Hoya Unable to Fully Meet Samsung's 2027 EUV Photomask Substrate Demand
nashnova research
Japan's Hoya has told Samsung it cannot fully meet the chipmaker's additional EUV mask blank requirements for 2027 — with over 90% of Samsung's supply dependent on this single vendor, the gap threatens its advanced-node production timeline.
How much does Samsung need — and why can't Hoya deliver?
Samsung's internal assessment projects annual EUV mask blank demand rising roughly 25% over the next three years; 2026 usage is estimated at about 5,500 blanks.
Hoya responded that demand from other chipmakers is climbing in parallel, making it impossible to exclusively fill Samsung's full request.
This means → the constraint is not Hoya's willingness but industry-wide capacity — Samsung's status as a top customer no longer guarantees supply priority.
Why is Samsung so dependent on a single supplier?
Samsung currently sources over 90% of its EUV mask blanks from Hoya, up from roughly 80% previously.
The other supplier, AGC, lost share after quality issues and specification changes.
Industry estimates put Hoya's global EUV mask blank share at roughly 70%–75%, with AGC at 20%–25%.
In plain terms = only about two and a half companies worldwide can make these blanks, and one of them stumbled on quality — Samsung had almost nowhere else to go.
Can a new factory close the gap in time?
In April 2026 Hoya announced a new EUV mask blank plant in Singapore, investing roughly ¥42 billion (about $263 million) — but production is not scheduled until fiscal 2028 (starting April 2028), with no new capacity in 2027.
CEO Eiichiro Ikeda disclosed that EUV revenue grew at least 25% year-on-year in fiscal Q2 2026, with full-year growth expected to exceed 20%. He added that rising demand may require "more than one new plant" per business line.
This means → demand is accelerating, but supply cannot catch up until 2028 — 2027 is a confirmed capacity gap.
Why does each new process node consume so many more blanks?
EUV mask usage jumps with every generation: Samsung's DRAM 1b node uses 4–5 masks, 1c rises to 6–7, and the 1d node — slated for mass production in H2 2027 — is expected to reach 10.
Foundry numbers are steeper: 4nm requires roughly 15–16 masks, 3nm about 20, and 2nm about 25–27.
Phase-shift masks (PSM) — a higher-precision but slower-to-produce mask type — tighten the bottleneck further: PSM blanks take 1.5 to 2 times as long to manufacture, and Samsung's 1d node is expected to shift most EUV masks to PSM.
In plain terms = every process step forward roughly doubles blank consumption, while the hardest-to-make blanks also tie up production lines longer — the supply–demand gap widens from both ends.
Where does the Korean domestic alternative stand?
Samsung is testing EUV mask blanks from Korea's S&S Tech, with sample qualification on the 4nm foundry line starting in early 2026.
Key thin-film properties have met Samsung's specifications, but particle contamination and other metrics still need improvement; the mass-production qualification target has slipped to early 2027.
S&S Tech's EUV center in Yongin was completed in October 2025; once operational, it is expected to roughly double the company's capacity.
This reflects a real push toward domestic alternatives — but the certification delay means S&S Tech is unlikely to contribute meaningful volume in the first half of 2027.
What does this mean for Samsung's production milestones?
The 1d DRAM node targets mass production in H2 2027; 2nm foundry is in a critical ramp-up phase over the same period.
Hoya's new capacity arrives no earlier than 2028; S&S Tech's qualification is incomplete — there is currently no confirmed solution to cover the 2027 supply gap.
This means → whether EUV mask blanks arrive on time has become the single most critical bottleneck variable for Samsung's next-generation node launches.
市场有风险，内容仅供研究参考，不构成投资建议。