| Manufacturing 15 Bird Mobile | MR Online Manufacturing #15, Bird Mobile Ningbo, Zhejiang Province, China, 2005 (Photo: Edward Burtynsky)

Chip wars or the crisis of late capitalism?

Originally published: Peoples Democracy (August 29, 2021 )  |

WITH the U.S. imposing technology sanctions on China, the world’s electronics industry is facing turbulent times. After the sanctions, Huawei has slipped from its number one slot as a mobile phone supplier (2nd quarter 2020) to number seven currently. Commenting on this slide, the Huawei chairman has said that right now, Huawei’s battle is for survival. On that count, Huawei is not only surviving but doing pretty well. It is still the world leader in the telecom equipment market with a hefty 31 per cent share, twice that of its nearest competitors Nokia and Ericsson, and profits of nearly $50 billion in the first six months of 2021. But will it be able to retain its market position without China catching up in chip manufacturing and design technologies?

It is not the Chinese companies alone that are facing tough times. With growing chip wars between the U.S. and China, the global supply chain for electronic chips has been affected and chip shortages have emerged. Semiconductor chips enter almost every product, from our lowly household equipment—microwave ovens and toasters—to the most advanced products in the industry. The car industry’s biggest bottleneck today is this chip shortage, which has badly hit their production. If the chip wars continue, the crisis of a chip shortage can affect other industries as well.

Is the crisis of the semiconductor industry a precursor to the fragmentation of the global supply chains? Will it lead to warring blocks, with the U.S. at one pole and China at the other? With this fragility of the supply chain, are we seeing the end of globalisation as a paradigm?

Chip making and the electronics industry is one of the most capital and research and development (R&D) intensive industries. No other industry has this characteristic. Power or steel plants are capital intensive; pharmaceuticals are R&D intensive. But not both. A little-known Dutch company ASML that produces the lithographic machines for chip manufacture has a market capitalisation more than that of Volkswagen, the world’s largest car manufacturer! This is due to the R&D costs that go into their machines: ASML is the only equipment maker that can deliver the machines that the most advanced chips require. A new fabrication facility to make the new generation of chips today cost $20 billion, more than that of an aircraft carrier; or a nuclear plant. Only two fabricators, TSMC in Taiwan and Samsung are capable of producing the most advanced chips that the industry uses.

The U.S. and China are in competition in areas such as artificial intelligence, computers, mobile networks and phones. The basic building block of all of these technologies is semiconductor chips. The more circuity we can pack into a chip, the more its computing power. The bulk of the market still consisting of older fabricators using 180 nm to 28 nm level technologies. Only two per cent of the chips, the most advanced chips, are below 10 nm (<10 nm) level. The only fabricators that can make such chips are TSMC and Samsung, the world’s largest fabricators. While SMIC of China, the third-largest chip fabricator in the world is also at 14 nm level, it has only recently moved from 28 nm to 14 nm. With Chinese government support, it is investing in production lines that can go below 14nm. Intel, once the world leader in chip manufacture, is stuck at the 14 nm level, though it also has plans for developing the next generation of chips.

The U.S. has chosen the electronics/semiconductor industry as a battleground for its geostrategic competition with China. It believes that here, it has a significant technology lead and a major market share. China is a late entrant and though has a comparable market share, is still dependent on certain core technologies that the U.S. and its allies—the European Union, Japan and South Korea—control. That is why the U.S. has chosen Huawei and Semiconductor Manufacturing International Corporation (SMIC), two major Chinese players as targets for its sanctions.

The U.S. is following up its sanctions on Huawei and SMIC with a plan to bar China from technologies it calls “foundational technologies” under its 2018 Export Reform Control Act. The argument that the U.S. is building is a simple one: “ we are ahead of China in certain critical technologies required for advanced chip manufacture; all we have to do to maintain our lead is to deny China access to these technologies; that will ensure that our lead for the future and ensure the dominance of the US.”

At first sight, it might appear that chips should be considered as foundational technologies and the target of U.S. sanctions. This is what the U.S. did when it barred Huawei from buying the latest 7 nm scale chips from Taiwan’s TSMC, the world’s largest chip fabricator.

The SMIC then tried to set up its fabrication line for 7nm chips and needed to import from the Dutch company ASML Extreme Ultraviolet Lithography (EUV) machines, each costing around 120-150 million dollars. Though the EUV machines are from the Netherlands, it uses software developed in its U.S. subsidiary and therefore has under the U.S. sanctions regime.

The U.S. sanctions mean that ASML cannot sell the EUV lithography machines to China, though it can sell other lithographic machines for lower-end chip production, keeping China out of the high-end <10 nm technology, and therefore a generation or two behind the market leaders.

