Solving mathematic problems with photonic integrated circuits
Open AccessFor centuries, scientists and engineers from various disciplines have strived to solve mathematical equations faster and more efficiently. While electronic digital circuits have brought about a revolution in equation solving in recent decades, it has become clear that the performance gains achieved through brute-force computational methods are reaching a saturation point. Consequently, there is a growing interest in exploring paradigms that capitalize on the inherent tendency of the universe to minimize a system's free energy, such as annealers or Ising Machines, due to their favorable complexity scaling. This paper introduces a programmable analog solver that exploits the formal mathematical equivalence between Maxwell's equations and photonic circuitry. The solver utilizes a network of nanophotonic beams to find solutions to partial differential equations. To illustrate its capabilities, the researchers designed, fabricated, demonstrated, and experimentally validated a novel application-specific photonic integrated circuit consisting of electro-optically reconfigurable nodes, achieving 90\% accuracy compared to a commercial solver. Finally, the performance of this photonic integrated chip is evaluated by simulating thermal diffusion on a spacecraft's heat shield during re-entry into a planet's atmosphere. The programmable light-circuitry presented here offers a straightforward approach to solving complex problems and holds significant potential for applications across a wide range of scientific and engineering fields.
- All rights reserved
Notice to Authors
If you are the author of this work and you have any questions about the information on this page, please use the Contact form to get in touch with us.