ASML Unveils Hyper NA EUV Technology, Enabling 5nm Patterning in a Single Exposure
nashnova research
ASML and Zeiss have jointly disclosed Hyper NA EUV, a next-generation lithography system that prints 5 nm features in a single exposure — over a third finer than today's High-NA EUV. This means chipmakers could slash redundant patterning steps and cut both cost and process complexity.
What exactly is Hyper NA EUV?
Engineers from ASML and Zeiss published a joint paper formally disclosing Hyper NA EUV — the next step in extreme-ultraviolet lithography, the technology that etches circuits onto chips with very short-wavelength light.
The headline number: single-exposure resolution down to 5 nm, more than a third finer than what today's High-NA EUV can achieve.
This means → patterns that currently require multiple rounds of exposure could be printed in one shot — fewer steps, lower cost.
Why do chipmakers care so much about fewer exposures?
Fabs like TSMC currently use 0.33 NA EUV systems at the 7 nm and 5 nm nodes, relying on multi-patterning — exposing the same layer several times — to hit smaller pitches.
Each extra exposure adds another round of alignment, inspection, and rework risk. Cost and complexity multiply with every pass.
In plain terms = Hyper NA EUV's value is not just "finer lines" — it is "one cut, done," eliminating redundant steps entirely.
From 0.33 to 0.55 to 0.75 — what is the roadmap?
ASML's current EUV lineup spans two generations: the 0.33 NA NXE series (7 nm to 2 nm) and the 0.55 NA EXE series High-NA EUV (2 nm and below, now in early adoption).
Intel has already deployed High-NA EUV on a commercial line; Samsung and SK Hynix plan to follow by 2028, with TSMC targeting 2030.
Hyper NA EUV is expected to push the numerical aperture — the optical parameter that determines how fine the system can resolve — to 0.75, boosting resolution while keeping the same 13.5 nm wavelength.
How big is the upgrade burden for fabs?
The paper states that the 0.75 NA optical module is expected to be only slightly larger than today's High-NA EUV scanner, compatible with existing cleanroom layouts.
Existing reticles and light-source systems can largely be reused — no wholesale retooling of the fab ecosystem is required.
This means → the upgrade threshold is lower than expected — more like swapping a lens than rebuilding a factory.
What is the biggest technical hurdle?
A higher numerical aperture sharply compresses depth of focus (DOF) — the vertical range within which the image stays sharp — demanding thinner photoresist, tighter focus control, and ultra-flat wafer stages.
In plain terms = the higher the NA, the thinner the "in-focus slice." Think of a magnifying glass: the stronger the magnification, the slightest shake blurs the image.
The paper explicitly flags multiple unresolved technical obstacles, and ASML has not committed to a commercial production timeline.
When could it arrive — and what does that mean for the industry?
Hyper NA EUV is expected to emerge within roughly a decade, targeting industry needs beyond 2035; until then, 0.55 NA High-NA EUV remains the workhorse for advanced nodes.
The paper also notes that even if future tools switch from today's laser-produced plasma (LPP) source to a free-electron laser (FEL) — a more powerful light source driven by a particle accelerator — the core illumination optics can stay the same. This provides a technical reference point for startups like xLight developing FEL sources.
This reflects ASML's strategy of locking in the next decade's roadmap through early academic disclosure. Whether Hyper NA EUV reaches volume production on schedule after 2035 will be the definitive test of this roadmap's viability.
市场有风险，内容仅供研究参考，不构成投资建议。