's a fucking mess because it can't break eggs to create omelette. Mao's mess was MASSIVELY beneficial, even if it led to excesses like CR and GLF, which bluntly are drop in the bucket tragedies on PRC historic scales.

>He is centralizing control and power.

A competent leader gaining power during interesting times isn't bad. It's not about efficiency but building comprehensive national power, many things simply do not get built or changes cannot be made without massive state directed attention even if less efficient. Recentralizing is not bad either - it's industrial policy - which is what made the west strong in the first place. Policies US abandoned and now is trying to rapidly regain. The ability to recentralize when needed is relative nimbleness the alternative is institutional stagnance west finds itself trapped in.

>Xi just copied a bad US habit ... > If it walks like a duck, and sounds like a duck; it's a duck.

Xi modernized the PRC military on a modest 2% budget. It's what NATO ought to spend. The amount of procurement PRC has made under Xi doesn't suggest there's massive waste - one of Xi's first efforts was cleaning out notoriously corrupt PLA brass, which Bo was associated with. Again he wasn’t up to the task, Xi was the best choice in the timeline we had. Matching US deterrence post pivot to Asia was incredibly difficult undertaking. Obama/Clinton already didn't buy Hide & Bide. Xi's a very impressive duck. Even centralizing power is a hard trick that requires acumen. Don’t have to praise it, but recognize that Xi got a lot of hard things done, fast.

> One China Two Systems was working

No it wasn't. HK existed for 23 years of NSL exception and being exploited by the west as the spy capital of Asia to undermine PRC interest - it was never a sustainable arrangement. HK only ever paid 1C2S lip service, and cultural shift as old demographics with memory of KMT/PRC relations die out, 1C2S wouldn't have been viable in TW either. No one had any pretensions that TW would return to PRC on a 10/20 year timeline. 1C2S ran its course like Deng’s development model or hide&bide and (eventually) no first use, 1C2S was a policy out of weakness, there will be alternative models as a result of PRC gaining strength. Xi is nimble enough to move past uneasy and untenuous policies. While the future of TW is debatable, Xi successfully secured HK, XJ, Tibet and built enough hard power to challenge the US when Trump decided to set eyes on PRC after covid drama. Wolf-warrior last year showed very few countries willing to step on PRC toes on internal issues, that PRC can effectively do what she pleases on core interests with relative impunity. Again, no small part thanks to Xi’s military modernization efforts. Xi has been great for PRC national interests.


> Xi had the qualities to at least direct PRC in right direction vs Bo.

I agree that Bo wouldn't have been much better than Xi, but that doesn't mean that Xi is a good statesman. His incorruptibility is marred by poor decisions that make little sense unless you're the Chinese version of Trump (Maduro & Chavez are probably better comparisons).

> PRC can trade lives for progress. It's harnessing the capital part of human capital. Bluntly frontloading a few million deaths to quickly set up the right conditions for future growth is the correct decision. Of course there were missteps, but overall Mao did the right things considering external factors. Again, look to India for a comparable alternative.

Just because India has its own mess, it doesn't condone what Mao did. India's problems also pale in comparison to the man made disasters of the Great Leap "Forward" and the Cultural Revolution. A lot of lives were lost, but no progress was gained unless you consider it as a step in controlling the overall world population. Instead, it just caused chaos and national humiliation. If anything, it was 1000 steps backwards.

> Recentralizing is not bad either - it's industrial policy - which is what made the west strong in the first place

It is not what made the West strong. Capitalism is about decentralizing control, and letting the more efficient markets decide where to allocate resources. It's imperfect and there's still waste, but overall it's far more efficient than centralized economic planning. Centralized economic planning is what caused the Great Leap and the Cultural Revolution. All it takes is a few bad ideas and absolute control. There's plenty of historical data to bolster my position.

> Xi modernized the PRC military on a modest 2% budget. It's what NATO ought to spend. The amount of procurement PRC has made under Xi doesn't suggest there's massive waste

Yes, I totally believe that coming from an authoritarian government with little to no transparency. /s

> one of Xi's first efforts was cleaning out notoriously corrupt PLA brass, which Bo was associated with.

The corruption elimination campaigns are just thinly veiled efforts to eliminate political rivals. If officials really wanted to eliminate corruption, you would need transparency. Transparency only comes with freedom of speech and a democratic government.

> Matching US deterrence post pivot to Asia was incredibly difficult undertaking.

While US is still powerful at the moment, only the Western equivalents of you and FooBarWidget would fail to admit that it's been on a steady decline for years now. Not to mention that it's extremely overstretched by being World Police, and politically polarized as a divided nation. The TPP failed. Literally the only thing China would need to do is wait. Nothing immediate needed to be done. The only thing Xi has accomplished is to galvanize rival nations to better cooperate with each other. Genius move just like his mishandling of HK.

> No it wasn't. HK existed for 23 years of NSL exception and being exploited by the west as the spy capital of Asia to undermine PRC interest - it was never a sustainable arrangement.

Is this a joke? The Lost generation was just a vocal minority. Key emphasis on minority. The new generation can barely even speak Cantonese anymore. Again all you had to do was wait for the inevitable to happen. Yet due to Xi's 'genius', he's shifted formerly indifferent public opinion to support the Lost gen.

> No one had any pretensions that TW would return to PRC on a 10/20 year timeline.

Like HK and given the controls and policies set by Xi's predecessors, it was only a matter of time when Taiwan was so economically linked to the mainland, that meaningful opposition would all but disappear. Instead, Xi with his usual genius moves just helped the separatists make their point with his other genius decisions involving HK.


> decisions that make little sense

You keep alleging this but much of his decisions made perfect sense and help contribute to building national power. Something Bo wouldn't do, and things Hu failed to do. So in that sense Xi is at least more competent than Hu while also keep in mind Xi's policies in many ways proceeded his tenure. The different between other populist is Xi's China is unequestionablly stronger in relative power by almost every metric. Xi's successful.

> pale in comparison ... no progress was gained

There was progress. Destroying the old and uniting society under common national identity was paramount to enabling subsequent reforms. Destruction WAS the progress. It was the point. Sometimes destruction was excess, but MORE destruction was better than less. India unable to destroy and renew itself = India STILL has hunger problems on level of North Korea today. The result is additional generations of Indians suffering subsistent poverty hundreds of millions of excess and prevental death. Cost of GLP and CR pale in comparison.

> /s

Lazy. Plenty of literature out there affirming PRC industry and purchase power = massive buildup cheap.

> you would need transparency

I mean no? You don't? You've drank the democracy dogma koolaid. CCDI eliminated 2M+, Xi doesn't have that many rivals. Of course he's going to use it to consolidate power, because eliminating that much corruption require firm power.

> Literally the only thing China would need to do is wait...

Waiting is how PRC got embarassed in 3rd strait crisis and Belgrade embassy bombing with inpunity. Again, learn some history, US pivoted to Asia before Xi. Manifested in Trump pushing unconstrained anti-PRC populism. Xi couldn't wait because US was already plotting against PRC pre Xi. He had to answer, like every PRC leader during their Sino US crsis. Except Xi chose respond to US as equal. And it worked. The reason why western MSM is losing their shit is Xi's PRC can't be talked to from "position of strength" anymore.

> galvanize rival nations to better cooperate with each other. Genius move just like his mishandling of HK.

