In Japan, the Government Is a Go-to-Market Channel

In April, METI sanctioned a further ¥631.5 billion for Rapidus, bringing cumulative government R&D support to roughly ¥2.35 trillion. In June, an additional ¥150 billion capital injection completed, funding prototype improvement and front-end wafer processing at the IIM-1 facility in Chitose.

Those are large numbers and they got the coverage large numbers get. The sentence that mattered was quieter, and it appeared in METI's own framing of the April announcement: NEDO will support semiconductor design projects alongside Fujitsu and IBM Japan, with the goal of boosting domestic demand for Rapidus's future products.

Read that as a market participant rather than a policy observer. The state is not only financing the factory. It is using public R&D money to create the customers.

What is actually happening

The logic is straightforward once you see it. Without a steady pipeline of wafer orders, an advanced fab cannot improve yields or justify its operating costs, subsidies notwithstanding. So the NEDO–Fujitsu–IBM design work is not peripheral to the semiconductor programme. It is a deliberate demand-creation strategy intended to ensure there are anchor customers ready to consume output from the first day of mass production, targeted for 2027.

Public money is flowing to both sides of a market that does not yet exist, in order to bring it into existence.

The technology standpoint

This matters technically because it determines where the design work happens.

Chip design decisions made in 2026 under NEDO-commissioned projects will determine what gets fabricated in 2027 and beyond. Design tools, IP blocks, packaging approaches and verification methodologies selected during those projects become defaults. A foreign company whose technology is present inside that design work is embedded in the resulting supply chain. A foreign company that arrives in 2028 to sell into it is selling against choices already made.

The commissioned projects themselves span 2nm integration, short turnaround-time manufacturing, and chiplet and packaging design — and several are explicitly framed around Japan–US collaboration, which is a meaningful detail for American companies in particular.

The business standpoint

Here is the reframe I would offer to any foreign deep-tech company looking at Japan.

Most treat government programmes as grant funding: non-dilutive capital, administratively burdensome, nice if you can get it, not central to commercial strategy. In Japan that framing leaves the most valuable part on the table.

A NEDO commissioned project is a contract to perform work. It brings revenue, but it also brings three things that are harder to buy:

Institutional validation. In a market where the central buying anxiety is whether a foreign vendor will still exist in five years, participation in a national programme answers that question in a way no reference customer can. It shortens every subsequent enterprise conversation.

Proximity to the companies that will buy. These programmes are executed alongside the large corporates in the sector. A year inside one puts you in working relationships with organisations whose procurement front door would otherwise take eighteen months to open.

Influence over what gets specified. The standards and defaults emerging from these projects shape the eventual market. Being in the room is worth more than any marketing spend available to a company of your size.

This is also the part most foreign companies misread. A NEDO or METI programme is not a grant application dressed up as paperwork — it is a competitive process that funds work which is simultaneously research and product development. Genuinely novel commercial work usually does qualify. The difficulty is recognising which parts of it qualify, and describing them in the language the programme is written in.

The market standpoint — what this means if you're selling here

Treat public programmes as a channel, and resource them accordingly. If you would assign a person to a major partner relationship, assign one here. Most foreign companies delegate this to a consultant and receive consultant-quality outcomes.

The inaccessibility is the opportunity. These programmes are documented largely in Japanese, structured in ways that do not map onto foreign grant processes, and have application cycles that reward people who understood the priorities before the call was published. That is precisely why competition for them from foreign companies is thin. Inaccessibility is a moat when you are on the right side of it.

Watch the policy direction before the funding lands. METI publishes strategic priorities well ahead of the money. The physical AI sector target announced in March is a statement of where public capital is going next. Companies positioning against those priorities now will be eligible for programmes announced in 2027.

Understand what you are taking on. Public money brings obligations: reporting, milestones, occasional expectations about local presence and employment. It also brings alignment constraints — participate in a national programme and you have made a statement about which side of certain geopolitical lines you sit on. For most companies that is fine. It should still be a decision rather than a discovery.

The broader principle

Japanese industrial policy is not a backdrop to commercial strategy here. It is an active market participant with a balance sheet, a set of priorities published in advance, and a demonstrated willingness to build demand as well as supply.

Foreign companies routinely model the competitive landscape, the channel landscape and the customer landscape, and then treat government as regulatory context — something to comply with rather than something to sell to or build with.

In Japan that is a modelling error, and in deep tech it is an expensive one.