Abstract

Light’s unparalleled speed and natural parallelism open new frontiers for computing. In this talk, I explore how gain-based computing—harnessing light–matter interactions in driven–dissipative systems—can tackle complex optimization tasks by encoding them in the system’s gain and loss rates. As such systems spontaneously evolve toward minimal-loss modes, they effectively solve Ising, XY, or even k-local Hamiltonians, offering novel approaches to challenging combinatorial problems. I will compare two major hardware paradigms: (i) Spontaneous Dynamics in polariton condensates, lasers, and atomic media, where interactions emerge from intrinsic nonlinearities. and (ii) Engineered Couplings via optical modulations and feedback loops, allowing more direct control over gain and loss pathways. Along the way, I will present proof-of-concept experiments that highlight the promise of these platforms and examine the key challenges in scaling them up to useful and potentially transformative computing systems.