Except... none of this is true outside of western MSM headlines. Wolf warrior demonstrated that PRC barking = no countries except US will affect PRC domestic concerns except with theatre. Every one of these "cooperative" alliance has stalled or failed to affirm intent to contain PRC. Essentially everyone expressed neutraility or explicitly won't commit to military against PRC interests, including TW. Xi's asserting revealed how weak bloc against PRC actually is.

> support the Lost gen

Lost gen had enough political capita to block NSL and patriotic education and were gaining support. Now they're gone. It's not Xi being genius but doing what eventually had to be done. NSL free HK is too much liability in age of great power competition, not worth "waiting" for.

> meaningful opposition would all but disappear

That's not how it works. I suspect you know how many countries have poor opinion of PRC but strong economic linkages. Did economic integration swayed PRC closer to US orbit? TW generations that grew up under democracy like HK generations that grew up under liberal educationisn't going to clamor for reunication with PRC under CCP... which it will be for foreseeable future. It doesn't take a genius to read the room and know where things are going and how to respond. Xi isn't some genius sweeping the west with grand strategy, but he's doing a demonstrably great job doing generally the right things and the hard things for PRC interests.


> keep in mind Xi's policies in many ways proceeded his tenure.

You mean preceded. Since Xi has been declared president for life, when his tenure will be proceeded is indefinite.


> You keep alleging this

I back up my assertions. You make sound like I just make random statements like, "Xi sucks" without offering a reason why

> Something Bo wouldn't do, and things Hu failed to do.

imo Bo seemed to be a clone of Xi with some slight variation. He's a little smarter, but more deeply flawed in terms of corruption. If my memory is correct, both want a return to Mao's style of rule.

Hu was superior to both of Bo and Xi. He is the last of the Deng era: rule by a moderate committee based in pragmatism and not Xi's bullshit theatrics.

> Destroying the old and uniting society under common national identity was paramount to enabling subsequent reforms. Destruction WAS the progress. It was the point.

Destruction is only 'progress' for insane psychopaths. Millions of Chinese people needlessly died due to a dangerous mixture of hubris and incompetence. There was no progress. Families were broken. Society was broken. Even the very people who engineered China's revival as a superpower were in mortal danger. The terrible ideas in the Great Leap Forward were comedic like getting villagers to smelt different grades of metal in homemade furnaces.

Mao's opus was his brilliant Modern Theory of Guerrilla Warfare, which still affects the world theatre. He was a legendary military strategist. Unfortunately, he was a really terrible statesman. I've already explained why multiple times. He would kill people who reported problems, allowing these problems to fester and get even worse. That alone is terrible enough. You have to be extremely delusional if not brainwashed to believe that the Great Leap Forward and the Cultural Revolution were successes.

> India STILL has hunger problems on level of North Korea today

This is a good one coming from someone who's defending Mao's accidental genocide. I'm surprised that you're not calling it "progress". India has a lot of problems, but they also have a free press and transparency. It does NOT have the hunger levels of North Korea. North Korea is basically what would happen if Mao was an immortal ruler and there were no pragmatists like Zhou Enlai to eventually overthrow his flawed train of thought.

> Waiting is how PRC got embarassed in 3rd strait crisis and Belgrade embassy bombing with inpunity.

Embarrassment? It was a very temporary setback that everyone else in the world has more or less forgotten. The only people who really remember it is the US Navy. Even if Xi wasn't leading, I think we can both agree that it wasn't a matter of if China would surpass US naval power in Asia, it was only when. The Belgrade embassy bombing was clearly an accident. It's also yet another issue that everyone in the world has forgotten.

Do you know what Chinese embarrassments that people still remember today? The Great Leap Forward, the Cultural Revolution, Hong Kong, and Xinjiang in reverse order. Those are embarrassments that the world will remember for generations to come. Half of them happened under Xi's watch. As a Chinese person, Xinjiang angers me the most. I will never forgive Xi for that, especially because millions of overseas Chinese will most likely be subjected to danger because of it for decades to come.

> Again, learn some history, US pivoted to Asia before Xi.

Where did I show my ignorance on this issue? The US pivoted to Asia immediately after WWII because Imperial Japan caught most US leadership by surprise. The humiliations in Vietnam and Korea bolstered this position. You make this pivot seem recent, but it's ancient. Everyone already knows about it and why it happened.

> Except Xi chose respond to US as equal. And it worked.

Trump is an idiot insurgent with a temporary hold on power, as designed in a democracy. All you have to do is wait until US elites regain their power. The US is another country that is essentially ruled by committee. The only time to worry is if Trump somehow miraculously holds on to power.

Regardless, the US is on a course for implosion. Only a "genius" would have the bright idea of moving into the blast radius. For some reason showmanship impresses is what impresses Xi's fans more than actual pragmatic leadership with real gradual progress. I guess I was right in calling Xi China's Trump.

> Except... none of this is true outside of western MSM headlines.

MSM headlines barely mention it. I don't even think Fox News glamorizes it because it's boring to the masses in the West, so I'm not sure what you're reading. No one outside of investment bankers and maybe tech company leadership care about TSMC foundries being built in multiple countries.

> Every one of these "cooperative" alliance has stalled or failed to affirm intent to contain PRC.

Yes, I've already mentioned that the TPP, the only true danger that Xi faced, failed because Trump unilaterally killed it. I find it comedic that the only true recent threat to China wasn't even mentioned until now and you didn't even name it. If Xi didn't emulate Trump's bravado and idiocy, new threats would emerge much more slowly. Instead, now we're seeing new cooperation amongst rival nations, even the very ones targeted by the TPP.

> Lost gen had enough political capita to block NSL and patriotic education and were gaining support.

They did not, until Xi broke the One China Two Systems policy which was pure 'genius'. No one cared about their little movement until everyone saw them being abused in the streets. This was exactly what the West wanted and the blame falls directly on Xi. Without Xi, their little movement would have died out to little fanfare. Of course, Xi loves theatrics even when it hurts China's standing and stability. He is Chinese Trump.

> That's not how it works. Did economic integration swayed PRC closer to US orbit?

Under Deng's policies it did. There certainly wasn't as much animosity as there is now, but that's a poor analogy because you're speaking of two nation states and not two provinces, or are you saying that Taiwan is a country?

> TW generations that grew up under democracy like HK generations that grew up under liberal educationisn't going to clamor for reunication with PRC under CCP... which it will be for foreseeable future.

Pro-PRC parties were gaining ground until the mess in Hong Kong. If HK was left in peace to meet their destiny of being fully swallowed by the mainland, Taiwan would have soon followed. China's economic tendrils are too deep. Now you have many countries going out of their way to provide more economic support to Taiwan. That is directly Xi's fault.

> Xi isn't some genius sweeping the west with grand strategy

That's putting it very lightly. The CIA report was very accurate. With Xi, the question isn't whether a disaster will happen due to his bumbling. The question is when and how big. His comparison to Winnie the Pooh is also accidentally accurate if you ever watch the cartoons.


>I back up my assertions.

Not really. Your assertions have been Xi sucks / his decisions are bad, then quote XYZ trope about authoritarian governance from western perspective. It's dogmatic like "at least India has free press and transparency"... because democracy gud??? Reality is Xi's performance has been overall good for PRC interests.

> 'progress' for insane sociopaths

It was never the less "right" progress. Mao was bad statesman for right goals. Like I said, he was excessive and did the hard things sometimes poorly, but it was important to do them none the less. Ditto with Xi. Zhou corrected Mao's excess. Someone will correct Xi's excess. But they will also all build off the "necessary" foundation Mao and Xi built. PRC simply could not be where it is if population trapped by old culture and social structure.

