DOING BUSINESS IN JAPAN · SEMICONDUCTOR INDUSTRY
Where Japan Actually Sits in the Global Semiconductor Supply Chain
Japan lost leading-edge logic but holds structural positions in equipment, materials and components. Understanding this asymmetry is the starting point for any Japan strategy.
← Doing Business in Japan’s Semiconductor Industry
The declining-industry story is a costly misreading
Japanese firms held roughly half of global semiconductor revenue in the late 1980s. Today the figure is a fraction of that. Every summary of the industry repeats this arc, and it is accurate as far as it goes.
What the arc obscures is which business Japan lost. Japan lost the capital-intensive, cycle-exposed, scale-driven businesses — DRAM first, then leading-edge logic manufacturing. It retained the businesses where the competitive moat is accumulated process knowledge rather than capital: photoresists, silicon wafers, specialty gases, CMP slurries, bonding wire, ceramic components, and a long list of manufacturing equipment categories.
Those retained positions are not residual. In several categories, Japanese suppliers hold shares that would be described as dominant in any other industry, and the qualification cycles are long enough that switching is measured in years rather than quarters.
The practical consequence: if you are negotiating with a Japanese equipment or materials supplier, you are frequently the weaker party, regardless of your company’s size. This is not the posture most foreign executives arrive with.
Four layers, four different negotiating positions
Layer 1: Manufacturing equipment
Japan is home to several of the world’s largest semiconductor equipment makers, with particular concentration in coater/developer, etch, deposition, cleaning, dicing, grinding, and inspection/metrology. Tokyo Electron is the flagship, but the depth matters more than any single name — Advantest in test, Disco in dicing and grinding, Screen in cleaning, Lasertec in mask inspection, Kokusai in thermal processing.
Several of these companies sit in categories where two or three suppliers serve the entire world. Lead times, allocation and roadmap alignment are therefore negotiated, not purchased.
Layer 2: Materials and chemicals
This is Japan’s deepest position. Photoresist, wafers, CMP slurry, high-purity chemicals, packaging substrates, sputtering targets and bonding materials are supplied heavily by Japanese firms — Shin-Etsu Chemical, SUMCO, JSR, Tokyo Ohka Kogyo, Resonac, Fujifilm, Sumitomo Chemical, AGC, Ibiden, Shinko Electric.
The moat here is unusual: it is not patents, and not capital. It is decades of process refinement, tight co-development with customers, and qualification lock-in. A leading-edge fab cannot change photoresist suppliers because it got a better quote.
Layer 3: Components and precision parts
Beneath the equipment makers sits a supplier base of ceramic parts, quartz, seals, valves, vacuum components, precision machining and thermal-management hardware. Kyocera, CoorsTek (クアーズテック), Ferrotec (フェローテック), NGK Insulators, Ebara and a long tail of mid-sized specialists.
This layer is largely invisible from outside Japan, and it is where a great deal of M&A activity is now concentrated — the companies are smaller, often founder-owned or facing succession issues, and the technology is genuinely hard to replicate.
Layer 4: Device manufacturing
Japan’s weakest layer relative to its history. Domestic strength is now concentrated in power semiconductors (Mitsubishi Electric, Fuji Electric, Toshiba, Rohm), automotive and industrial microcontrollers (Renesas), image sensors (Sony, where Japan remains globally dominant), and NAND flash (Kioxia).
Leading-edge logic is the gap that Rapidus and TSMC’s Kumamoto investment are intended to address. Neither has yet proven out.
What the asymmetry means in practice
If you are buying from Japan, you are usually in the weaker position, and Japanese suppliers allocate on relationship history and forecast reliability rather than on price. Foreign buyers who churn purchase orders, revise forecasts aggressively, or run reverse auctions are systematically deprioritised — not out of resentment, but because the supplier’s planning system genuinely cannot absorb the volatility.
If you are acquiring in Japan, the interesting targets are usually in Layers 2 and 3, not Layer 4. They are also usually the ones most likely to be caught by inbound investment screening, because these are precisely the sectors Japan has designated as economically critical.
If you are competing with Japan, note where the incumbents are actually vulnerable: leading-edge logic, EDA, IP cores, GPU-class design, and anything requiring rapid product-cycle iteration. Japanese firms are structurally strong at long-horizon process improvement and structurally weak at fast pivots.
The direction of travel
Japan’s government has set an explicit target of roughly ¥40 trillion in annual domestic semiconductor sales by 2040, up from approximately ¥8 trillion currently, extending a previously stated ¥15 trillion target for 2030. Whether those numbers are achievable is debatable. What is not debatable is that they signal sustained public capital deployment into this sector, which changes the calculus for anyone considering a Japanese footprint.
The strategic question for a foreign company is not “is Japan rising or declining.” It is “which layer am I dealing with, and what does that layer’s structure do to my leverage.”
Related: How Japanese Semiconductor Companies Really Work · [Japan’s National Semiconductor Strategy] · [Mapping Japan’s Clusters]