A Chinese startup has demonstrated a prototype extreme ultraviolet (EUV) light source producing 15 watts of output power, built with roughly 90% domestically sourced components, Tom's Hardware reports. The demo is real. The catch: the device remains years away from viability as a lithography-grade source.

Fifteen watts sounds like a number. In context, it is a starting point. A source at that power level can prove the physics and the engineering chain — the plasma generation, the debris mitigation, the optics — but it cannot yet drive a production scanner through the wafer throughputs that leading-edge fabs demand. The company itself frames the work as early-stage, with commercial viability still years out, according to the report.

Why does a light source matter so much?

EUV lithography runs on 13.5-nanometer light, generated today by firing a laser at tin droplets tens of thousands of times per second. The light source is the beating heart of an EUV scanner. It is also one of the hardest subsystems to build, and the reason only one supplier — ASML, with Cymer's laser-produced-plasma technology — ships production EUV tools at scale.

So when a startup anywhere demonstrates a working source, the industry pays attention. When that startup is Chinese, and when 90% of the components come from domestic suppliers, the demonstration carries an extra layer of meaning.

How much of the stack is actually domestic?

The reported figure — around 90% domestic components — is the standout claim. It signals that China's supply chain for precision optics, power electronics, and control systems has matured enough to assemble a credible EUV source prototype without leaning heavily on imported subsystems.

That matters commercially. Export controls have tightened access to advanced lithography equipment for Chinese fabs. A domestically sourced EUV light source is a step toward reducing that dependency, even if the full toolchain — source, optics, scanner, photoresists — still has gaps.

A prototype is not a product. The gap between 15 watts in a lab and a source qualified for high-volume manufacturing spans reliability, power scaling, and cost. The startup has cleared the first hurdle. The remaining ones are taller.

What changes, and what doesn't?

For now, the competitive picture doesn't shift. ASML remains the sole merchant supplier of production EUV scanners, and nothing in this demonstration alters near-term capacity or pricing for leading-edge lithography tools.

What it does change is the trajectory. Each demonstrated subsystem — a source here, optics there — shortens the distance between China's current capabilities and an indigenous EUV platform. The startup's own timeline, per the report, puts viable deployment years away, which keeps any commercial impact comfortably beyond the current investment cycle.

Still, the direction is clear. A working prototype with near-total domestic content is a proof point that the barrier to entry, while enormous, is no longer theoretical. Watch the power figure: if future iterations climb from 15 watts toward the hundreds of watts that production lithography requires, the conversation moves from curiosity to competition.