> India does NOT have the hunger levels of North Korea.

Go look up hunger indexes over last 20 years. India is at North Korean level. Or excess death studies due to poverty. Hundreds of millions, mostly children dead. Because India is a fragmented multicultural society with broken free press and transparency that can't coordinate for their citizens well being. India is what happens when you don't have "immortal rulers" doing the hard, dirty but necessary work. Because democracy sucks for development. Which is why I'm fine with Xi doing what he does. Authoritarianism is the worst form of government for developing countries... except for all the others.

>everyone else in the world has more or less forgotten.

...and PRC? The actual party that matters. Consistently being preoccupied with non-PRC interest and non-PRC POV when discussing the PRC is asinine. Why does it matter what others or diasphora are embarassed by? They don't matter. XJ terrorism stopped. HK is firmly under PRC control. This makes PRC citizens happy, and forwards interest of PRC. GLF and CR at least got that 30% bad label. This applies to the rest of yoru comments re: conflating Pivot to Asia policy to... history of US in east asia, TPP as magic China containment meme, Xi not worrying about Trump 2nd term. You can call/compare Xi to who/whatever. Still doesn't change the fact that has been massive boon for PRC interests. He's not looking for your forgiveness.

>Under Deng's policies it did.

And how did that work out? A bunch of political and people-to-people linkages that ultimately revealed the extend of US infiltration into PRC political systems. Why do you think the new animosity exist? Because PRC under Xi, eliminated foreign NGOs, intelligence networks and other sources of foreign influence that dangerously upset the system - the cost result of drawing closer to US orbit. Security dilemma trumps trade at end of day. Again see HK. I don't care if people refer to TW as country like it's a gotcha.

> Now you have many countries going out of their way to provide more economic support to Taiwan. That is directly Xi's fault.

Who didn't affirm 92 consensus? PRC or TW/DPP. Or TW Sunflower movement in 2014 that protested against trade integration with PRC before Xi did anything. The undercurrents were there, Xi didn't create it. So far, TW support has been marginal, and only beacause TSMC dominantes (for now). It's something to be managed, but ultimately there's no cultural or economic victory with TW.

>CIA report was very accurate

Except... he hasn't bumbled outside of this imaginary CIA report. He may very well in time, because he's tackling the hard things previous leaders couldn't. And from PRC perspective of his track record, there's a good chance he'll do well. The amount of irrationally angry western reporting on diifcult reforms CCP can tackle but west systemically can't will probably tell you how well.

Look we're clearly not going to agree on Xi. I'll eat my hat if PRC collapses in 2028.


> Not really. Your assertions have been Xi sucks / his decisions are bad, then quote XYZ trope about authoritarian governance from western perspective.

Like you've done anything differently (Xi is great / his decisions are good, then quote XYZ troupe from PRC's perspective)? At least I'm doing it from the world perspective instead of just one country's POV.

> Like I said, he was excessive and did the hard things sometimes poorly, but it was important to do them none the less.

Mao didn't have to do anything. He could have at the very least delegated it to someone else more pragmatic.

> Go look up hunger indexes over last 20 years. India is at North Korean level.

You're right about the hunger index. I should have checked it. I can believe India's score. However, how is it even close to accurate for North Korea? Data isn't exactly easy to collect from the Hermit Kingdom. It could be much worse than the index score.

> Because democracy sucks for development.

India is a developing nation that doesn't represent all democracies. The US would be a much better example. Even then, it would be more accurate if we also look at the EU.

Less centralized economies, which is representative for democracies, are actually very good for development. One could see this at work before and after Deng's policies took effect.

> Authoritarianism is the worst form of government for developing countries... except for all the others.

Sometimes it works because of good leadership. Deng & his successors are the best example of this. I'm sure there are plenty more since dictators dominated most of human history. The issue with authoritarianism is that when you have bad leadership it's near impossible to get rid of them. They are there for life. Also any bad decisions get magnified 100x because there's no checks and balances of power. Democracy has a control for that: voting and term limits. For example, the US was able to neatly depose of Trump in four years without blood. You can't same about bad dictators.

> Why does it matter what others or diasphora are embarassed by?

Because the PRC obviously cares about saving face. Only barbarians would be indifferent to the developed world's uproar. The fallout for overseas Chinese is just a side effect, but it's still important.

> This makes PRC citizens happy, and forwards interest of PRC.

That's very debatable. How do we truly know when there's no free speech and free press? Data is not accurate within an authoritarian regime.

> Except... he hasn't bumbled outside of this imaginary CIA report.

He did and I've already listed his mistakes. Even you and the other guy admit that he doesn't seem very bright.

I think we're going around in circles. Your arguments are not going to move me, and vice versa. The second that you mentioned that the Great Leap Forward was "progress" is the second that this conversation became a waste of time for the both of us. We're grounded in different realities.


Okay I've been watching your back and forths for a while now and you guys seem to deadlock. I'll add some verifiable data.

Exhibit 1: "From Yao To Mao" by Kenneth James Hammond, history professor at New Mexico State University https://www.youtube.com/watch?v=9VMEFNcbR0o&t=8m30s This history lecture states that The Great Leap Forward actually achieved many of its goals. One famous failure was the famine. But you better listen to his very interesting explanation of that instead of reading a summary by me.

Exhibit 2: life expectancy from 1850 to 2020. https://twitter.com/CarlZha/status/1439608272940519430?t=4-p... You see a couple of interesting things. From 1949 onwards, there was a steep gain. Even the famine only managed to slow the gain rather than reversing it. That same thread also refers to the Barefoot Doctor system which was a major contributor to raising life expectancy.

Exhibit 3: Kevin Tellier, a very good China watcher (corroborated by another HN user in this submission), discusses the nature of Xi's power centralization. https://twitter.com/kevtellier/status/1441774309652025346 Rather than grabbing power for the sake of power, Xi's power centralization served a real purpose that is in the interest of the nation, namely removing entrenched interests that keep the country back. The problem of entrenched interests influencing policy is unfortunately an unsolved problem in many democracies.


Kenneth James Hammond isn't objective. He's funded by the Confucius Institute. imo his opinion is compromised by money i.e. he is a paid CCP apologist. It's no different from researchers who deny climate change and are conveniently funded by the oil industry. I don't see any change in mental gymnastics just because it's coming from a caucasian. Incentives tend to drive the outcome of anything like someone's public opinion.

Mao was a brilliant military strategist, but he was a complete disaster as a statesman. He is not vindicated. He made Zhou Enlai's work much harder to accomplish and much longer to take effect. There are few statesmen who were worse than Mao like Stalin. I have to question either your logic, or whether or not if you're commenting from inside the mainland.

> Exhibit 2: life expectancy from 1850 to 2020.

Again, how can we trust the data from yet another secretive authoritarian government?

> Exhibit 3: Kevin Tellier, a very good China watcher

I don't know much about Kevin Teller, but if he has a conflict of interest like Hammond, then his opinion is worth less than more objective sources.

For the record, I am neither anti China or even anti CCP (I don't even think it really exists anymore since it's a dictatorship again); nor am I a pro US zealot. I am only anti-Xi, and I am not a fan of Mao's governance.

If something stinks, I'm not going to pretend that it's nice perfume.


I am in fact not writing from the mainland. The insinuation that the only way I can have a different opinion, is because I am forced, is part of my critique: that western views of China are entirely unrealistic and overly warped by biases and preconceived notions that people are unwilling to let go.

