← ALL FRONTIERS
// FRONTIER.AFTER-THE-CHIP

After the Chip

Post-Silicon Computing (Photonic, Neuromorphic, DNA)

4 companies · 2 sourced backers · 26 milestones on record

// What this frontier is about

This frontier explores alternatives to conventional computing hardware, including thermodynamic computing and optical data movement. The companies aim to reduce the energy needed for AI computing and make light-based connections standard in data-center hardware, with thermodynamic chips still at early stages and optical systems targeting commercial deployment at scale.

// How real is it?

Where the 4 companies stand today, from their profiles and 10-year milestones. Tap a phase to see who is there.

← labavailable to all →
// What usually holds it back
  • ▸Energy limits on scaling AI computing
  • ▸Moving thermodynamic hardware from early demonstrations to finished products
  • ▸Scaling optical interconnect production for major systems
  • ▸Integrating optical data movement into standard AI hardware
// Who's backing it
AMD Ventures· Celestial AISamsung Catalyst Fund· Normal Computing

Publicly listed: 1 company (owned by public shareholders).

// Inside the terminal
  • ›Every sourced story on all frontiers, filterable by company and news type
  • ›10-year milestone records for each company: demo vs shipped
  • ›Explained: plain-words readiness diagrams for real breakthroughs
  • ›Signal Trends of 39 frontiers covering 378 companies over time.
  • ›The events calendar: launches, approvals and deadlines ahead
Create a free account
// Companies in this frontier
  • Ayar Labs

    To make in-package optical interconnects a standard building block of AI data-center hardware, shipping in volume in major systems.

    Trials · publicly listed

  • Celestial AI

    To make optical (light-based) data movement a standard part of AI computing hardware, deployed commercially at scale.

    Some users · backed by AMD Ventures

  • Extropic

    To make AI computing dramatically more energy-efficient by commercializing thermodynamic hardware, so that energy stops being the limiting factor in scaling AI.

    Lab

  • Normal Computing

    To reach multiple-orders-of-magnitude (targeting ~1000x) energy-efficiency gains for AI data-center silicon by scaling thermodynamic computing.

    Trials · backed by Samsung Catalyst Fund

// WEEKLY.UPDATES

Keep following breakthroughs like these. Every Monday, get up to 5 major breakthroughs and launches explained in plain English.

Free. No account needed. One email a week. Unsubscribe anytime.

9 NOV 2026

Ayar Labs Wins Demo of the Year Award at 2026 TSMC North America Technology Symposium

14 OCT 2026

Wiwynn unveils CPO compute system with GUC and Ayar Labs

1 OCT 2026

First Sparks of Thermodynamic Recursive Intelligence

21 SEP 2026

Mark Wade: The $50 Billion Optical Bet That Must Precede AI's Revenue Boom — BigGo Finance

12 AUG 2026

Cerebras Q2 2026: core revenue $209.9M (+103% YoY), cloud nearly quadruples, raises 2026 core outlook to $880-890M