Semiconductor wars and China’s huge IPO
Even $7.6 billion may not be enough to break the Western tech monopoly
Enjoy this extra edition of Contention News! It should take just about five minutes to read. Follow us on Twitter and Facebook. If you like it, please subscribe:
China’s largest semiconductor manufacturer, SMIC, debuted in Shanghai last week and tripled in price almost immediately, raising $7.6 billion for the company. This was the largest IPO in China in 10 years, and brings hope to the country’s plans for independence in the crucial industrial sector.
A full breakdown of the strategic picture, however, makes it clear that this is at most a small advance, with years of challenges and uncertainty still to come for the People’s Republic.
At the heart of recent economic friction between the United States and China: Made in China 2025 -- a set of goals to bring more advanced production within the country’s borders by 2025, including 70% of China’s semiconductor market.
SMIC and other manufacturers can easily make mature technologies, but their most advanced design, N+1, is still less powerful than technologies used by TSMC, Samsung, and others since 2018. Those companies are now moving beyond that point, and China is destined to be a buyer of those chips, not a producer.
Two key strategic choke points block China from catching up: the machines which foundries use to print the chips, and the software used to design them.
Most state-of-the-art chips and all currently planned technologies can only be made with extreme ultraviolet (EUV) lithography. Only one company in the world sells EUV equipment -- ASML, based in the Netherlands. There are fewer than 50 of these machines globally, and the minimum capital cost for an operation using EUV is $1 billion.
SMIC tried to buy an EUV machine in 2018 only for the United States to block the sale with high-level pressure on Dutch PM Mark Rutte. China is investing in developing its own EUV technology but is likely years away from anything viable.
Electronic design automation (EDA) software is even more directly under U.S. control. Three firms dominate the field, all headquartered in the United States -- Synopsys, Cadence Design Systems, and Mentor Graphics. Mentor is owned by Siemens, a German company, but based in Oregon.
You can't map out billions of transistors manually -- EDA is necessary to make chips. Software is obviously much easier to pirate, and China recognizes Western intellectual property as a rent-seeking obstacle to global technological advancement. But this is not a sustainable solution, and China’s domestic EDA industry is starved for talent and other key resources. Recent U.S. restrictions on high-skill immigration could help reverse this trend, but it won’t mean much if they can’t get the equipment to print the designs.
China has opportunities for reversing their conundrum, each with potential drawbacks.
They could replicate the model that worked for SMIC: identify small but effective private companies in the equipment and EDA sectors, pump them up with substantial state-backed investment, and then take them public to add even more capital from domestic retail and foreign institutional investors.
They are midstream in this strategy with fabrication equipment maker AMEC, but for EUV in particular it’s not just a matter of money -- this process will still take years, and with rapidly escalating aggression from the United States they may not have that time.
That same political chaos might open up the second possibility: ASML, Siemens, or other EUV and EDA companies could defy Washington to sell equipment and necessary technologies to Chinese customers. These companies’ governments are foregoing substantial revenue to maintain their strategic relationship with the United States. Ongoing U.S. political failure could make foregoing the sales a bigger risk than losing the relationship.
The third possibility is much higher stakes, but may already be under way: China can leverage its monopoly power at the other end of semiconductor supply chains, in rare earth materials.
Just last week two stories laid this leverage bare:
China raised production quotas for the 17 minerals under direct state control, which comprise 80% of the global supply for these strategic elements.
U.S.-based rare earth miner MP Mining raised $1.47 billion through a reverse merger with Fortress Value Acquisition Company, a SPAC. The money is meant to finance new refining infrastructure for the company’s minerals, 100% of which are now sold to Chinese refineries.
China seems hesitant to fully exercise such an aggressive option, which eliminates the fourth and most dramatic possibility -- ending Taiwan’s de facto independence to gain control of TSMC and other chip-making industries.
In all cases it appears that China will fall far short of their Made in China 2025 goal for semiconductors, but they have nonetheless raised the stakes in this sector. Will China be able to break the U.S. strategic monopoly in the tech industry, and thus their power to favor their largest investors over literally everybody else in the world? The billions flowing into SMIC last week think they very well might.
Contact us with feedback or stories you think we should see.