How realistic are China's semiconductor ambitions?
When US President Joe Biden landed in South Korea last month, he headed straight for Samsung's massive semiconductor factory outside Seoul. There, he met with South Korean President Yoon Suk-yeol and Samsung Electronics vice-chairman Lee Jae-yong, and praised the construction of a US$17-billion Samsung semiconductor factory in Texas. The economic and strategic importance of semiconductors could not have been made any clearer.
During the Covid-19 pandemic, semiconductor supply disruptions forced a range of industries -- from automobiles to consumer electronics -- to slow or halt production. Reliable semiconductor supplies, it became clear, are vital to a country's economic resilience. For the United States and China, they are also central to a strategic competition in which leadership in cutting-edge industries plays a crucial role.
As it stands, the US has a bigger slice of the global semiconductor pie, owing to its strength in chip design and in the fabless segment of the industry. But the vast majority of chips are manufactured far from America's shores, including at the Samsung factory in Chinese President Xi Jinping's hometown of Xi'an.
And China -- the world's biggest chip market -- is investing heavily in the sector as part of its effort to boost indigenous innovation. So, is the US set to lose its semiconductor edge?
So far, China has struggled to catch up. First, the typical latecomer strategy -- focused on building cheaper, low-end products -- cannot be applied to semiconductors, because a more advanced "next-generation" memory chip tends to cost the same or less than its predecessors. Less advanced chips are thus virtually worthless.
This is not to say that the incumbents' positions are unassailable. After all, South Korean firms like Samsung managed to overtake more established Japanese companies like Toshiba in semiconductors. The key is a "leapfrogging" strategy: developing more advanced versions of a technology before the incumbent can.
Such a strategy requires that the development of a technology follows a relatively predictable path -- in the case of chips, upgrading from a capacity of one kilobyte to 2K, then 4K, and so on -- and that companies have access to technology from abroad.
South Korean companies like Samsung never made the lowest-capacity chips. Instead, they used equipment and facilities imported from Sharp in Japan, and circuit designs licensed from Micron Technology in the US, to begin developing 64K chips immediately upon market entry.
Later, Samsung set up a research and development facility in California's Silicon Valley, in order to develop designs for high-capacity (256K) chips ahead of Japanese firms. Its use of the "stacking method" for increasing chip complexity -- rather than the "trenching method" used by firms like Toshiba -- helped to propel progress. But Samsung continues to rely on high-tech components, parts and supplies from Japanese and other foreign sources, and on software from the US.
At a time when China's access to foreign technologies and equipment is increasingly constrained, the country will have a hard time replicating this leapfrogging strategy. Semiconductors and other cutting-edge industries are dominated by a very small number of firms. In some cases, just one or two companies can provide a particular input or piece of equipment.
These firms are concentrated largely in the US and Europe. One Dutch company, ASML, is the only producer of extreme ultraviolet (EUV) lithography machines, which are vital to the chipmaking process, and US firms dominate software.
This is not to say that China has no chance to develop an advanced -- even world-leading -- semiconductor industry. While this will certainly not happen overnight, there are opportunities to boost China's prospects.
For starters, while the market for memory chips is uniform -- lacking high- or low-end segments -- the market for system chips (or application-specific integrated circuit chips) is segmented according to application.
Automobile companies, for example, do not use the most advanced fabrications, manufactured through cutting-edge sub-10 nanometre (10nm) lithography. Instead, they use 20nm or 30nm technologies, for which technology transfer is not tightly controlled. In this segment, the Chinese foundry maker SMIC is reaping huge profits, which can be channeled towards investment in advanced or future-generation chips.
True leapfrogging success, however, will probably depend on China's ability to carve a new technological path that diverges from the path taken by industry incumbents and thus relies less on Western technologies.
For example, Micron Technology claims that next-generation chips can be developed using a last-generation process machine -- "deep ultraviolet lithography" (DUV) -- instead of EUV. This kind of alternative thinking could go a long way towards boosting China's semiconductor prospects.
Here, China's rapidly growing scientific capabilities will work in its favour. From 2013 to 2018, China's share of IT journal articles rose from 22.4% to nearly 40%, while America's fell from more than 20% to 16%.
In any case, restrictions on China's access to foreign technology might soon begin to be loosened. Some argue that, by constraining supply, restrictions on Chinese manufacturers, including chipmakers, are contributing to rapidly rising US inflation. The US Innovation and Competition Act is supposed to counter this effect by funneling $50 billion in subsidies to semiconductor firms, possibly including foreign-origin firms like Samsung or TSMC.
But critics point out that companies could waste the subsidies by using them for, say, stock buybacks, rather than investing in factories. And the Act might not be implemented at all.
The Biden administration faces a dilemma. If it loosens restrictions on Chinese firms, including chipmakers, it could help offset inflationary pressures, while potentially enabling China to make progress on a leapfrogging strategy. If it does not, semiconductor shortages will likely continue to exacerbate inflation, and China might ultimately find its own alternative technological path and leapfrog anyway.
Keun Lee, vice-chair of the National Economic Advisory Council for the President of South Korea, is Distinguished Professor of Economics at Seoul National University. ©Project Syndicate, 2022. www.project-syndicate.org