Theses›Semiconductor manufacturing›Photonic multimaterial foundry

Photonic multimaterial foundry

A photonic foundry built ground-up for multi-material integration beats the silicon-first lines that bolt heterogeneity on

  • Bear
  • Early
  • $15B by 2035

The callA foundry built for several photonic materials on one flow could be the TSMC of photonics, but it costs $1B+ and the incumbents are adding materials to existing lines.

D1D2R1R2C1foragainst D1D2R1R2C1foragainst
Fig. 1The balance. Drivers and catalysts against restraints, each block sized by its weight; catalysts hang dashed because they have not happened yet, sized by their odds. It tips toward the restraints.
  1. D13.2T and 6.4T optical engines need lithium niobate, BTO or indium phosphide modulators next to silicon, because silicon modulators stall near 400G a lane.
  2. D2CMOS-first photonic lines struggle to host hot, exotic recipes, so today's engines are assembled from parts made in different fabs.
  3. R1A ground-up fab is $1B+ and $50-200M per process node, with no seed-sized way in; the only European build is sovereign.
  4. R2imec, Tower and GlobalFoundries are layering heterogeneous integration onto existing lines, which is the street's expected path.
  5. C13.2T and 6.4T engines move modulation off silicon in volume, forcing one qualified multi-material flow rather than a stack of single-material lines plus packaging.2031
Rests onthe beliefs underneath
Supported by · confidence
What would prove itdated tests, open and resolved
202720282029today123watch 202720282029today123watch
Fig. 2Each ring is a dated test: dashed while open, struck through when overdue, ticked or crossed once resolved.
  1. 1A ground-up multi-material team financed ahead of the fab, which would also reopen the OPE conversation.Dec 2027 · open
  2. 2GlobalFoundries or Tower ships an open multi-material design kit co-qualifying III-V, silicon nitride and lithium niobate or BTO: the kill.Dec 2028 · open
  3. 3A design win that only a co-integrated line can serve, or supply chains refusing to design to any new platform.Dec 2028 · open
Expressionslisted, private, technologies
Private · 10 private companies, 3 of them in the US.

X-Celeprint · nanoPHAB · OpenLight Photonics · Intel Foundry Services (Intel IFS / Silicon Photonics) · POET Technologies · Quintessent · EDWATEC · Photon IP · LIGENTEC · CompoundTek

12 more private companies tracked, not shown

Technologies
Who has already bet on itprivate rounds, from public sources

$9M raised by 3 companies, $15M of it in the last three years. Wave building: Money is arriving: recent rounds and tier-1 funds are in, but no runaway leader yet.

  • $9MLIGENTEC
    • Founded 2016
     Accelerator, Sep 2024 (Leading European Tech Scaleups)
  • ·CompoundTek
    • Founded 2017
     Rounds not itemised
  • ·nanoPHAB
    • Founded 2015
     Rounds not itemised
What changedthe record, newest first
  • 10 Sep 2026A founder answered the foundry-fallacy piece with the asset-light version of the argument: put the photonics in the CMOS back end, license the module and the design kit, never build the fab. The physics is simulated, the material's lifetime is an open question, and the pitch is honest about both. It is the first time the "process node in front of the fab" shape has turned up as a company rather than a construction phase.