Semiconductors are used in weapons, aircraft and space crafts. But they are of a different grade than those used in consumer devices. Those chips are hardened against a variety of stuff such as radiation, thermal, etc. But those chips are usually built with very matured processes such as 150nm, 250nm and are designed to run on lower frequencies unlike high-end smartphone chips, which requires the most leading edge fab processes.
Chinese semiconductor industry
- Thread starter Hendrik_2000
- Start date
- Status
Semiconductors are used in weapons, aircraft and space crafts. But they are of a different grade than those used in consumer devices. Those chips are hardened against a variety of stuff such as radiation, thermal, etc. But those chips are usually built with very matured processes such as 150nm, 250nm and are designed to run on lower frequencies unlike high-end smartphone chips, which requires the most leading edge fab processes.
Strong arm AMD to fab in the Arizona, along with possibility of miTSMC Arizona is going to be DOA. I pick 28 nm as the benchmark because it is basically the cheapest 'advanced' semiconductor node and the last planar node before FinFET. 28 nm is good enough for really good microcontrollers, automotive chips, signal processing, etc.
Semiconductors are not like cars. Cars are a discrete product while semiconductors are a process product. That is to say, it is possible to make a single car, and charge accordingly, or make 1 million cars, and charge accordingly. There's flexibility of economy of scale. Semiconductors aren't that. It's like oil. Can you make a single liter of fuel oil? No. You have a refinery and either make 1 million liters or 0 liters, the economy of scale does not support anything in between. The profit margin on a single semiconductor chip is tiny, so if your costs are increasing 6x, you better have a market that can justify it.
Any market in the entire western hemisphere that can justify multiple 5 nm fabs?
AMD processors or military chips maybe?TSMC Arizona is going to be DOA. I pick 28 nm as the benchmark because it is basically the cheapest 'advanced' semiconductor node and the last planar node before FinFET. 28 nm is good enough for really good microcontrollers, automotive chips, signal processing, etc.
Semiconductors are not like cars. Cars are a discrete product while semiconductors are a process product. That is to say, it is possible to make a single car, and charge accordingly, or make 1 million cars, and charge accordingly. There's flexibility of economy of scale. Semiconductors aren't that. It's like oil. Can you make a single liter of fuel oil? No. You have a refinery and either make 1 million liters or 0 liters, the economy of scale does not support anything in between. The profit margin on a single semiconductor chip is tiny, so if your costs are increasing 6x, you better have a market that can justify it.
Any market in the entire western hemisphere that can justify multiple 5 nm fabs?
Hendrik_2000
Lieutenant General
Good article excerpt as I said many times the tech embargo is blessing in disguise China should know that from past experience with Soviet Union . It accelerate domestic defense industry Semi is no difference. the biggest hurdle is not technical but acceptance of Chinese private company toward domestic semi. So this article is right on. But I totally disagree with their assertion that China aerospace industry like comac is a failure. No it is not a failure but starting from very low base she has no choice but using component from the west.
Enhanced U.S. export-control measures have made that decision for them and united China’s government and its leading firms in a shared goal: to pursue technological and industrial self-sufficiency so that no Chinese firm is at the mercy of U.S. trade policies. By imposing restrictions on American products, the U.S. government has inadvertently done more than any party directive to incentivize private investment in China’s domestic technology ecosystem.
Washington is right to target Chinese firms that are obvious or complicit in human rights abuses. But the sweeping nature of the Trump administration’s sanctions did not suggest a careful selection process. Rather, they gave the impression that the United States would punish any Chinese company that achieved success. Chinese firms are no longer sure if they can depend on U.S. products—or if they will be added to another opaque government blacklist and face potential collapse.
THIS TIME IS DIFFERENT
China’s true Sputnik moment has been its realization that it cannot count on the United States to supply its technology—and that it must cultivate domestic alternatives. Washington bristled at Beijing’s ambitions for semiconductor self-sufficiency and then proceeded to punish Chinese companies naive enough to depend on American technologies. U.S. companies are now uncomfortable questions on whether they can be counted on to be reliable suppliers. For all the complaints about Xi’s efforts to drive “offensive decoupling,” it is the United States, not China, that is forcing Chinese firms to abandon American products—and now these companies are pursuing domestic self-sufficiency with a vengeance.The combined efforts of China’s state drive and its innovative industry will accelerate the country’s technological advancement. In the 1960s, integrated circuits were developed when the National Aeronautics and Space Administration was willing to pay any price for technology that could send astronauts to the moon and bring them safely back. Today, the U.S. government is putting Huawei in NASA’s position: a cash-rich organization willing to pay for critical components on the basis of performance rather than cost. Smaller Chinese companies that previously never stood a chance of selling to Huawei are now sought after as vendors, and they receive infusions of cash and technical expertise that will accelerate their growth.
