Abstract: As conventional electronics approach their limits in speed and energy efficiency, photonic systems are emerging as attractive candidates for neuromorphic computing. In particular, platforms operating in the strong light–matter coupling regime offer a unique combination of optical nonlinearity, coherence, and efficient signal transport. Exciton-polaritons, as hybrid light–matter quasiparticles, provide a promising basis for implementing neural-network-inspired hardware. A major challenge, however, has been that most polaritonic platforms demonstrated so far require cryogenic temperatures, which limits their technological relevance.
I will present a broader perspective on room-temperature photonic hardware for neuromorphic computing, focusing on perovskite nonlinear optical materials and microstructured platforms capable of supporting polariton condensation under ambient conditions. Unlike conventional architectures, this platform does not require external cavity mirrors, which simplifies fabrication and improves compatibility with integrated photonics. We observe polariton lasing at wire edges and corners, together with signatures of mutual coherence and long-range propagation.
Altogether, these developments point to promising hardware platforms for tasks such as classification or pattern recognition, and bring polaritonic neuromorphic computing closer to practical implementation.Zoom details
Zoom link: [LINK]
Meeting ID: 842 4891 1743
Passcode: 394021
