Whoever controls computer chips will control the world. As we saw during COVID, its almost impossible to build anything without chips. And we can all make guesses where AI is going. The future will require advanced chips and lots of them.
Advanced chips are incredibly difficult and complex to make and are impossible to manufacture without an input of American technology at some point in the chain. Last Oct. 7 the Biden Administration announced export controls of American chip technology to China. Notably, Japan, the Netherlands, and Taiwan have signed on as well. Without this technology, it will be extremely difficult if not impossible for China to develop its own advanced chip industry.
But there are risks as well. One reason China has struggled to develop its own chip industry is because manufacturers had access to superior Western chips. But if Chinese manufacturers banded together, they possibly could create the capability to build their own advanced chips, which is the nightmare scenario everyone wants to avoid. And China is very motivated to do so.
Though delivered in the unassuming form of updated export rules, the Oct. 7 controls essentially seek to eradicate, root and branch, China’s entire ecosystem of advanced technology…C.J. Muse, a senior semiconductor analyst at Evercore ISI, put it this way: “If you’d told me about these rules five years ago, I would’ve told you that’s an act of war — we’d have to be at war.”
Sure, technically anything is probably possible. But look at who makes all the advanced chip making equipment. Primarily ASML (Netherlands based but lots of US sites) and several US companies.
Who makes the software that is used to design the most advanced chips? I think they are all US companies. This software has to be updated to use the latest equipment and the secret sauce that each FAB uses for each smaller process node.
So they would need the advanced software (CAE/EDA), they would need the advanced equipment (i.e. EUV litho equipment and all its supporting tools), they would need an advanced FAB with an advanced process and they would need some advanced designs.
Not doable in 5 years. Even Intel, who has stumbled hasn’t been able to catch up in 5 years. Possible in 10+ years if they have EUV work underway
Absolutely. The article emphasizes how extremely difficult this is. That’s why there are so few players, why the technology embargo is so damaging, and why getting players like the Dutch and Taiwan involved was so important. Without advanced chips China’s goose is cooked, so they have little choice but to try. Succeeding is a different story.
A teacher of mine, decades ago, commented on the Soviets/Chinese, and others gaining nuclear weapons. He said words to the effect physics works the same everywhere. Once one party has proven it possible, others will, eventually, work out how to do the same thing.
There is the complexity of the atom but we do not need the extremes of that to split the atom. Meaning we did not know much about subatomic particles to split the atom. Physicists might have but the actual bomb designed by physicists was an engineering problem.
There is the complexity of DNA. That is more complex than a semiconductor. Both though are more like spaghetti code than the complexity of the atom.
There are the basic chips and then more and more complex chips along with the firmware and software to run them. Without that as well you have nothing at all. With all that complexity starting mid stream but from scratch is even more difficult.
The real problem though where are the Chinese design companies? By the time the Chinese get neck deep into creating the design companies all of it is a nightmare. Human beings want a gravy train.
There is an undergrowth of technical people who can not perform well. As China is doing rushing into design on all levels that undergrowth is too thick.
The Chinese five years ago were estimated to have 1000 extreme experts in AI. Yet the country throughout had hired hundreds of thousands of people to work on AI.
The AI budgets in Shanghai alone were something like $16 Billion I vaguely remember about five years ago. The country went into a massive building spree for AI. Since then the US and Chinese budgets may have come into parity. The main chatbots so far are US. But I admit I do not know if a Chinese company has a chatbot. Attaining a Chatbot is not the highest hanging fruit.
Okay so I looked the early powerful Chinese chatbot was Ernie. The thick underbrush is horrifying. The chatbot has to act as a censor of all sorts of things it has no clue it could say. Entire bureaucracies will have oversight.
Just another note on China’s “progress” in developing their own advanced chips
A former ASML worker accused of stealing trade secrets for advanced chip-making equipment from his employer is now suspected of spying for the Chinese government.
Citing sources familiar with an internal ASML probe into the alleged theft, Bloomberg reports the employee, who is apparently based in China, had “potential” ties to a Beijing-backed spy ring and may have been stealing that data on its behalf.
ASML, for those who don’t know, is headquartered in the Netherlands, and makes the highly specialized equipment used by Intel, TSMC, and others, to fabricate silicon chips in their own fabs.
The blueprint theft, ASML’s second in less than a year, comes as the US is pressuring its allies to starve out China’s domestic chipmaking efforts. This campaign includes barring the export of lithography machines built by ASML and others that are used by foundries - such as TSMC and Samsung Electronics - to produce advanced semiconductors used in smartphones, laptops, and servers.
