SUMCO: Japan's Other Wafer Giant — How Two Companies Control Most of the World's Silicon

Read this alongside our analysis of wafer leader Shin-Etsu Chemical for the full picture of the duopoly. The broader pattern of Japanese dominance in chip materials is covered in why most chip materials are made in Japan (in Japanese), the industry's division of labor in the six roles of semiconductor companies (in Japanese), and the overall industry flow in how the semiconductor value chain works (in Japanese).

What kind of company is SUMCO?

SUMCO Corporation is a Japanese manufacturer specializing in silicon wafers, the substrate on which semiconductors are built. The name derives from SUMITOMO MITSUBISHI SILICON CORPORATION: the company was formed in 1999 by combining the silicon operations of Sumitomo Metal Industries and Mitsubishi Materials. A Sumitomo-group materials company, it is headquartered in Tokyo with its main production sites in Kyushu, including Imari City in Saga Prefecture.

Where Shin-Etsu is a diversified chemical maker earning from both chemicals and wafers, SUMCO is almost entirely a wafer pure play. That makes it a high-purity proxy for the wafer business itself — its fortunes rise and fall directly with the semiconductor cycle.

Consolidated revenue for the fiscal year ended December 2025 (FY2025) was 409.7 billion yen. SUMCO's name never appears on a finished chip or smartphone, but its products are embedded deep inside as the foundation disc (SUMCO Corporation).

How is SUMCO's revenue split by product and region?

Nearly all of SUMCO's revenue comes from silicon wafers. Products divide mainly by diameter: 300mm (12-inch) wafers for leading-edge logic and memory, and 200mm (8-inch) and smaller for mature applications such as automotive and analog. In recent years 300mm has been the center of revenue and profit, while commodity wafers of 200mm and below are in structural decline.

Results and product trends for FY2025 were as follows.

(Source: SUMCO Corporation)

By region, overseas sales account for the large majority. Demand centers on East Asia, where front-end chipmaking (foundries and memory makers) is concentrated, with Taiwan the largest single market. As of 2025 the regional split is roughly: Taiwan about 37%, Japan about 19%, China about 13%, South Korea about 10%, and the U.S. about 9%. The expansion of TSMC and other Taiwanese foundries keeps pushing Taiwan's weight up (Positen).

Note that in FY2025, although revenue edged up, operating profit shrank to 1.3 billion yen and the bottom line fell to a net loss of 11.8 billion yen. Impairment losses on fixed assets tied to production restructuring and inventory write-downs took their toll. Even with AI demand strong, weak commodity wafers and cost burdens crushed profitability (SUMCO Corporation).

Where does SUMCO's technological strength lie?

SUMCO's technological strength lies in engineering the purity and flatness of large-diameter 300mm wafers at mass-production scale. A silicon wafer is the substrate on which chip circuits are printed — silicon sliced into thin discs and polished to a mirror finish. Quality starts with the precision of crystal pulling: dipping a seed crystal into molten silicon and slowly rotating and lifting it to grow one enormous single-crystal ingot.

The slightest bump or particle on a wafer's surface makes the circuits built on it defective and drags down yield — the ratio of good chips. And the more advanced the node (the circuit generation), the smaller the tolerated defects, so requirements only tighten. That is why SUMCO has spent decades refining this quality in volume production, building a world-class record in 300mm wafers.

The flagship advanced product is the epitaxial wafer — a polished wafer on which an additional thin, high-quality single-crystal layer is chemically grown. With extremely few surface defects, epitaxial wafers are adopted where quality demands are highest, such as leading-edge logic and image sensors. As data-center demand (in Japanese) for AI chips grows, demand for these advanced wafers grows with it.

What role does SUMCO play in the supply chain?

SUMCO sits at the very top of the supply chain, supplying wafers to nearly every foundry and IDM — the integrated device makers that both design and manufacture chips. TSMC, Samsung, Intel and every other major chipmaker must procure wafers before they can build a single circuit, and SUMCO stands alongside Shin-Etsu as one of the two dominant sources.