This brings us to what is a foundational technology. Clearly, chips, though the key driver of electronics is not as foundational as the machines that produce these chips. If we want to be at the cutting edge of technology we have to master not only the technology of chip production but also the technology that produces the machines that run the production lines. That is why ASML’s lithography machines are the bottleneck for China.

What then drives the advances in key technologies of the machines and chip production? As Marxists know, it is the knowledge that drives the productive forces, in this case, the advances in chip design. This knowledge is captured in the software design tools and of the lithography machines. They are both highly knowledge-intensive and require people with very specialised skills.

The U.S. and its universities are still the major source of knowledge development, the key to the advances in this sector. But here is its long-term problem. The research programs of U.S. universities are mostly staffed with foreign students, the bulk of them from China, India and other developing countries. A large number of them stay back in the U.S. and provide the human power required for the advances in knowledge that the U.S. has today.

If Chinese students and researchers are not welcome in the U.S., this source of knowledge development is going to weaken. Unfortunately, countries like India have not made the investments in high-quality education and research laboratories to provide a substitute for the stream of Chinese students that enter the U.S. universities. On the other hand, China has invested heavily in its universities and research institutions, overtaking the U.S. in producing a larger number of PhDs in science and technology. It is also building a strong pipeline of innovations from the universities/research institutions to the industry.

The fear of the U.S. semiconductor industry is that China is the biggest market for the U.S. for its chip design software. The U.S. companies also design high-end chips which are then manufactured in Taiwan. In the short run, the U.S. sanctions will certainly damage the Chinese in advanced chip production and the production of electronics devices based on such chips. But it will also mean that the U.S. companies will lose a significant part of their revenues that it now receives from the Chinese market for its sale of design tools. It also means a loss of revenue for advanced chips that the U.S. companies like Qualcomm and Nvidia design and then manufacture in Taiwan’s TSMC.

For the high-tech U.S. companies, the loss of this income means less money for R&D and the slow erosion of its position as the global knowledge hub. If the U.S. companies lose the Chinese market and therefore a significant part of their revenues, their ability to compete in the future will be seriously impaired. They may gain in the short run, as they are doing with Huawei losing its number one spot in smartphones, but in the long term, the loss of revenues will mean less ability to produce the knowledge that gives the U.S. its edge in technology. Less money in research means an eventual loss of leadership, because, unlike other countries, the U.S. increasingly does not produce the chips or the machines, but the knowledge that goes into both.

In a series of submissions to the U.S. department of commerce, the U.S. semiconductor industry has argued that the U.S. companies de-linking from the Chinese market will mean a huge loss of revenue and eventually giving up the U.S. leadership in electronics. Already, the U.S. sanctions have led the Chinese companies to take out U.S.-designed components from their product lines. Sanctions, therefore, have not only hit Huawei and other Chinese suppliers, but they have also hit the U.S. companies who were their suppliers.

How long will China take to erase the lead that the U.S. and its allies have in semiconductor technologies? Analysis Mason, a leading consulting company says in its May 21 report that China will be able to attain self-sufficiency in semiconductors in three-four years. The Boston Consulting Group and Semiconductor Industry Association have modelled the impact of breaking up the global supply chain of China and the U.S. delinking their supply chain and markets. It predicts that with such a policy, the U.S. would still lose its leadership to China. According to the Semiconductor Industry Association’s submission to the U.S. department of commerce, the only way that the U.S. can preserve its lead is if it allows exports to China—except in the military sector—and uses its profits from these sales to develop the new generation of technologies. Of course, along with hefty subsidies from the U.S. government!

Where is India in semiconductor manufacturing? India missed the semiconductor manufacturing bus when it decided not to rebuild the Semiconductor Complex in Mohali after it was destroyed in a fire in 1989. Its policymakers decided that India should leverage its strength in software and systems and not worry about manufacturing chips. Vinnie Mehta, formerly the executive director of Manufacturers’ Association for Information Technology (MAIT), had said to Mint ten years back, “A nation without silicon (technology) is like a person without a heart.” That heart is still missing in the Indian technology ecosystem.

If the U.S. wants to be a world leader, it has to match China in investing in knowledge generation for future technology. Why then is the U.S. taking the sanctions route? Sanctions are simpler to implement; building a society that values knowledge is much more difficult. This is the crisis of late capitalism.

Monthly Review does not necessarily adhere to all of the views conveyed in articles republished at MR Online. Our goal is to share a variety of left perspectives that we think our readers will find interesting or useful. —Eds.