If you want elaborate, nuanced comments from me, check my commenting history, such as this lengthy thread I wrote 2 weeks ago. https://news.ycombinator.com/item?id=28632820 I am sorry if I don't turn every one of my comment into an essay.

As for Deng Xiao Ping, I am in agreement with dirtyid's comment.


> The insinuation that the only way I can have a different opinion, is because I am forced, is part of my critique

China doesn't have freedom of speech. Chinese citizens can be punished for posting innocuous things such as a cartoon bear who loves honey, so it's not a stretch to assume that you will write differently when inside the mainland vs outside of the mainland.


Censorship exists, but that is not at all the same thing as forcing people to have a certain opinion.

This Harvard paper explains that, instead, censorship's role is to silence movements. Pro-government posts are censored when they have the potential to go viral. Anti-government posts that don't go viral, aren't censored. "How Censorship in China Allows Government Criticism but Silences Collective Expression" https://gking.harvard.edu/publications/how-censorship-china-...

Furthermore, censorship exists simultaneously with government responsiveness. That is, they actively monitor social media for grievances and they actively address these grievances (as in actually addressing, not arresting people for arresting). Nowadays social media is one of the major channels for government feedback. New law proposals are regularly formed based on mass complaints on social media (which may or may not also end up getting censored). This is a relatively new thing, developed in the past 15 years or so.

Punishments are not that common. You'd have to reach Julian Assange level for that to happen. Censoring without any further consequences is the norm.


> Censorship exists, but that is not at all the same thing as forcing people to have a certain opinion.

I agree, but it's a moot point. When you censor all opinions that you don't like, it's only natural that if you limit available data, people will form very specific, limited opinions based on that data.

> That is, they actively monitor social media for grievances and they actively address these grievances

I kind of agree with you. In some cases, Beijing fixes problems that local provincial governments either cannot or will not, but that is not always the case, and due to lack of overall transparency; we don't know the ratio of fixing problems vs punishing people for reporting grievances.

> Punishments are not that common. You'd have to reach Julian Assange level for that to happen. Censoring without any further consequences is the norm.

Honestly, I do not feel that people broadcasting the events in Wuhan were Julian Assange level. Neither are the people who were speaking about Xinjiang. Yet they were still punished. Tax evasion imo is also not Assange level enough for a rendition and a resource intense redaction of media credits. Of course, you can argue that it's better than not punishing tax evaders at all like how virtually no one in Wall Street went to jail for the 2008 financial crisis.

I know the social credit system is still currently a mess implementation wise, but if it ever gets standardized and fixed nationally, we can revisit this subject.


Then you aren't reading the right "China-watchers". Some names from a quick review of my Twitter feed to check out if you're actually interested: Kevin Tellier, Peter Hessler, Xifan Yang, Michael Pettis, Manya Koetse, Sabina Knight, Amy Qin, Keith Bradsher, Yuen Yuen Ang, Chaoyang Trap (podcast), Matthew Casey, Jonathan Cheng, Yiling Liu, US-China Perception Monitor, Paper Republic, China Media Project.

There are tons more that I'm leaving out because I don't feel like searching for them. Enjoy. Or just continue to dismiss all Western journalism because you'd rather do that than look for quality.


I am in fact following a couple of the people you mentioned. The critique isn't that no good China watchers exist. It's that most of them are like Gordon Chang instead of Kevin Tellier.

I don't see the need for you to resort to confrontational hyperbolic strawmans. The fact that good China watchers are a small minority, is very bad for western understanding of China, and thus very bad for rational policy making that actually addresses issues rather than stirring up emotions or manufacturing consent for a war. This isn't just my own opinion; Kishore Mahbubani, ex-Singapore diplomat, ex-UN security council head, has spoken extensively about this problem.


Sure, the quality of mainstream news is always, and has always been, a problem. More education and improvement in understanding is always a good thing. But you're not making any interesting points there, and you have no way to quantify that "most of them are like Gordon Chang". And the problem I see is that people have a tendency to find poor sources of information and use them to dismiss entire categories of information and positions.

By the way, your statement was about "Western China watchers," not some or even most. You may not have meant it that way, but that's how you said it.


Okay fine. Although I did mention MSM.


It would be a hell of a lot easier for the west to understand China if China would let their citizens interact with the wider internet.


Check out quora.com, It must be whitelist or something ... it's a real melting pot ...


That isn't my experience. There are in fact quite a lot of Chinese people that jump over the firewall. Most of the time, their opinions are quite different from the mainstream western perspective. But when they voice such opinions on western platforms, they tend to be labelled as shills. Only negative, dissenting opinions about China are accepted.

I am in favor of relaxing the firewall. Yet I hold no illusions about that yielding a better outcome when it comes to inter-cultural understanding: it think all in all it's just going to lead to more flamewars.


From their viewpoint, TSMC is theirs anyway.


So obviously some chips fabricated by TSMC go to China for further manufacturing into consumer devices or whatever. But I was also surprised by the grandparent comment to this one. Does TAMC have fabs in mainland China. I would have thought that would make the US (and Taiwanese) government nervous as Chinese manufacturing being so reliant on external fabs discourages them from doing anything silly with Taiwan.


TSMC has fabs in China (Nanjing and Shanghai), but at the much older 28nm node.

1. search for "Fab 16" and "Fab 10" on following page:

https://www.tsmc.com/english/aboutTSMC/TSMC_Fabs

2. search for "TSMC Nanjing node" and "TSMC Shanghai node" and you'll find articles mentioning 28nm


TSMC is good at this subtle diplomatic tightrope dance.


TSNC has a plant at Nanjing, China (TSMC Nanjing Company Limited, Fab 16).


There is a Tsmc plant in china for old generation chip.


> Wait, China uses TSMC? I'm surprised PRC govt allows that.

They have any other alternative? China imports more microchips than OIL in trade value.

TSMC going black is an instant lights out for their industry.


China tried to subsidize a national semiconductor manufacturer to rival TSMC but after billions wasted they gave up.


> they gave up

That's a good thing, as there are still people trying to play the catch up (e.g. SMIC), without subsidize.


So no TSMC in Europe ? Or any equivalent ?


I think an issue is tax breaks. In many European countries (AFAIK) it would be against the law to offer a company a tax break if they would choose to establish operations there.

> The total investment in the project is estimated at 800 billion yen ($7 billion), with the Japanese government expected to provide up to half the amount.

That definately wouldn't fly in many EU countries.


State aid in EU is limited to 15% of the total value of the investment, but sometimes it is circumvented by offloading certain parts of the investment onto public budgets (e.g. infrastructure, or training programs)


They mentioned wanting to open one in Europe, iirc Saxony is a possible location. Bosch has built one there recently, Global Foundries, Infineon and Xfab have fabs too.


Cross our fingers that India is able to pull this off. China got old before it got rich, hopefully India can bridge the income trap.


Pursuing high value but low employment industries like IT services for Indian upper cast is why India is so behind PRC. Semiconductor industry is more of the same, but worse. Even less employment or and revenue once famine cycle of semi demand kicks in after every bloc gets their fabs up. India needs cheap dirty manufacturing that employs 100s of millions. But automation, desire for self-sufficiency, and rise of China means countries won't risk concentrating global supply chains in single place like India again. Especially India due to geography.

>India can bridge the income trap.

