Skip to main content

Envisioning is an emerging technology research institute and advisory.

LinkedInInstagramGitHub

2011 — 2026

research
  • Reports
  • Newsletter
  • Methodology
  • Origins
  • My Collection
services
  • Research Sessions
  • Signals Workspace
  • Bespoke Projects
  • Use Cases
  • Signal Scanfree
  • Readinessfree
impact
  • ANBIMAFuture of Brazilian Capital Markets
  • IEEECharting the Energy Transition
  • Horizon 2045Future of Human and Planetary Security
  • WKOTechnology Scanning for Austria
audiences
  • Innovation
  • Strategy
  • Consultants
  • Foresight
  • Associations
  • Governments
resources
  • Pricing
  • Partners
  • How We Work
  • Data Visualization
  • Multi-Model Method
  • FAQ
  • Security & Privacy
about
  • Manifesto
  • Community
  • Events
  • Support
  • Contact
  • Login
ResearchServicesPricingPartnersAbout
ResearchServicesPricingPartnersAbout
  1. Home
  2. Research
  3. Substrate
  4. LightGen Photonic AI Chip

LightGen Photonic AI Chip

A Chinese all-optical chip with 2 million photonic 'neurons' runs 100x faster than NVIDIA's top GPU on image generation — using light instead of electricity
Back to SubstrateView interactive version

Researchers at Shanghai Jiao Tong and Tsinghua universities built LightGen, a chip that processes AI workloads using photons rather than electrons. With over 2 million integrated photonic neurons, it generates high-resolution images and 3D scenes at 100x the speed and energy efficiency of NVIDIA's flagship GPU.

The strategic significance is sanctions-proofing. Photonic chips don't use the transistor-based fabrication that US export controls target. They can be manufactured with different equipment and different materials, potentially rendering the entire silicon chip embargo irrelevant for certain AI workloads.

The constraint is generality. LightGen excels at specific tasks (image/video generation) where parallel photonic processing has natural advantages. It's not a general-purpose replacement for electronic GPUs — yet. But if photonic architectures can be extended to training and inference broadly, China could leapfrog the silicon paradigm entirely.

TRL
5/9Validated
Impact
3/5
Investment
4/5
Category
Hardware

Book a research session

Bring this signal into a focused decision sprint with analyst-led framing and synthesis.
Research Sessions