Inversion Semiconductor

What do they actually do

Inversion Semiconductor is an early-stage R&D company developing tabletop, particle‑accelerator–based light sources intended to power future chip lithography. Today, they run high‑power laser experiments and laser‑wakefield accelerator prototypes in a small lab, measuring and stabilizing short‑wavelength (EUV/soft x‑ray) light. They have no commercial tool in fabs yet; current work is internal experiments and joint tests with research partners (YC profile, company site, EE News).

They’re collaborating with Lawrence Berkeley National Lab’s BELLA Center (BELLA‑LUX) to stabilize and test their prototypes, and are working toward a tunable high‑power source (“STARLIGHT”) plus a demo scanner (“LITH‑0”) to prove silicon patterning. Near‑term milestones they’ve stated include achieving about 1 kW of soft x‑ray light in the ~20–6 nm range and then choosing a commercial route (sell sources, build scanners, or offer a beam‑time facility) (YC profile, Tom’s Hardware, EE News).

Who are their target customer(s)

• Large-volume logic and memory fabs (e.g., TSMC, Samsung, Intel): They need higher wafer throughput and finer patterning but face throughput, uptime, and cost limits with today’s EUV sources. Any new source must run 24/7, integrate cleanly with existing scanners/processes, and be serviceable in production (YC, Tom’s Hardware).
• Lithography equipment makers (ASML, Nikon, Applied Materials): They need a proven, stable, industry‑grade light source with a clear integration and controls path; committing to redesigns or new subsystems requires long co‑engineering cycles and rigorous optics/control qualification (EE News).
• Materials and resist suppliers: They require predictable wavelength, power, and stability to develop and qualify resists, pellicles, and optics; shifting spectra or instability can waste development cycles and delay qualifications (Tom’s Hardware).
• National labs and academic groups: They want compact, tunable EUV/soft‑x‑ray sources but rely on large facilities with limited access. A tabletop source must be stable and easy to operate for routine experiments (e.g., BELLA‑LUX collaboration) (EE News).
• Fab operations, maintenance, and reliability teams: They need energy‑efficient, maintainable tools. High‑power lasers/accelerators add power, cooling, uptime, and service complexity that must be engineered away before factory deployment (Tom’s Hardware).

How would they acquire their first 10, 50, and 100 customers

• First 10: Run subsidized beam‑time and joint pilots with national labs, university groups, and 1–2 equipment‑maker partners (using STARLIGHT + LITH‑0) to produce reproducible patterning data and public technical reports; embed Inversion engineers on‑site for hands‑on support (YC, EE News).
• First 50: Convert validated pilots into paid integration contracts and lease/demo units for pilot fabs and materials vendors; offer a centralized beam‑time facility for those unwilling to install prototypes and package standardized integration kits and training (YC).
• First 100: Transition to commercial source offers (leased or sold units with SLAs and certified integration) and secure strategic co‑development with scanner OEMs and major fabs, tied to proven uptime/throughput and resist qualification from earlier pilots (YC, Tom’s Hardware).

What is the rough total addressable market

Top-down context:

Near term, the realistic target is the EUV light‑source subsegment of the overall EUV lithography market. Industry estimates peg EUV lithography at roughly $12B in 2024 and growing toward the low‑$20Bs by decade’s end, with light sources a major component (MarketsandMarkets, Yahoo Finance press release).

Bottom-up calculation:

Start with ~$12.2B EUV lithography revenue (2024) and apply an estimated ~40–46% light‑source share from industry breakdowns, implying an EUV light‑source TAM of roughly $4.8–$5.6B/year today; the broader lithography equipment market itself is about $27.7B in 2024 (MarketsandMarkets, Mordor Intelligence, Fortune Business Insights).

Assumptions:

• EUV market estimates and component shares are used as proxies and vary by source; we use a conservative ~40–46% light‑source share for illustration.
• Initial TAM focuses on EUV light sources; it excludes scanners and services unless Inversion later sells a scanner or beam‑time at scale.
• Fab adoption depends on multi‑year qualification and capex cycles; market sizes do not imply near‑term revenue capture (WFE context: equipment billings were ~$117B in 2024) (SEMI).

Who are some of their notable competitors

• ASML / Cymer: Incumbent supplier of production EUV scanners and integrated LPP light sources (CO2 laser + tin droplets). Any replacement must match their uptime, integration path, and installed‑base relationships.
• xLight: Startup pursuing accelerator/FEL‑style EUV sources aimed at replacing laser‑produced plasma in semiconductor manufacturing—directly comparable long‑term positioning to Inversion.
• Energetiq Technologies: Provider of compact EUV/soft x‑ray and related light sources for R&D, metrology, and inspection; competes for early research users who need stable EUV light now.
• Lyncean Technologies: Commercializes compact accelerator‑based light sources (CLS) for industrial and research users; a credible alternative for customers seeking compact x‑ray/EUV capabilities.
• KMLabs: Supplies tabletop EUV/soft x‑ray sources using high‑harmonic generation for labs and applied research—competes for research and prototyping budgets before fab‑grade tools exist.