India @ 1/5th of PRC economic size would have to sustain ~25% growth just to maintain with China growing at modest 5%. Indian pre-covid single digit growth can't even generate enough jobs to keep up with population growth. Reality is India's demographic divident in her most productive states are going to fizzle in 10-20 years, comparable to PRC. Overpopulated and underdeveloped states skews overall trend to 2050s, but these are provinces with level of hunger, malnourishment, stunted and wasted developement on par with North Korea. These are the human capital comparable to uneducated folks that compromise ~200M stuck in informal economy in PRC because they can't be integrated into modern economy. India is going to have to deal with demographic bomb while much poorer than PRC before there's light at end of tunnel.

>China got old before it got rich

Middle income trap is a myth, and if it wasn't PRC is upper middle income and steps away from high income of ~$12,500 anyway. Competent countries "trapped" in middle income still typically grew faster than developed economies as well. The question is really can PRC get powerful before it gets old, and all signs indicate to yes.


>The question is really can PRC get powerful before it gets old, and all signs indicate to yes.

The question is whether the CCP can weather the political transformation when transitioning past the middle income stage. That's the real preoccupation with policy makers in China right now.


> India @ 1/5th of PRC economic size would have to sustain ~25% growth just to maintain with China growing at modest 5%.

If India is 20% of China's GDP, and growing at 5%, and China is growing at 5%, India will remain 20% of China's GDP.


Base effect.

5% of 100 is 5.

5% of 500 is 25.

India needs to grow at 25% of 100 to match China growing at 5% of 500. China will eventually only grow at 1-2%. India may eventually catchup by sustaining 10%+ growth. But it's a long road and forces not in India's favor.

E: to clarify, I mean maintain same amount of growth as China, not maintain same proportion of growth. Same proportion increases actual gap.


Same absolute gap is kind of a weird metric, though, isn't it? If China is 500 and India is 100, and China grows to 50,000 and India grows to 49,600, you would call that "the same gap" but I would call it near-parity...


It's not a useful metric but illustrates how much India is actually lagging vis-a-vis PRC. The gap is widening not converging. The actual situation is even worse if you factor in things like forex trends.


Same thing. India will be old by the end of this decade, and we are not even as rich as China.


If you read the original article, Taiwan is not keen on giving India a semiconductor fab, but who cares, a narrative has been built.


Yes, clearly everyone is preparing for the inevitable - one way or the other China will finally get some control over Taiwan.

Military-wise China is almost ready to be able to take over Taiwan by force (don't get fooled by the numbers that says that US is still spending more on military - every dollar spent in China can produce three times more military equipment than dollar spent in the US).

AUKUS keeps developing its marine forces, however it is very difficult to provide supply chain for large military operation over the Ocean while your enemy can do this using trucks and trains.

I don't think China wants military solution, their leaders are very practical people, so, probably, we will face "Hong kong-isation" of Taiwan at some point.


> China is almost ready to be able to take over Taiwan by force

I keep seeing this argument. Let me explain why Taiwan will be free from China's attack for the next 30-50 years (long enough for Xi JinPing to expire)

- US protection. If there's one thing both parties in US agrees on, it's against China. Biden just emphatically committed protection for Taiwan and peace in the south east Asia. There are many US warships sailing up and down Taiwan straits. There are many military deals and cooperations with Taiwan. TSMC is a critical part of the world's electronic supply chain. Taiwan is a critical part of the island chain strategy to contain China, since it acts as an unsinkable carrier against China.

- China's military is only starting up. It doesn't even have capability to produce its own jet engines or carriers. Its strength lies in the large # of soldiers, which have to be transported. And its large # of ships, which are mostly coast guard/civilian level, and can be sank relatively easily. China's military is about 3-4 generations behind US, and currently is no match for the 10X power from US and its allies

- United democracies. US and its many allies are now coordinating naval exercises in the Region. Japan. Australia. India. UK, France, and Germany now have warships in the region as well. South Korea and Taiwan are now increasing its military defenses. These firepowers are 90% of the world's military right now.

- Taiwan as a fortified island. Fortified islands are hard to take, as evident in the modern war histories. There are only a few times in the year that China can safely cross the straits without fearing typhoons or rough sea. And landing only on a few spots. Especially if the island is armed with the latest war technologies from US, and has the missile capability to strike Beijing, or the three gorges dam which will then wipe out millions of Chinese. Taiwan is now producing domestic submarines, long range missiles, and ship missiles. You'll want to remember that Taiwan has very sophisticated electronic capabilities, so the arms it is starting to produce will be very good.

- China only gets one shot at attacking. After which, it will get economically sanctioned into oblivion by the world economies. Because otherwise Japan, South Korea, and other island nations near China can be attacked with similar attempts as well. Since China mainly imports most of its resources, and is export focused, it will sank into a deep economic depression filled with local riots.

以前大陆民众被灌输的是“血浓于水”的观念,其实并没有了解李登辉以来台湾的真实变化,等到两岸口岸开放,大家可以自由往来的时候,现实的观感和预期落差很大,自然产生了矛盾。从台湾人的观点来看,大陆的提升对他们产生了威胁,态度发生改变,加上过去三十年来的灌输,对中国大陆的认同感低落和仇视也是什么出乎意料的事。台湾俨然成为美国的战略桥头堡,整个局势到如今无可挽回。


> United democracies.

Between Hong Kong and the Ukraine those democracies have shown little willingness for action. TSMC may be Taiwans trick to change that but with all these Fabs outside Taiwan who knows if that will remain the case.

> There are only a few times in the year that China can safely cross the straits without fearing typhoons or rough sea.

Thyphoons don't come out of nowhere and there are usually at most a couple per year.

> And landing only on a few spots.

Could you expand on that? Almost the entire west coast is flat with normal beaches that don't seem particular incessible.

I really hope that Taiwan will remain independent, but I don't think relying on Taiwan's defenses or western military support will guarantee that. The economic argument is a better one - there is a lot of chinese investment around the world that they probably don't want to risk and the world is (slowly) starting to realize that it might be a good idea to have manufacturing capacity outside of china.


Large areas of the west coast facing China are reclaimed land filled with very narrow roads and aquaculture ponds. Think of the bocage country in Normandy and then fill all the central fields with water. There are really only four coastal areas with the roads needed to move an army off the beach at speed.


I have been in Taiwan, concretely in Taipei.

It is one of the most lovely places in the world. It is developed, it is safe. If China took control of it would be terrible news to me.

In fact, Taiwan stands as one of my first destinations to settle in the next years.


I can see and support a war against China if they try that. If not then, when?


So, on reading this I imagine that this is purely for "mature" process nodes, and not definitely for cutting-edge nodes, plus the fact that some of the semiconductor infrastructure is already there, so 7 billion US dollars might be in the correct ballpark.

Well, at least Japanese manufacturers (probably more of automobile suppliers like Denso rather than Sony's mobile division) will have a domestic source.


Wikipedia lists about a hundred fabs in Japan already.

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

This might not achieve the very latest cutting edge nodes, but there are diminishing returns for going very far down-node.


The article says, "Japanese chipmakers had dropped out of the race for large-scale chip development by the 2010s and instead contracted out the production of cutting-edge semiconductors to companies like TSMC." and the new plant "would be TSMC's first chip production operation in Japan".


Sure, but the basic infrastructure needed to support fabs is still there.


Yes, I knew that, but the chip crunch also affected mature nodes, so the capacity for mature nodes is also clearly not enough.


