Tech

Hybrid light-matter quasiparticles promise faster, cooler AI chips

May 29, 2026 · From the archive, written in the Professor's voice before our October 2026 redesign from sciencedaily.com

What happened

Penn researchers coupled photons with excitons in a 2-D perovskite microcavity to form polaritons that perform matrix multiplies at the speed of light. Their prototype executes a 1024-by-1024 multiply in 120 femtoseconds while drawing 40 femtojoules per operation, two orders of magnitude below electronic SRAM. The device is fabricated with standard lithography on a silicon substrate.

Why it matters

You begin evaluating workloads for optical versus electronic paths instead of assuming all computation stays on transistors. This reframes model architecture choices around latency and heat budgets rather than FLOPs alone.

Who's doing it

Professor Ritesh Agarwal's lab at the University of Pennsylvania ran a 4-by-4 polariton array for 72 hours continuous inference at 1.2 mW total system power, publishing the trace data and mask layouts under an open-source hardware license.

Try it

1. Download the open PDK and simulation files from penn-agarwal-lab.github.io/polariton-pdk.
2. Use the Cadence Virtuoso template to instantiate a polariton coupler sized for your target matrix dimensions.
3. Run the Ansys Lumerical script to extract energy per MAC and confirm sub-100 femtojoule operation before taping out.

Read the original at sciencedaily.com