Programmable photonic circuits of reconfigurable interferometers can be used to implement arbitrary operations on optical modes, enabling a platform for optically accelerating tasks in artificial intelligence and signal processing. A major obstacle to scaling up these systems is static fabrication error, where small component errors within each device accrue to produce significant errors in the circuit computation. Mitigating this error usually requires numerical optimization dependent on real-time feedback from the circuit. In this talk, I discuss our recent work on a deterministic approach to correcting these circuit errors. We apply our approach to simulations of large scale optical neural networks, finding they remain resilient to component error well beyond modern day process tolerances.

Saumil is a PhD student in the Quantum Photonics Laboratory at MIT. He received his S.B. and M.Eng. in Electrical Engineering from MIT in 2017 and 2018, respectively. Prior to starting graduate school, Saumil worked on silicon photonics as a Senior Photonics Engineer at Elenion Technologies. His current research centers on programmable photonic systems for signal processing, quantum information, and artificial intelligence.