Private and state-owned chip manufacturers are ramping up their operations. Once siloed industries now collaborate in the service of tech innovation: the Chinese Academy of Sciences, for example, has begun coordinating regular sessions that bring together math professors and private companies. China is now undertaking a whole-of-society effort to improve domestic technology, specifically around what Chinese leaders think will drive not only economic growth but also geopolitical power.
Is all of this enough to make Chinese industrial policy work this time around? It is likely that in a decade, China will have made greater technological advancements under the U.S. export-control regime than it would have had the United States not forced China’s leading companies to buy from weak domestic firms.
Had the United States implemented but measured reforms—strengthening the Committee on Foreign Investment in the United States and prosecuting intellectual property theft—and stopped there, Made in China 2025 would have likely played out in the usual way, with inefficient state-owned enterprises and government ministries taking the lead rather than innovative tech firms.
But this time is different. True, China has big technological hurdles to overcome, including weak basic research, ambiguous intellectual property protections, and excessive bureaucratic meddling. Yet the United States cannot assume that China’s leading firms will stay down forever: companies are rushing to fill the demand that U.S. firms can no longer supply. Chinese firms have to reinvent only certain wheels, with many simply working to recreate technologies that already exist. And no U.S. restriction can change the fact that China is an enormous market loaded with entrepreneurial talent and technical expertise.
The ripple effects of Chinese technological success will be felt beyond China. For one thing, they will shape American politics. A Beijing less dependent on U.S. products will feel less apprehensive about retaliating against American firms, giving it license to respond to perceived affronts. For another thing, technological dominance will shift the Chinese leadership’s calculations on Taiwan. Beijing knows that any armed invasion of the island would prompt U.S. sanctions that could inflict great pain on the Chinese economy. Greater self-reliance would deflate the threat of those sanctions and remove a deterrent against military action.
The economic consequences of Chinese technological dominance on the United States would be no less significant. For the most part, U.S. technology firms have stayed a few steps ahead of their Chinese competition. But they might fall back as their sales dip and as Beijing launches a more powerful drive to replace them. If China comes to dominate semiconductor production in the way it has dominated solar panels, then the United States will have lost its last crown jewel in manufacturing as the products become commoditized and profits disappear.
At this point, no effort on behalf of the U.S. government can deter China’s state from its end goal of industrial self-sufficiency. But Washington can still change the calculations of private Chinese tech companies. Many of these businesses would rather not have to reinvent their tools and find new suppliers and would likely stick with U.S. technologies if given the chance. The United States should therefore roll back its most punitive restrictions on the Chinese technology sector, lest it force some of the most innovative companies in the world to work within their domestic tech ecosystem. At stake is the future global center of technological innovation: Washington should know better than to fuel its greatest competitor.
@escobar Sir a mea culpa moment, but the author seems convinced the righteousness of their action until the Chinese had found a way out of their predicament...lol. As always of US think tank publication , they think the Chinese can't lived without them. Now the prospect of an independent Chinese IC really scared them hahahaha. I think the optimism in China IC and the recent rebuff on American overtures really bring home the futility of using sanction as a weapon. Now let see if they will implement a total ban, with inflation raging in the world with the approval of the $1 TRILLION infrastructure spending, China will have 5 years to navigate this problem and develop itself until the US recover. Within that time frame all of our aspiration will be achieved. 2025 will be an interesting year and I'm looking forward to it.
I have never laughed this much.. The Anglos know what's coming but they are too arrogant to admit their mistakes and also because of their fractured politics.
A China not dependent on American tech is a China they don't wanna face.
Funny thing is that they have got everything to lose.
@Topazchen bro that think tank is so gung ho in supporting Trump in sanctioning China now they blame him...lol. You know they force Trump to act against his instinct, Trump is against full sanction on Huawei until force to, I remember a staffer inside the commerce department was reprimanded and force to resign for recommending it...lol
you use redundant and hardened chips. Most of those are on older processes like 180 nm but use exotic architecture, materials and processes.
Every fab needs volume. The closer to 100% the better. Any unused capacity means every single chip from the fab is more expensive.
If they price in the increase then all their products are more expensive and other customers may go to Samsung or SMIC. If they specifically increase the price for AMD the AMD product line loses market share. Although the increase may be taken up by Intel or Apple, it could also be that non business customers just use an ARM tablet instead.
Since there's no local electronics market they need to export anyways, so China is going to be a buyer. But that also means China now holds those fabs in it's hands: retaliatory tariffs would kill them since they can't sell to anyone else at volume but Chinese fabs can sell into their local market.
That means subsidies.
They will need to subsidize these zombie fabs forever, or they'll simply collapse.
It is rare to see such a well balanced article in western media. Anyhow, even if Biden rolls back every sanction and blockade I do not believe China companies will trust relying on American technology anymore. It is independence or death. Nothing less.
That's newly grad white collar work for you. If you work in state owned corporations it might be slightly better because you get additional perks and vacations.
- Status