By mature, they probably mean 7nm or even 5nm. Semiconductor manufacturers are always keeping old nodes spun up for technology that doesn’t require (or wouldn’t even function properly) on smaller nodes. Spinning up a 7nm/5nm fab allows them to free up their flagship fab for 3nm.


There are levels of "mature." Many chip plants, some of which were built to nurture home grown tech, have fallen far short of what is old hat to TSMC.


With the rise of Tata, India, the US and Japan all have substantial automotive concerns that might be very interested in their suppliers having shorter supply chains. Especially after the black eye they've already gotten. Expect automotive lobbyists to be smoothing the way for these agreements to get approval from the respective State Departments.


It's very likely at this point that the next 'mature' process from TSMC will still be interesting from a competitive landscape sense. If the timeline works out so that the current node gets built in a few years in Japan or India, that won't be 'cutting edge' but it'll still be pretty sharp, if some other fabs don't get their shit together and soon.


Somebody needs to build cutting-edge node fabs given the rhetoric from China in the past few days regarding forced "reunification" with China.


Taking over Taiwan will not give China cutting edge node fabs. TSMC is working on a western equipment plus chemical stack. So it will give China absolutely nothing, China is rich enough to make Taiwanese talent offers they can't resist. So much so that Taiwan had to shut down Chinese recruitment bureaus and certain Taiwanese talent are not allowed travel or exit the country.

China can just bide its time they will be the biggest economy at the end of the decade that will psychologically change the world so much regarding the West and Asia. You can already see a disconnect between western and Asian elites/diplomats view on China and the Region. Why Western ASEAN anti-china recruitment the last few years failed really hard.


> China can just bide its time they will be the biggest economy at the end of the decade that will psychologically change the world so much regarding the West and Asia.

Does this take into account China's demographics? Will be to able to 'grow rich before it grows old'?


I'm not a demographer, so only time will tell how big of a problem it will become. The Chinese government is worried about it, already was couple of years ago. But I don't see it becoming a roadblock that will prevent China from becoming the biggest economy in the world, hell I wouldn't be surprised that the Chinese market will be twice that of the combined western market.

China can always play the Asian immigration card. To soften the blow of a graying demographic.


China has been pushed towards being a monoculture for the past 70 years or so. I don't think Asian immigration would be an easy option at all.


For certain definitions of "monoculture". Even Han culture between different regions can be very different.


Sure, China is not as homogeneous as what it could be. Yet is it preserving diversity? Through its enforcement of standards and nationalism, the CCP has been clear that it has no interest in making China more multicultural.


That isn't the reality I see on the ground. Take a look at for example this Tibet documentary: https://youtu.be/zkYhSsQqRhI This Tibetan woman has 3 husbands. Look inside her house: full of Tibetan and religious artifacts. And this is just a normal Tibetan household, after we've been told for decades that Tibetans are being culturally genocided.

It looks to me that Tibetan scripts and languages, Mongolian scripts and languages, etc are better preserved than over the border in India and Mongolia.

Yes there are also plenty of examples where they tear down a cultural site. Yet I think there is a fundamental difference between western and Chinese interpretations of such acts. Western societies have developed a sort of mythical worshipping of traditional ethnic culture. To the Chinese, to whom development and modernization have come late, development and modernization is number 1: there is no pride in living a traditional, ethnic life if that means being in the middle of poverty. Some sites are torn in the name of development. See this discussion from 1999, back when the China topic wasn't so politicized: https://www.theatlantic.com/magazine/archive/1999/02/tibet-t...


> It looks to me that Tibetan scripts and languages, Mongolian scripts and languages, etc are better preserved than over the border in India and Mongolia.

You've been drinking the koolaid. Tibetan as only been preserved through a heroic effort to smuggle books over the border to Nepal and India for preservation and duplication.


Seeing Tibetan and Mongolian scripts in videos of Tibetan and Mongolian locations is "drinking koolaid" now? Are you saying those videos are completely doctored and the signs put up only for the purpose of making those videos? Are you saying they don't actually teach Tibetan and Mongolian language in schools in China? Are you saying that woman in the video is an actor? What exact facts are you disputing?

In the mean time, a large part of the Tibetan separatist diaspora in India have lost the ability to read, write and speak Tibetan; and outer Mongolia has only recently begun their efforts to restore traditional Mongolian script, after decades of using Cyrillic.


Sure, but immigrants would be non-Han.



Not just Taiwanese talent -- China has also been recruiting South Korean semiconductor talent.


At the current rate, China won’t become the biggest economy in nominal GDP for a few more decades, and that assumes growth continue at the current rate

China might become number one in GDP (PPP) in a decade, but that also assumes continued growth at the current rate. China is overdue for a recession, as they need massive restructuring of the economy to shift away from an export driven economy and to pop its housing bubble (which does drive GDP growth now but not in a way that actually makes the country wealthier)

Why do you think Xi Jinping is doing such massive reforms right now? And clamping down harder on dissent and access to information and culture that’s not in line with CCPs agenda? He knows what’s coming, and to an extent he’s accelerating it with new lending restrictions and such. He knows it’s necessary.


China already has the largest GDP (PPP) in the world: https://en.wikipedia.org/wiki/List_of_countries_by_GDP_(PPP)


It was kind of a shocker too me when I discovered that in the PPP metric China has surpassed the US back in 2014. Also a thing i heard today was that Mercedes for examples sold 770k plus cars in China compared to only 220k and 260k in the US and Germany. And people still expect western companies to decouple.


> So it will give China absolutely nothing

It essentially gives them tremendous power over the world. All they have to do is threaten to shut it down, or ban exports to the west, or put very high tariffs on them. In addition to gaining all IP from customers. What do you think that will do to all companies that rely on cutting edge TSMC nodes such as Apple? Tesla? Nvidia? These new plants won't likely be putting out M1 chips, etc., the cutting edge fabrication is held physically in Taiwan. It would be a great way to collapse the economies of the world as you are rising, if one were aiming at doing such a thing.


And why would they do that? They themselves also need advanced chips, which they can't get when the west boycotts supplies. In the end, such actions are doubled-edged swords, nobody benefits.

I find all this decoupling rhetoric/fear very weird and contradictory. On the one hand people say we should decouple from China, on the other hand we fear decoupling from China and we get furious when they initiate decoupling (see Australian trade). In the mean time actual Chinese officials keep saying that a trade war benefits nobody and keep calling for rational decision making.


Trade war benefits nobody, but if you can gain slight advantages by manipulating currency, stealing IP, etc, you can get ahead while 'saying the right thing.'


China depends on global commerce as much as the US or Europe. It is not on their best interests to create a global tech crash.


China must obviously gain cutting edge node fabs or associated knowledge and processes from seizing Taiwan given the claims you've made about Taiwanese being prevented from being hired by the Chinese Government.


China is already domestically on 14nm pushing for 7nm at SMIC. They went from a decade behind to like 3~5 years behind in like 18 months.

Time is on China's hand, Taiwan and south Korea should be worried about their so called ally the US stealing their tech. Because without TSMC what is Taiwan for the west a annoying island that we need to backstab some time in the future like the western backed regime in Afghanistan. At the end of the day its just business.


Past few days? Where have you been for the last 70 odd years? Difference lately is it's become more credible for bunch of reasons around politics and capability.


What the world has in shortage is those mature process.


Let me know when they actually break ground and start building one :-).

I fully support TSMC's efforts to diversify the jurisdictions that have semi-conductor fabs though. That is a huge win for them and makes the Chinese mainland threat to assimilate Taiwan less? More fab capacity is a win.


