Atomic Beam Lithography Advances with Lace Lithography’s €34.5M Raise Challenging EUV Dominance

Atomic beams enter the lithography race as Norway’s Lace raises €34.5 million

Lace Lithography, a Bergen-founded semiconductor equipment startup, has secured €34.5 million to advance a chip patterning platform that replaces photons with beams of neutral atoms. The company’s approach targets smaller features at lower energy than light-based systems and is positioned as an alternative path for extending Moore’s Law. The round includes deep-tech investors and Microsoft’s venture arm M12, underscoring strategic interest in diversifying the world’s lithography supply chain and reducing single-point dependencies at the heart of advanced chip production.

Lace originated from European research programs and is expanding across Norway, the Netherlands, Spain, and the U.K. Its development roadmap sits alongside Europe’s concerted push-via instruments such as the EU Chips Act-to reinforce sovereign capability in critical chipmaking tools and to ensure that next-generation patterning capacity is not exclusively controlled from a single geography.

How atom-based patterning works

Lace’s platform develops Beyond‑EUV (BEUV) lithography that forms patterns with neutral atoms rather than extreme‑ultraviolet light. The technology is being matured through the EU‑funded FabouLACE project, which focuses on metastable atom beams and dispersion‑force masks.

• Source and beam: a high‑brightness atom source and beamline form a finely collimated neutral‑atom beam under ultra‑high vacuum.
• Pattern transfer: nanostructured, dispersion‑force‑based masks modulate the atom beam to write features on the resist-coated wafer.
• Resolution outlook:
  • Demonstrated lab features: ~26 nm (project baseline).
  • Project target: ~16 nm by 2028 under FabouLACE.
  • Longer‑term objective: enabling ~2 nm features with future system iterations.
• Potential advantages to validate: reduced exposure energy, alternative materials stack, and different stochastic behavior than photon‑based resists.

Company materials emphasize industrialization beyond the research stage, with attention to beam uniformity, resist chemistry, overlay control, mask durability, contamination management, and step‑and‑scan mechanics. For policymakers and procurement teams in charge of national fabs, the key question is whether these lab demonstrations can be translated into stable, repeatable processes at 300‑mm wafer scale within the reliability envelopes expected of front-end production tools.

Rival bets: new EUV light sources aim to upgrade ASML scanners

In parallel, U.S. startup xLight has raised US$40 million to build a free‑electron‑laser (FEL) light source intended to slot into existing EUV scanners. The effort speaks to a broader wave of U.S. and European attempts to boost EUV throughput and reliability without replacing the entire photolithography stack, potentially offering fabs a way to meet foundry roadmaps and national incentive targets with incremental upgrades rather than all‑new platforms.

As xLight’s chief executive put it: “This is the most expensive tool in the fab. It’s what drives the cost of the wafer more than any other tool in the fab, and it’s what drives capacity more than any other tool in the fab.”

Pat Gelsinger, who chairs xLight’s board, framed the strategic backdrop bluntly: “There was a terrible mistake made giving Cymer the ability to become a European-owned and controlled company.” He added: “We can build that here, or it can be built elsewhere. China is investing heavily in this space. There’s an extraordinary backstory here that says, ‘Let’s get this one right.’” The comments echo a growing view in Washington and Brussels that light sources, as much as scanners themselves, are now a core element of industrial and security policy.

ASML’s position is entrenched-but the edges are being tested

ASML remains the sole provider of EUV scanners at leading nodes and is ramping High‑NA EUV for sub‑2‑nm patterning. EUV’s maturity, ecosystem depth, and proven high‑volume manufacturing throughput are formidable moats. Yet the cost, power demand, and single‑vendor concentration of EUV keep the door open for credible alternatives or complementary tools-whether atom beams that change the patterning physics or upgraded EUV sources that improve availability and cost‑per‑wafer without upending fab workflows.

This is increasingly a governance question as much as a technology one. Export‑control regimes, subsidy programs, and state-backed capacity plans all assume that enough advanced patterning tools will be available on commercial timelines. Any disruption-technical, geopolitical, or supply‑chain driven-would ripple through national AI strategies, cloud build‑outs, and industrial policy commitments.

Feature and maturity comparison across next‑generation patterning paths

For corporate strategy teams and public agencies allocating chip subsidies, this matrix is less about picking a winner today and more about understanding which paths are likely to be complementary, which could be lock‑in risks, and where early de‑risking capital might have the highest leverage.

