Issue:
March 2025 | Letter from Hokkaido
The official start date for Hokkaido’s ambitious 2 nm semiconductor project is still anyone’s guess
Decades after Japan lost its leadership in the semiconductor industry, and following years of preparation, Rapidus is gearing up to produce the world’s first 2 nm semiconductor chips at its plant in Hokkaido.
Pilot production is reportedly due to start “soon” – the quotation marks deliberately used to emphasize the loose timeline suggested by Rapidus, Hokkaido officials and media, as well as visitors to the plant.
Not surprisingly, the likely start date for pilot production comes with contradictions and often confusion.
I will avoid dredging up debates about whether or not Rapidus will succeed, and focus instead on scheduling issues.
While some reports last year suggested pilot production would kick off in April, it appears that preparations are running behind schedule. The delays, we are told, are to be expected in any project of this scale and complexity. Keep calm and carry on, seems to be the message.
The April starting date seems pretty optimistic. For pilot production to begin next month, a lot of things on the foundry floor will have to go smoothly at the first attempt, which would be a statistical anomaly more akin to a miracle.
It is more likely that pilot production will begin in late May or even June, and that’s assuming that the current issues get ironed out without further delays and no major new, unforeseen problems arise.
What happens once pilot production begins? A lot of things, but they’re all aimed at ensuring that full production of the chips is underway by 2027, and that they are produced at a pace and level of quality that ensures there are enough to meet customer demand in a highly competitive market.
As soon as pilot production begins, the next-generation foundry in Chitose will see engineers and technicians, many of them from abroad, arriving to test the equipment and the 2 nm chips themselves. Everything must pass rigid inspections before full production can begin.
Semiconductor engineers and industry experts are aware of the role of Murphy’s Law in these endeavors. That’s why they are hesitant to announce even short-term schedules with any confidence, especially before testing begins or while it is taking place. Unfortunately, Hokkaido’s politicians and business community, as well as Rapidus executives and the Japanese government, have their own views on what “soon” could, and should, mean.
In fairness, Rapidus is under immense pressure to get up and running, and be successful. Prime Minister Shigeru Ishiba’s government is pumping money, and investing political rhetoric, into reviving Japan’s semiconductor industry under the assumption that the Rapidus project will proceed as scheduled – which means by 2027.
But Rapidus is doing itself no favors in Hokkaido with a communication strategy that leaves even senior local officials out in the cold. Rapidus and its supporters try to excuse the lack of communication as necessary to protect corporate secrets. Or they simply point to their press releases and invite people to contact them for information.
But that smells like a lot of what one encounters at a Hokkaido cattle ranch on a hot day. To my knowledge, Rapidus has yet to set up an office or information center in Sapporo where people from Hokkaido and beyond can learn more.
In his column on February 6, Hokkaido University President Kiyohiro Houkin addressed those who ask the question: What if Rapidus Fails?
“Non-specialists tend to hold such views, which could have collectively been the architects of Japan’s lost decades. Their concerns are often vague and unfounded, leading to the typical emotional fears and, therefore, the avoidance of challenges in those generations,” he wrote.
That kind of statement comes easier if you are a tenured academic with guaranteed lifetime employment, rather than a corporate leader responsible to your shareholders, an engineer worried about job losses if Rapidus fails, or a Hokkaido politician concerned about the transparent, efficient allocation of increasingly limited public funds for Rapidus-related projects.
Sadly, it is symbolic of the communication approach many in Hokkaido feel that Rapidus is taking, even though pilot production is set to begin. When will that happen? Soon, of course.
Eric Johnston is the Senior National Correspondent for the Japan Times. Views expressed within are his own and do not necessarily reflect those of the Japan Times.