>That is a huge win for them and makes the Chinese mainland threat to assimilate Taiwan less? More fab capacity is a win.

No the opposite, TW is giving up their silicon shield due to pressure from various blocs. The chief global interest in defending TW evaporates and PRC calculus for reunification increases when enough critical semi productions moves off island. DPP / TW has been balancing a tight rope of placating demands for TSMC fabs outside of east asia and trying to trade it for some geopolitical wins. But really TSMC and TW has no choice, EU/US/JP all have sticks like tech sanctions of supply chains TSMC relies on, the carrots of subsidies and temporary positive geopolitical messaging around TW recently are consolation prizes.

I think TW hopes naively if they spread TSMC supply chain in enough countries, it would increase mutual dependancy and security. Instead, in scenario of PRC reunifying forcefully, host countries will just nationalize TSMC fabs on their soil. The entire point of TW's silicon shield is to concentrate fabs on TW to sustain uneasy detente. The more production moves off island the more opportunities for global demand to factor out TW fabs... which unlike Japan, Korea, NATO countries where new fabs are being proposed, explicitly does not have US security guarantee, i.e. TW fabs will become the most vunerable suppliers. This point is key. For PRC, having critical fabs concentrated on TW is a geopolitical gift, even if US can leverage them in peace time via sanctions, PRC can leverage them more during war time via disruption. Everyone knows this, no one will want to depend on TW fabs once suppliers on secure soil are established. TSMC / TW writing their suicide letter. I anticipate TSMC will do everything it can to delay these projects to buy more time.


I wonder what the situation with production capacity is. I suspect TSMC will build a fab here and there that might suffice to make enough to supply the military, but not for anything else. Large sectors of the economy would still be cut off from chip supply, and we've seen how that went with the car industry.

If PRC would conquer TW, they would not have an incentive to stop chip exports I think. Too much mutual dependency with the trade partners, and the party does not like domestic instability - unemployment would be unwelcome. Would TW blow up the factories though? It is said they can do this with a press of a button.

> EU/US/JP all have sticks like tech sanctions of supply chains TSMC relies on

Wouldn't that be self-sabotage? TSMC is in the strong position, everyone wants capacity, they have it. Intel and Samsung can't compete on the latest nodes.


I still think this is good for Taiwan’s security. The world is very aware of Chinas growing ambitions, and everyone views Taiwan favorably. Taiwan needs to focus on being viewed as the better partner to liberal democratic countries. I think USA will keep putting troops near or at Taiwan in the coming years. Very slowly as to not create a huge provocation, but still steady.

Taiwan doesn’t need the chip card to be considered worth supporting because a war between China and Taiwan will devastate the world economy in a hundred different ways. And regardless, it’ll take many decades before Taiwan isn’t such a large share of world chip production that they’d actually lose that card. Losing even 10% of the worlds cutting edge chip production would be utterly devastating.

If China attacks Taiwan, and American troops are harmed, the US will have the justification they need to join the war, and the public support theyd need. I don’t think the US wants to go to war. But they need a threat that’s as credible as possible to deter China.


> Losing even 10% of the worlds cutting edge chip production would be utterly devastating.

The alternative is losing ~80% of all global chip production on TW, JP, SKR soil if US decides to fight PRC over TW. That's the leverage PRC has in war scenario. Losing leading node is survivable, losing east Asian fab sets globe back decades, and hurts US/western tech industry the most.

A fundemental shift in region is US can't deter PRC over TW anymore. And as much as TW is viewed favourably, no one has committed to TW defense and US tried recruiting hard. Even US itself won't give TW security guarantee, and if you follow policy space, strategic thinking has shifted away from TW defense to TW armorer and blockading PRC past 1st island chain - increasing consensus and acknowledgement that TW is simply not defensible. Ultimately, no one wants to goto war and US can't deter PRC. So we have US and other blocs trying to maintain status quo for 10/15/20 years while shifting risk off TW sufficiently to mitigate esclation towards regional semi mutually assured destruction, a crisis much larger than just losing TSMC. Is this good for TW security? IMO it's not.


So what actually happens to fabs outside of Taiwan if China do take Taiwan and presumably also TSMC the company.


Really good question. I don't know.

What I would do if I were leading TSMC is create a wholly owned subsidiary which held all of the non-Chinese assets and headquarter it in a tax neutral country like the Netherlands. Let's call it TSMC International. Then in the event of an assimilation by China, sell all of the assets that TSMC the parent company owns, to TSMC-i for $1 or something. And in fact have the paper work at the TSMC-i headquarters all filled out just waiting for being dated and executed by TSMC-i executives.

It is a variation on a poison pill defense.


I think the going in position is that the fabs in Taiwan are not going to exist if China takes the island, so the only fab business will be outside fabs. I assume they have a continuity plan for one of the other offices to become the new HQ https://www.tsmc.com/english/aboutTSMC/business_contacts


Something like what happened with Arm China.

https://www.extremetech.com/computing/326447-arm-china-seize...


p.sm this was retracted. The new company is specifically specialising in extensions and add-ons. Or that was what they were told to re-write.


>Or that was what they were told to re-write.

Especially when they are in the middle of going thought M&A. Along with trying not to anger China which could retaliate.


Good question, but we'll probably have bigger problems, like a potential hot war, if that happens. How these problems get resolved ie "who wins" is the answer to what happens to TSMC.


Theoretically; something similar to what happened to the Cuban cigar producers after the revolution. They spin off into their own company/companies no longer recognizing the “invalidated” company charter.

But, realistically, China will press their global influence to keep them under Sino-TSMC’s control under a new Chinese corporate charter.


It will depend heavily on the amount of global sway China has at that point. I imagine the countries TSMC builds the fabs in would vastly prefer to use the invasion as an excuse to say the fabs belong under the control of a entity without connections to China.


Well one possibility is that those fabs will simply be nationalized by the country that they are in.

That would be an effective economic retaliation against a country that invades someone else like that.


They'll be spun off into different companies. That's usually what happens when multinationals get cut across by war.


The headline and the first line of the article are not all that much in sync. Joining or considering to join:

Taiwan Semiconductor Manufacturing Co., the world's largest contract chipmaker, and Sony Group are considering joint construction of a semiconductor factory in western Japan amid a global chip shortage, Nikkei has learned.


Thanks for pointing it out, it’s just ridiculous how many posters only pretended to have read the article.

I don’t usually agree with heavy handed moderation but in this case it would be nice if the title was changed to reflect the first line of the article… and if the submitter was given a warning.

The mass political derailing will probably block the actual info from rising to the top but oh well at least we can try.


So when Xi invades Taiwan and forces unification, who owns these fabs? Will there be a new tsmc-jp and tsmc-us?


Possession is nine tenths of the law.

(I almost worry about Taiwan; if they are the sole source of iPhone and PC chips, then they will receive a lot of protection from other countries in the event that China invades. If we have TSMC factories in Japan and the US, then people won't care. Good for buying GPUs, bad for democracy.)


I also worry for Taiwan. If China chose to invade and the first invasion wave were not successfully repelled by the military assets already in the immediate area, other nations might try to intervene but few would be in a position to act immediately. "Defending Taiwan" is a tricky prospect if the PRC already invaded it yesterday; and as you say, possession is nine tenths of the law.


I'm not sure what would be left in Taiwan after an invasion but I'm sure it doesn't include the hyper-sensitive latest-node semiconductor factory.


