How well or badly is the US-China Chip war going?

After the Biden Administration announced its industrial strategy to revitalize domestic manufacturing, create jobs, strengthen American supply chains and accelerate future industries, the CHIPS and Science Act of 2022 will give as much as $280 billion over the next decade to support US semiconductor chip industry.    

Goodbye free markets, hello industrial policy.   The outcome of the US-China rivalry hinges on the technology edge and how to use such technology.    

In 2018, the Trump administration ban on Chinese tech giant ZTE to buy sensitive components and software from American companies signalled the start of a tech war.  The same year, Huawei was put on the “Entity list”, requiring US Government approval to buy US technology.  This story is rivetingly told in Chip War, written last year by historian Chris Miller. 

In a recent interview, Miller cited three reasons why the US invoked the Chip War.  First, the U.S. concern over Chinese intentions over Taiwan, as Taiwan Semiconductors Manufacturing Corporation (TSMC) is the leading producer of cutting-edge semiconductor chips.  Nano-sized chips are essential in the next generation military and intelligence capabilities. Second, China has certain advantages in narrowing the US-China tech gap.  Third, as export controls and sanctions have limited effect, such as those against Russia proved, there was no point in waiting to restrict Chinese access to foreign technology. 

With Taiwan as a “choke point” in US-China competition, the Biden administration is adopting a two-prong strategy to compete or contain China in technology.  The first part would shift production partly out of Taiwan to “onshoring” or “friend-shoring” allies willing to de-risk dependence on Chinese production.   The second part would bring American semiconductor production back home, which has fallen from 37% in 1990 to about 12% of global output.  Moving TSMC and Samsung chip production to US soil are efforts in that direction. 

The real killer is the “choke point” strategy, which means that you strangle your rival at his most vulnerable supply chain points.  China has always been vulnerable to energy imports, hence the strategic importance of the Malacca Straits.  Avoiding this explains China’s remarkable shift to home-based solar energy, installing 413 gigawatts of solar capacity or 44% of its own electricity usage – electricity being the key driver of the digital economy.  As Miller shrewdly points out, “China now spends more money each year importing chips than it spends on oil”.

Why are semiconductors so critical in the new Great Power rivalry?  “Last year, the chip industry produced more transistors than the combined quantity of all goods produced by all other companies, in all other industries, in all human history.  Nothing else comes close.” 

Semiconductors have become so small and so fast, with so much computing power, that they are the foundations of anything “smart”.   My iPhone has 10 million times the processing capacity of the 3 tonne IBM 360 mainframe business computer that I used in the early 1970s.  As the Ukraine war showed, handphones with GPS determine the next smart tool to fight precision-drive drone wars.    

We have also moved from hardware to software, because it is the software applications that ultimately create the Artificial Intelligence (AI) computations beyond the capacity of mere human beings.  The Chinese are great with hardware, but in the software side, there is still a huge gap with the Americans, partly because China has not yet created the complete tech start-up eco-system that exists in Silicon Valley. 

It is illuminating to see that even though China has its own versions of ChatGPT and NASDAQ equivalent in the Shenzhen stock market, ChatGPT sparked the equity revival of the US Magnificent Seven tech stocks (Alphabet, Amazon, Apple, Meta Platforms, Microsoft, Nvidia and Tesla) with a combined market cap of over $11 trillion.  Nothing like that happened in China this year.   

One clue is that Chinese regulators worry about stock market bubbles, whereas the 2000 Nasdaq tech bubble taught the Americans that tech bubbles are not systemically fatal, but their wealth creation, if ploughed into the next generation of start-ups, create new commercial (and military) technologies.  Funding market tech innovators is key to next generation technology.

However, the real choke holds over the Chinese are the extreme ultra-violet (EUV) lithographic equipment, which are so precise that they can etch nano-chips into highly compact integrated circuits that are free of bugs and technical flaws.  These need such clean contamination-free manufacturing environment that all workers have to wear space-suits with diapers.   Unfortunately for the Chinese, only the Dutch ASML can manufacture the high quality EUV lithographic machines that TSMC and Samsung need to fabricate the most advanced logic chips.  In essence, the US has “weaponized interdependence” hold on the Chinese, who are at least a decade or more away from creating their own EUV machines. 

Make no mistake.  The Chinese can manufacture or buy the less advanced chips that are the workhorse of consumer “Internet of Things” products.  But as AI exponentially demand more computing power, China will be stunted in her cutting edge technology without access to such nano-chips that are specifically designed and fabricated for dedicated usage.  If quantum computing becomes commercially viable, the demand for high end chips will be even more critical.

The Chip War is really about scale and imagination in industrial policy, with gorilla-sized resources and talent thrown at how to wrestle the opponent to the ground through key choke points.  Techno-nationalism means that whoever has the best eco-system of innovation, talent, funding and production dexterity will have the edge over the others.  This is not a 100 meter sprint but a bruising, brutal and ugly Long March to techno-superiority.  So far, the US and her allies has the edge, but China has the market scale.  If by 2030, China accounts for one-quarter of the market, versus 10 percent for the US, as the Semiconductor Industry Association has estimated, who knows who will really have a decisive edge in the Chip War? 

Get real.  Tech wars are ultimately deadly.

We are no longer on the ring-side watching fantasy New Roman Coliseum tech gladiators killing each other.  The next Kill generation game players may end up with Generation Dead.   

Contributed by

Andrew Sheng