Russia’s EUV Roadmap to 2037 — Ambition vs Reality in Litho Sovereignty

Russia’s researchers sketch out a plan to replace DUV with EUV by 2037. But the technical hurdles and geopolitical clampdowns make the timeline more aspirational than executable.

What’s in the roadmap

The Institute of Microstructure Physics published a multi-stage lithography plan spanning 2026 to 2036. The goals: build first a 40 nm capable tool (2026–2028), then a 28 nm to 14 nm mid-tier system (2029–2032), and finally a sub-10 nm production unit using six mirrors (2033–2036). The target wavelength is 11.2 nm, using hybrid solid-state lasers and xenon plasma instead of tin droplet sources. (From the Tom’s Hardware article) :contentReference[oaicite:0]{index=0}

Why this is wildly optimistic

• Tooling complexity: Each jump in mirror count, overlay precision, vibration control, and mask stability is a steep slope. Even established firms struggle.
• Materials and optics: The mirrors (Ru/Be) and reflective coatings for 11.2 nm are not commodity — polishing, cooling, and alignment are all huge risks.
• Supply chain and IP blockade: Western sanctions bar access to critical components, metrology equipment, upgrades, and training.
• Throughput expectations: The proposal envisions 5 wafers/hour at first, scaling to 100 wafers/hour later — in practice, real commercial EUV tools must achieve much higher yield and throughput to amortize cost.

Strategic angle and possible value

Even if the project falls short, it signals Russia’s push for tech sovereignty. It moves R&D on alternative lithography forward, forces conversation around export control reform, and gives political cover. On the ground, though, small fabs may adopt parts of the first-stage system, but full node parity with ASML remains distant.

Bottom line

Russia’s EUV roadmap is provocative and symbolic — a long-term bet, not a near-term threat. The gap between ambition and hardware execution is wide. But in countries under tech blockade, narrative and direction matter. I’ll watch whether any of the early sub-10 nm prototype efforts survive to 2030; most will fold before then.

Source: Tom’s Hardware