October 7
@
12:00 PM
–
1:00 PM
Accelerated Integrated Photonics Design Using Advanced Computing
Xinzhong (Tom) Chen
Abstract:
The growing complexity of photonic devices and integrated circuits demands design workflows that are both faster and more automated. This seminar will present recent progress in accelerating photonics design through advanced computing and AI, spanning both device- and circuit-level design. The first half will focus on GPU- and cloud-accelerated electromagnetic simulation for rapid full-wave modeling of photonic structures, with emphasis on large-scale simulation, adjoint-based inverse design, and multiphysics modeling. The second half will discuss AI-assisted photonic circuit design automation, including schematic- driven design, layout generation, and circuit-level workflows. Together, these capabilities illustrate an integrated approach to photonics design that connects fast physics-based simulation with higher-level automation, helping reduce development time and enabling more scalable photonic engineering.
Bio:
Xinzhong (Tom) Chen, Ph.D., leads Tidy3D development at Flexcompute. He holds a doctorate in physics from Stony Brook University and completed postdoctoral research at Columbia University. His work spans optical spectroscopy, photonic device design, and GPU-accelerated electromagnetic simulation, with recent focus on applying agentic AI across the full photonic design flow, from device optimization to circuit layout and verification. Tom has also built much of Flexcompute’s user-facing technical content, including its Example Library and training materials used by engineers and researchers worldwide.