Gunjo · Business Intelligence for the AI Era
← Sticker Wall JOURNEY · DETAIL

DISCO: From Hiroshima Grinding Stone Factory to Hidden Monopoly Hegemon of Chip Dicing and Grinding

Founded: Mitsuo Sekiya · DISCO Corporation

JOURNEY

Key Fields

FIELD STAMPS
IndustryAI / LLM
RegionJapan
ScaleGiant
ChannelOther

Origin

In 1937, Mitsuo Sekiya founded Dai-ichi Seitosho in Kure, Hiroshima Prefecture, initially producing industrial grinding stones for naval arsenals. After the war, military demand vanished, forcing the company to pivot to the civilian market. Through its survival struggles, it established the technical pathway of 'grinding stones to ultra-thin limits.' In 1956, it launched ultra-thin resin bond wheels for cutting fountain pen nibs to validate the process. In 1968, it released a diamond ultra-thin dicing blade only 40 micrometers thick, stepping into the realm of semiconductor precision processing and embarking on its transformation from a local grinding stone factory into a global monopolist.

Milestones

1937
Inception Turning Point
Mitsuo Sekiya founded Dai-ichi Seitosho in Kure, Hiroshima Prefecture, initially manufacturing industrial grinding stones for naval arsenals. Postwar military orders dropped to zero, forcing a pivot to civilian markets. While searching for new customers, the company established the technical direction of 'making grinding stones extremely thin,' laying the groundwork for its future entry into semiconductor precision processing.
1968
Tech Breakthrough Inflection Point
The company launched Microncut, an ultra-thin dicing blade containing diamond powder with a thickness of only 40 micrometers, pushing grinding stone technology into semiconductor wafer dicing applications. Building on the 1956 development of ultra-thin resin bond wheels for fountain pen nibs—which validated the feasibility of extreme thinning processes—this completed the technological leap from consumer goods to semiconductors, marking 1968 as the official starting point of its entry into the semiconductor industry.
1969
Export Failure Failure
In 1969, the company attempted to directly export high-precision ultra-thin dicing stones to the United States, encountering major setbacks: the U.S. market lacked matching high-precision dicing equipment capable of handling ultra-thin stones, causing them to break frequently during use and resulting in massive financial losses. This failure forced management to abandon the 'consumables-only' model and seriously consider how to integrate tools and hardware systems.
1975
In-house Equipment PMF
In 1975, because the market lacked dicing equipment compatible with its ultra-thin stones, the company independently developed its first semiconductor dicing saw series, the DAD-2H, transitioning from a pure consumables vendor to an integrated 'equipment + consumables + process' manufacturer. In 1974, the company had also cut lunar rock samples brought back by the Apollo program for the University of Tokyo, earning widespread external recognition for its cutting precision and technical reputation.
1999
Renaming & Listing Turning Point
In 1977, the company officially changed its name to DISCO Corporation (derived from Diamond Industrial Saw Company). In 1999, it was listed on the First Section of the Tokyo Stock Exchange (ticker: 6146) and joined the Nikkei 225 index. Following its listing, supported by capital markets, it continuously stepped up R&D and international deployment. As early as 1998, it established DISCO Hi-Tec (China) Co., Ltd. in Shanghai, deeply cultivating the Chinese market for over twenty years.
2024
Mass Expansion Growth
In January 2024, DISCO announced an investment of over 40 billion yen (approx. $276 million) to build a new factory in Hiroshima Prefecture, planning a total of three new factories by 2035 to increase overall production capacity by 14 times. According to company disclosures, replacement parts already accounted for 20% of its total revenue at the time, with the 'equipment + consumables' repeat-purchase model continuing to generate high profits, targeting the demand boom driven by AI and advanced packaging.
2025
Earnings Peak Growth
For FY2025 ending March 2026, DISCO achieved revenue of approximately 436.9 billion yen and net profit of about 135.5 billion yen, hitting record highs for the sixth consecutive year. Overseas sales accounted for 87.6%, gross margin was around 65.5%, and net profit margin was about 34%. Driven by HBM and Chiplet, the company turned 'dicing, grinding, and polishing' into an overwhelming cash-flow business, pushing market capitalization and profitability to historic highs.

Turning Points

  • The 1969 U.S. export failure made management realize that grinding stones and dicing systems must be integrated, directly giving rise to the in-house equipment development strategy.
  • The 1968 launch of the 40-micrometer diamond dicing blade Microncut transitioned grinding stone technology from civil use to semiconductor backend processing, defining the company's core business for the next six decades.
  • The 1975 debut of the first self-developed dicing saw, the DAD-2H, transformed the company from a pure consumables vendor into an integrated 'equipment + consumables + process' manufacturer.
  • The 1999 TSE First Section listing provided capital ammunition for global expansion and continuous R&D investment.

