Rapidus Begins 2-nm Chip Pilot Production, Vital for Japan's Tech Future

このページを日本語で読む

Rapidus is leading Japan's push to establish domestic production of next-generation semiconductors. In April, the company began prototyping chips at its factory in Chitose, Hokkaido. This marks a crucial step in tackling two of its biggest challenges: attracting customers and securing funding.

To get the business off the ground, Rapidus must prove it can deliver high-performance chips — and demonstrate that capability to the world.

The 2-Nanometer Challenge

In semiconductors, performance improves as circuit line widths become narrower. Japanese manufacturers once held a majority share of the global semiconductor market, but they fell behind in the race toward miniaturization. Today, Japan can only produce general-purpose logic semiconductors with 40-nanometer line widths.

Now, Rapidus is attempting to domestically produce next-generation 2-nanometer semiconductors. No company in the world has yet mass-produced chips at this scale. So why is Japan, which once lost its edge in miniaturization, now positioned to take on such an ambitious challenge?

The answer lies nearly five years in the past. Tetsuro Higashi, former president of Tokyo Electron and now chairman of Rapidus, was approached by IBM with a proposal: "Can Japan manufacture semiconductors using IBM's 2nm technology?"

*Chairman Tetsuro Higashi of Rapidus on April 27. (©Sankei by Hideo Iida)*

Formation of Rapidus

Behind this question were growing geopolitical concerns about semiconductor supply chains. In the field of advanced semiconductor contract manufacturing, Taiwanese companies — particularly TSMC — held over 90% of the global market share.

A Taiwan contingency could seriously disrupt the supply of advanced chips. Securing production bases outside of Taiwan had become a strategic priority.

After evaluating IBM's technology and confirming its viability, Japanese stakeholders decided to move forward with the domestic production of 2-nanometer semiconductors. Rapidus was established in August 2022, about two years after IBM's proposal.

Fortunately, Japan still had a pool of experienced engineers from its semiconductor heyday. Rapidus recruited many of them, sending around 150 to IBM's research labs in the United States to master the new technology. Around 70 have since returned and are now working on prototype development.

Mass Production Challenges

The first prototypes are expected between mid- and late-July. Rapidus will begin by producing standard chips using IBM's technology. Customer-specific prototypes will follow, with the goal of delivering them within fiscal 2025.

To improve yield — the ratio of usable chips — Rapidus is using a method called "batch-type" or "single-wafer processing," in which wafers are handled one at a time. While not ideal for mass production, this approach allows real-time feedback to the design team. Defects can be quickly addressed, enabling faster improvement of the yield rate.

Still, mass-producing these advanced chips is a formidable task. At a press conference on April 1, Rapidus President Atsuyoshi Koike admitted, "Developing mass production technology is extremely difficult. Confirming yields and ensuring reliability are major obstacles." Whether the company can begin mass production by its 2027 target remains uncertain.

Can Rapidus truly mass-produce 2-nanometer semiconductors? Some companies remain skeptical. The prototypes will be key to dispelling doubts and gaining trust.

Government Support and Economic Stakes

So far, the Japanese government has pledged over ¥1.7 trillion JPY (about $11.6 billion USD) in subsidies to Rapidus. On April 25, a new law was enacted to expand financial support. This includes direct investments and debt guarantees. A ¥100 billion (about $658 million) investment is planned for later in 2025.

Such massive government backing has drawn criticism. If domestic production fails, the burden will fall on taxpayers — making failure unacceptable.

On the other hand, if Rapidus can quickly supply cutting-edge semiconductors tailored to the specific needs of industries — such as autonomous vehicles — it could have a significant ripple effect across Japan's broader industrial base. The scale of government support reflects how strategically important next-generation semiconductor production has become.

"If Rapidus fails, semiconductor equipment and materials manufacturers [areas where Japan still has an edge] could begin relocating overseas," warns Hideki Wakabayashi. He is a Distinguished Professor at Kumamoto University and a member of the Ministry of Economy, Trade and Industry's Semiconductor and Digital Industry Strategy Committee.

"In that case, Japanese companies would lose their edge in products that use advanced logic chips," he continued.

Rapidus's mission goes beyond reviving Japan's semiconductor industry. In a country still searching for a new growth engine to follow the automotive sector, the success of this initiative could shape Japan's economic future.

(Read the article in Japanese.)

Author: Shunichi Takahashi, The Sankei Shimbun

このページを日本語で読む