Just a couple days ago I was told on these very pages that there isn’t anything China could invest in that would be worthwhile, and therefore its economy is sunk, ghost cities and declining population and all.
Why couldn’t (wouldn’t) they “invest” in a Manhattan Project to further their technology industries and make these advanced chips available to themselves? Oh, I know, it’s terribly complicated, with layers of chip fabs and techniques and softwares and supporting industries required to build one on top of another.
Then again, the Manhattan project had to invent a host of technologies out of thin air: atomic research, shaped explosives, fission piles, and so on. And they had to make them practical: isolating U-238 atoms, perfecting multiple timing triggers, metallurgical questions, and more. We created three cities out of nothing but our imagination, invented technologies that never existed, and produced the atomic bomb in a little over 3 years.
It cost $2B in 1945 dollars, which scales to about $35B in today’s equivalent. Certainly China could afford a project on this scale - or larger. Would it work? Well, there was no guarantee the Manhattan Project would work, either (and in fact Germany was pursuing an atomic project, but took a wrong fork in the road and was later starved of resources, so nothing is guaranteed.)
Will they? They - and other Asian (and directed) economies have shown an ability to target an industry or a segment and then run full force at it. Will this one be next? If “chips” is as important to the future as some here believe, then China would be foolish not to.
China has spent billions of dollars in recent years trying to catch up to the world’s most advanced semiconductor makers.
Two foundry projects, led in part by a little-known entrepreneur then in his 30s, help show why China has yet to succeed.
The projects, in the Chinese cities of Wuhan and Jinan, were supposed to churn out semiconductors nearly as complex as the more-sophisticated chips made by industry leaders Taiwan Semiconductor Manufacturing Co. and Samsung Electronics Co., which have decades of chip-building experience.
Chinese officials kicked in hundreds of millions of dollars to support the upstarts. But it quickly became clear the plans had been too ambitious, and local officials had underestimated how difficult—and costly—it is to make complex high-end chips.
The two foundries, Wuhan Hongxin Semiconductor Manufacturing Corp. and Quanxin Integrated Circuit Manufacturing (Jinan) Co., burned through cash, yet never commercially built any chips.
Over the last few years I’ve seen articles purporting that Chinese scientists are leaving the US, and taking their skills back to China.
Here’s an example:
Noted structural biologist Yan Ning became the latest prominent addition to a growing list of Chinese scientists returning to China when she announced on Nov 1 that she would resign from her tenured professorship at Princeton University and return to China to help build the Shenzhen Medical Academy of Research and Translation.
Is the China Daily a reliable source?
Last year, over 1,400 US-based ethnic Chinese scientists …
$Snip> Many … had relinquished tenured positions at top US universities, the report said. They worked in key disciplines, including physical and life sciences, engineering, mathematics and computer science.
A…$snip>.s, but there has been a noticeable outflow of Chinese scientists leaving the US to return to China in recent years.
“This is the result of China’s rising recognition and support for quality talents on the one hand, and the US prosecution and underutilization of Chinese researchers on the other,” the scientist said on condition of anonymity for fear of being targeted by the US government.
There are more examples from 2022, 2021.
Here’s a NYT article. Pay walled.
IF these stories are true, when will the people with the right skill set for advanced chips take those skills “back”?
@Goofyhoofy mentioned a Manhattan project effort by China.
The US Manhattan project was staffed by refugees from Hitler’s shadow?
That 3 year success was made possible by a brain n skill drain to the US?
Even when we are lied to in the US generally we do not go that deep full on liar.
Except the orange one.
@rainphakir The software industry a few coders do better than a thousand coders. It is more efficient to work in smaller formations. Hardware takes a larger interwoven approach to corporate structures. The Chinese can not just take home a few scientists and make it work.
For each Chinese scientists we have here in the US over the next decade the fewer are working in China. We are talking the very top people. Those folks have free reign to stay in the US. Go home to China will they be charged with treason? Isn’t being Chinese in the first place and living in China now the same as treason? Fair question when you have a dictator like Xi.
Several articles have talked about highly educated foreign nationals leaving the US due to the increasing xenophobia, and increasing difficulty becoming a permanent resident, and on a path to citizenship. Canada recently had a test offer, aimed directly at recruiting people in the US on an H1B visa. The 10,000 slots in the pilot program were filled in two days.