Failures & Pitfalls

  • The 1969 failure of direct exports to the United States: the local market lacked matching dicing equipment capable of handling ultra-thin stones, leading to frequent breakage and massive financial losses.
  • Postwar military orders dropped to zero, causing the grinding stone factory to temporarily lose its core customers and forcing a complete pivot to the civilian market to survive.
  • The early pure consumables model was unsustainable: single sales of grinding stones offered limited gross margins, and the company could not control quality during use, compelling the in-house development of dicing equipment.

关键成功要素

  • 'Equipment + Consumables' deep bundling: high-margin dicing saws lock in high-frequency repeat purchases of specialized blades and grinding wheels, with replacement parts steadily contributing about 20% of revenue.
  • Extreme thinning technology moat: 40-micrometer diamond blades and sub-50-micrometer thin wafer processing capabilities make it difficult for latecomers to replicate.
  • Focus on dicing, grinding, and polishing without branching out: product boundaries remain strictly locked on three backend steps, achieving a global monopoly on single-point processes.
  • Long-term process trust and equipment qualification lock-in: semiconductor fab qualification cycles are long; once its equipment and consumables combination is adopted, deep integration is formed.
  • Forward-looking capacity deployment: investing 40 billion yen in capacity expansion starting in 2024 to meet AI and HBM demands, planning a 14-fold capacity increase by 2035.

Lessons

  • If a consumables manufacturer stakes its fate on someone else's equipment, it will be punished by the market whenever matching parts are missing; tools and main units must be integrated.
  • The path of a hidden champion is extreme single-item focus: making grinding stones 40 micrometers thin builds a monopoly far better than spreading out product lines.
  • Failure is the best strategic mentor: the 1969 U.S. export loss directly defined the 'equipment + consumables' business model for the subsequent fifty years.
  • Although backend processes seem unassuming, they are the final gate for realizing wafer value. The more monopolistic the link, the higher the profit, with gross margins reaching over 65%.

Core Data

  • FY2025 Revenue：436.9 billion yen (Company disclosed figures as of 2026, independent review unverified)
  • FY2025 Net Profit：135.5 billion yen (Company disclosed figures as of 2026, independent review unverified)
  • Gross Margin：65.5% (Company disclosed figures as of 2026, independent review unverified)
  • Net Profit Margin：34% (Company disclosed figures as of 2026, independent review unverified)
  • Global Dicing Saw Market Share：70% to 80% (Company disclosed figures as of 2026, independent review unverified)
  • Global Grinder/Thinner Market Share：60% to 70% (Company disclosed figures as of 2026, independent review unverified)
  • Overseas Sales Proportion：87.6% (Company disclosed figures as of 2026, independent review unverified)
  • Replacement Parts Revenue Proportion：20% (Company disclosed figures as of 2026, independent review unverified)
  • Hiroshima New Plant Investment：40 billion yen (Company disclosed figures as of 2026, independent review unverified)
  • 2021 Global Dicing Saw Market Size：$2.0 billion (Company disclosed figures as of 2026, independent review unverified)

Competitors / Peers

DISCO's primary competitor is fellow Japanese enterprise Tokyo Seimitsu, with both companies long dominating the global dicing saw and wafer thinning equipment markets. In addition, major equipment makers like Tokyo Electron compete indirectly in adjacent segments. In recent years, China's domestic substitution camp—including CETC 45, Power Automation, and Huazhuo Jingke—has also been accelerating catch-up efforts in dicing, thinning equipment, and high-end grinding wheel consumables. However, relying on its closed-loop of equipment, consumables, and processes, alongside a global dicing saw market share of over 70% and a gross margin exceeding 65%, DISCO's moat remains extremely deep, making it difficult to shake the market landscape in the short term.

Sources

Related

Same industry / scale
AI / LLMJapan

Kioxia: The Turnaround Journey from Toshiba Memory Spin-off to NAND Giant

AI / LLMJapan

Tokyo Electron: The Hidden Champion with a 90% Monopoly in Coater/Developer Equipment

AI / LLMGlobal

Arm Holdings: Chip IP licensing model covers global mobile chips, now penetrating server and automotive semiconductors

AI / LLMUS

Figma's Rise from Browser-Based Editor to Adobe Acquisition Block

⚠️ This entry draws on self-reported accounts and public reporting, not independently verified, and is not investment, startup, or legal advice; revenue figures are case-specific and not representative of typical results.