System design checkpoints that will make or break atom‑beam lithography

• Beam physics and stability
  • Uniform atom flux over large fields without decoherence or scattering losses.
  • Dynamic beam shaping for step‑and‑scan exposure.
• Mask technology
  • Durable dispersion‑force masks with low defectivity and manufacturable CD control.
  • Pellicle‑equivalent contamination management in vacuum.
• Resist and materials stack
  • Chemistries with predictable line‑edge roughness and sensitivity to atomic exposure.
  • Etch selectivity compatible with advanced interconnect and device layers.
• Overlay and metrology
  • Sub‑nanometer overlay with existing alignment schemes or new atom‑compatible fiducials.
  • In‑situ monitoring to control stochastic variation.
• Tool availability and fab fit
  • Mean‑time‑between‑service and uptime comparable to EUV toolrooms.
  • Integration with automation, FOUP handling, and vacuum cluster tools.

These checkpoints are also where early public‑private partnerships will likely concentrate: shared pilot lines, metrology benchmarks, and procurement pre‑commitments can all shorten the time between a physics demonstration and a tool that regulators and institutional buyers are prepared to qualify for mission‑critical production.

Regulatory and industrial context shaping the next nodes

• Strategic funding
  • EU research and innovation programs backing atom‑beam patterning and allied process technologies.
  • U.S. incentives aimed at upgrading EUV sources and onshoring critical lithography sub‑systems under national semiconductor initiatives.
• Export controls and concentration risk
  • Restrictions on advanced lithography sales to China elevate single‑vendor exposure for allied economies.
  • Diversifying core subsystems (light sources, patterning methods) is emerging as an explicit policy goal, informing how governments evaluate and sometimes co‑finance tools like atom‑beam platforms or FEL‑based sources.
• Standards and interoperability
  • SEMI equipment and wafer‑handling standards will pressure novel platforms to prove compatibility-or to justify new standards.
  • Mask infrastructure, inspection, and pellicle analogs must reach EUV‑class reliability metrics before institutional buyers fold them into long‑term capacity plans.

For regulators and export‑control authorities, the arrival of credible non‑EUV patterning routes will complicate permit decisions: restrictions that today hinge largely on ASML’s tools may need to be reframed around capabilities-resolution, overlay, throughput-rather than specific vendors.

Enterprise impact: AI capacity, cost per wafer, and energy

• AI compute supply: Any technology that increases patterning throughput or shrinks features without exploding cost directly expands available GPU/AI accelerator supply.
• Cost per wafer: EUV source upgrades promise better tool utilization; atom beams promise lower exposure energy and potentially simpler optics-both attack cost drivers from different angles.
• Energy profile: Reducing exposure energy and rework rates can moderate data‑center supply chain emissions tied to chip manufacturing.

Boards signing off on multi‑billion‑dollar AI infrastructure programs are therefore indirectly betting on how quickly these lithography innovations mature. The further EUV cost‑per‑wafer can be pushed down-or the sooner a viable alternative appears-the more headroom there is for governments and enterprises to sustain aggressive AI roll‑out plans without running into capital or energy constraints.

Signals to watch in 2026

• Independent metrology of sub‑20‑nm atom‑beam features with credible line‑edge roughness and defectivity data.
• Demonstrations of large‑field uniformity and multi‑die overlay on 300‑mm wafers.
• EUV source pilots showing sustained uptime and dose stability improvements inside production scanners.
• Early joint development agreements with top foundries or IDMs for pilot layers.
• Mask ecosystem progress: manufacturable dispersion‑force masks plus inspection/repair tooling.

For institutional investors, export‑control bodies, and national chip offices, these are the milestones that separate a compelling lab story from a technology that must be modeled into long‑term capacity forecasts and regulatory frameworks.

The takeaway

EUV is not standing still, but neither is the competition. Lace’s atom‑beam program offers a fundamentally different physics path that could sidestep EUV’s cost and energy ceiling if key engineering hurdles fall. FEL‑driven EUV sources seek a faster win by upgrading the heartbeat of today’s scanners. Both efforts test the long‑term edges of ASML’s dominance-and, if they deliver, could widen the bottleneck that constrains global AI capacity. The next two to three years will show whether neutral atoms and new light sources remain policy‑driven hedges, or graduate into tools that ministries, regulators, and corporate boards must plan around as core infrastructure.