Japan’s Chip Startup Is Right to Shoot for the Moon

Could a little league baseball team win the World Series in five years? Japan’s state-backed chip startup Rapidus has a similarly long-shot ambition. In a country that currently produces 40 nanometer semiconductors, the two-year-old venture has the goal of leapfrogging generations of innovation and producing bleeding-edge 2 nanometer chips in 2027.

It would be easy to use a collection of data points to string together an argument that its efforts are preposterous. But in shooting for the moon, Rapidus has reinvigorated Japan’s public and private sectors to cooperate in a way not seen for generations — and over a strategically critical national security goal. This makes the startup’s success a matter of collective national pride, and a reason to be bullish that it can pull it off. 

A shortage of these crucial components during the pandemic highlighted the importance of semiconductor manufacturing to the global economy. Rising geopolitical tensions, particularly Beijing’s posturing toward chip-producing powerhouse Taiwan, has made onshoring a national security imperative from Washington to Tokyo. This led to a coalition of eight Japanese firms, with government support, to launch Rapidus in 2022 with the goal of restoring the country’s tech glory.

Once home to more than 50% of the global semiconductor manufacturing industry in the 1980s, Japan now accounts for some 10%. It did not pivot to more-advanced production during the PC era in the 1990s, putting its current outputs technologically behind rivals by roughly a decade. More recently, the explosive growth of the AI sector and the ensuing data-center boom has driven an insatiable demand for the most-advanced chips.

This all propels a new sense of urgency for Rapidus to meet the moment. The phrase “last chance” has been tossed around by many stakeholders — including company chairman Tetsuro Higashi, the Japanese government, and even the general manager from Kajima Corp. who is overseeing the construction of its Hokkaido factory.

Kajima is one of Rapidus’s secret weapons. It did not pause work for a single day through the brutal winter in Japan’s snowy northern island and has contingency plans for virtually every scenario to ensure construction meets deadlines. Kajima has quietly lead the development of at least 10 chip plants over the past decade, including the high-profile Taiwan Semiconductor Manufacturing Co. fab in Kyushu that stands stands in stark contrast with the company’s delayed factory being constructed in the US.

A significant question looming over Rapidus’s ambitions is if it can attract customers. Firms might be hesitant to place orders with an unproven startup. But the coalition of companies behind its launch — including Sony Group Corp. and Toyota Motor Corp. — could also be its first clients. Moreover, revenue from semiconductors for data centers is projected to grow 12% per year through 2030. Some of the world’s biggest data center providers like domestic titan KDDI Corp. and US-based Amazon Web Services are expanding in Japan. Rapidus set up a sales outpost in the Silicon Valley area earlier this year, and a weak yen has has made doing business in Japan more attractive for potential customers. 

Still, the startup will need to be able to mass produce the state-of-the-art chips it’s promising, a major feat in itself. But it’s partnering with IBM Corp., which was the first company in the world to develop 2 nanometer semiconductor technology. It also has research and development support from Belgium’s Interuniversity Microelectronics Centre. 

Perhaps the biggest anticipated hurdle for Rapidus is a talent shortage in Japan that developed as chipmaking moved away from the country. This issue is expected to be exacerbated by a broader shrinking workforce. Some government estimates put the shortfall of semiconductor engineers to be as high as 40,000 over the next decade. Policymakers should help by attracting foreign workers to the sector, and streamlining the visa and immigration process for this critical talent.

Rapidus is currently collaborating with higher education institutes like Hokkaido University on training. But the government should expand these efforts to start recruiting even younger students, and encourage high schools to include lessons on the importance of domestic chip development for national security reasons. Japanese people have shown a tendency to pull together in times of crisis. If students can be attracted to careers in this industry by recognizing the high stakes, this can help provide more skilled workers in the long-run. 

And it’s making up for a lack of people in its semiconductor ecosystem with a raft of other critical components. Japanese firms hold a remarkable market share on many of the materials required for chip production — another factor that gives the nation an advantage compared to efforts to onshore manufacturing in Europe or the US.

Taiwan, currently the global leader, was able to achieve this through decades of cooperation between government, companies, as well as research institutions and universities. Together, they mastered technological knowledge and dispersed it throughout the supply chain. Japan should learn from its neighbor’s success. 

Rapidus’s efforts are a moonshot, but just because something looks incredibly hard doesn’t mean it’s not worth trying, and by aiming high, it at least ensures Japan won’t fall further behind. For this young startup, even being able to compete globally with big league players would be a win.

More From Bloomberg Opinion: 

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

Catherine Thorbecke is a Bloomberg Opinion columnist covering Asia tech. Previously she was a tech reporter at CNN and ABC News.

/opinion

This article was generated from an automated news agency feed without modifications to text.