This is where the stickiness of a consumables business kicks in. Wafers are not a one-time purchase: they are consumed every time chips are made, so orders recur as long as production runs. Chipmakers also spend long qualification periods evaluating a wafer before adopting it, which means that once a supplier is designed in, switching is rare. High switching costs turn into a steady accumulation of stable business.

Because the structure is close to a duopoly, Shin-Etsu and SUMCO are hard to replace for chip fabs everywhere. When demand for advanced data-center chips expands, it ripples all the way up to the wafers underneath.

What is SUMCO's market position and competitive structure?

SUMCO holds roughly 20% of the global silicon wafer market, second to Shin-Etsu Chemical's roughly 40% (around 2024, research-firm estimates). The top five suppliers hold about 80% of this oligopolistic market, and the two Japanese companies alone exceed half (Business Research Insights).

A five-forces view of the market looks like this.

The biggest entry barrier is purity plus proven volume. Only a handful of companies worldwide can stably supply 300mm wafers at leading-edge-node specifications, and without massive capital spending and years of qualification, no customer will adopt a newcomer. Large customers are also often tied in through long-term supply contracts, leaving almost no room for entry. Because the market does not compete primarily on price, the incumbents' positions are hard to dislodge.

What are SUMCO's weaknesses and risks?

The biggest risk is exposure to the semiconductor cycle. As a wafer pure play, SUMCO cannot smooth the waves with a chemicals business the way Shin-Etsu can; chipmakers' capex and inventory swings hit its results almost directly. In FY2025, even with AI-driven 300mm demand strong, weak commodity wafers and cost burdens pushed it to a net loss.

Heavy capital spending is another burden. Expanding 300mm capacity requires investment in the hundreds of billions of yen, and if demand undershoots, depreciation weighs on profits. The impairment losses booked in FY2025 alongside production restructuring illustrate how hard these investment calls are.

Dependence on power and raw-material costs cannot be ignored either. Refining silicon, pulling single crystals and polishing all consume large amounts of electricity, leaving profits exposed to energy prices and decarbonization costs.

What are SUMCO's recent moves?

SUMCO is pressing ahead with 300mm capacity investment centered on AI demand. In 2024 it designated its sites in Imari, Saga Prefecture and elsewhere as the world's largest production base for advanced semiconductor wafers, announcing major investment in Kyushu. With TSMC's arrival in neighboring Kumamoto, the semiconductor materials supply web across Kyushu keeps thickening (Nikkei).

At the same time, it is concentrating its portfolio. SUMCO plans to end 200mm wafer production at its Miyazaki site around the second half of 2026, redirecting resources to high-value 300mm for AI — doubling down on the leading edge while exiting structurally shrinking mature products (Business Research Insights).

On major new fab investment, however, it has signaled caution on construction timing, citing uncertainty in the chip market and tariff policy. SUMCO rarely makes headlines — but the more AI-era chips the world builds, the more quietly indispensable the company holding their foundation becomes. For a memory maker whose demand ultimately flows down to these wafers, see our Kioxia deep dive.

Key takeaways

• SUMCO is a pure-play silicon wafer maker ranked No. 2 worldwide (roughly 20% share), one half of a near-duopoly with Shin-Etsu Chemical that controls more than half the global market.
• Consolidated revenue for the fiscal year ended December 2025 was 409.7 billion yen; AI-driven 300mm grew, but restructuring impairments and weak 200mm-and-below demand pushed it to a net loss.
• Its strengths are the purity and flatness of large-diameter 300mm wafers and mass-production technology for advanced products such as epitaxial wafers.
• It supplies nearly every foundry and IDM from the very top of the supply chain, and because wafers are consumables, customers rarely switch.
• Its risks are exposure to the semiconductor cycle, heavy capital-spending burdens, and dependence on power and raw-material costs.