Rumor has it the TSMC fabs are rigged with explosives in case of invasion. The beaches are also lined with pipes to spray fuel and prevent landing troops.


I thought the way the island is set up is that is that it was very difficult to make an initial amphibious push on the island with a few chokepoints that made defending it much easier for Taiwan that one might expect. And if China DID manage to break through one of those points the Island was effectively lost.


The decision whether the West will fight for Taiwan will not be affected by TSMC.

HN crowd absurdly overestimates the need for the smallest node. At worst, the West will have a mildly smaller amount of slower chips for a few years until sufficient investment allows catching up.


TSMC is a better defense than owning nuclear weapons. But also way more expensive to develop (so not really a good option for North Korea or Iran...). But for Taiwan it is vital to keep the smallest proccess on their land.


If war did not end up quickly, all of the trade in the Western Pacific might be blocked. Japan as an upstream manufacturer of semiconductor industry might be affected by this, and needs to seek for a new channel for shipping (maybe Northeast Passage). TSMC-whatever would have raw material shortage.


Wasn't there a Venezuelan oil company that had its American operation "secede"? I imagine a similar thing will happen then, what's HQ going to do about a rogue office that has the support of the host country?


From the legal perspective, the shareholders still own TSMC (by definition) and therefore still own the fabs, same as before, and the shareholders are spread out across the entire globe.


But non-Chinese can’t own a Chinese company. All the Chinese companies on American stock markets are basically just shell companies with a “profit sharing agreement” with the underlying Chinese controlled asset.

That’s why this is an interesting question to me.

It doesn’t feel like we can trust the CCP to maintain the status quo any longer.


It's not quite that simple, A-shares are only for mainland investors but H-shares (listed on HK exchange) are available to trade by anyone. US based ADR's have received a lot of attention recently but they're not the only channel for foreign investment in China.

BlackRock is also running a mutual fund of mainland equities.


Under which legal system? I thought that China was generally able to influence who owns domestic companies pretty much at a whim by nationalizing them, and that chip manufacturing seems like something that many countries might consider nationalizing based on how expensive they are and how critical they are.

Although it seems convoluted enough that if it's not truly critical things get left alone. But chip fabs are probably critical enough that they won't get left alone.


Article in the original language:

[0]: https://www.nikkei.com/article/DGXZQOUC10E680Q1A910C2000000/


Not an insider : isn't $7bn relatively cheap? Could it be this plant is for image sensors?

I couldn't read the article.


The article mentions:

> Sony will also help prepare the factory site. Its aim is the stable procurement of semiconductors for its image sensors.

> The company controls half of the world's market share for sensors used in smartphones and cameras, with manufacturing bases in Kumamoto and Nagasaki prefectures. The sensors are manufactured in-house, but the semiconductors that process images are procured from third parties, including TSMC.

> Sony CEO Kenichiro Yoshida previously said that the ability to steadily procure semiconductors is important for maintaining Japan's international competitiveness.


I was interested to see it was in Western Japan, and after looking it up in context, it isn't surprising when almost all of Sony's foundries are all over Kyushu.

Always interested to understand why things are where they are (e.g. proximity to Korea/China, sheer expense of Tokyo or somewhere like Osaka for a large industrial operation).


Once you get out of the major metro areas the rural towns which are alive all have various major factories.

Tokyo and Osaka houses headquarters, manufacturing is almost always in some small town. A big benefit is Japan has a flat consistent tax system. Towns and prefectures cannot compete with tax policy or tax cuts. Thus factories are much more diffuse. The large corporations can bring their support networks with them but when recruiting high school graduates a big plus is to be local. Thus a town might have a couple large factories which can consume the high school graduate supply.

Occasionally you'll see various R&D centers. Those tend to be within an hour or two from the major metro. Far enough away the companies can build large campuses and employees can buy large houses for their family.

For example HAL Labs, the company which makes Kirby games, has an office in Yamanashi. Which made them one of the rare non publisher video game devs who do not rent.


No idea for your question, but you should be able to read the archived article here https://archive.is/TM8fk


So, on reading this I imagine that this is purely for "mature" process nodes, and not definitely for cutting-edge nodes, plus the fact that some of the semiconductor infrastructure is already there (for use in image sensors), so 7 billion US dollars might be in the correct ballpark.


Sounds cheap but they made a couple of billion on the ps4 alone. Sounds like a reasonable investment.


I thought semiconductor factories couldn't work in places with lots of seismic activity.


Taiwan have loads of earthquakes too, people just work harder.


Intel has multiple fabs in Oregon. They have seismic requirements for buildings and tools.


I'm a little bit confused why it has to be so expensive and so hard to produce chips... Is it because of the scale required to sell them at a low enough price to be competitive? If so, is there maybe a market for smaller fabs that only produce small scale batches of specialized chips?


A few reasons:

1. Extremely expensive equipment, both directly related to IC fabrication (e.g., lithography) and indirectly related (e.g., air filtration, water processing, stabilization).

2. A large amount of skilled labor onsite to operate and maintain the complex equipment.

3. Constant supply of resources, including power, water, and sand.


The machines that can make modern chips are extremely expensive. On one hand they are sensitive enough that shipping them costs millions of dollars since they need to stay in a vacuum. On another hand they need a room-sized laser installed nearby.


And water. They need lots and lots of water. And everything has to be ultra clean and to start that way in creation, shipping and in use.


While I don’t disagree, saying that expensive equipment is expensive doesn’t really explain why it’s expensive in the first place.

The next two statements are more helpful.


>I'm a little bit confused why it has to be so expensive and so hard to produce chips.

You are talking about something that is forever edging towards atomic size manufacturing. $7B is hardly expensive in today's terms. Not to mention the scale. Annual Smartphone sales is 1.2B unit. You have multiple silicons within a single smartphone.

>If so, is there maybe a market for smaller fabs that only produce small scale batches of specialized chips?

Smaller Fabs will just be as expensive. You built larger fabs ( as you have noted ) to try and amortised over larger volume. ( Lower Unit Cost ) Which is the problem why NAND and DRAM capacity planning are hard, their Fab size have grown to a scale that it is hard to build additional one without some careful consideration. ( High Cost, High Risk )


foundry is a capital intensives business.

the processing steps that is involved is like 300+. that's why you hear the term "yield" in the semis foundry business. each step that went wrong is going to affect your yield.

here is a short video that show the semis fab process.

https://www.youtube.com/watch?v=c9arR8T0Qts


The more times you have to fiddle with something, the more opportunities you have to break it. It's compound interest essentially, and that can add up real fast. If you have 300 steps and by some freak coincidence, each step had exactly the same error rate, so that you lose N% at each step, then here are the overall yields:

1% loss: 4.9% yield

0.5% loss: 22.2% yield

0.1% loss: 74% yield

0.01% loss: 97% yield

You have to have an error rate in the range of X failures per thousand per step to get above 50% yield, and X/10000 for a really good yield. And all of these things are so small that a spec of dust causes failures.


It takes on the order of a thousand steps to build them, each of those steps is supported by million dollar equipment, hence the cost.


A new fab would be competing with old fabs that are already amortized/depreciated, and thus have lower capital costs.


"Manufacturing consent" has to be the seminal term that defines the age we live in. The dust hasn't even settled on the Afghanistan disaster, and already the propaganda machine is spinning at full speed trying to build the next enemy - China.


Sony?? This is very promising.




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

Search: