HMNTL Master Calendar

Electrical and Computer Engineering Faculty Candidate Seminar

by Zhengqi Gao, PhD Candidate, Electrical and Computer Science Department, Massachusetts Institute of Technology

Wednesday, April 1, 2026, 11:00 am-12:00 pm

B02 CSL Auditorium or Online via Zoom

Title: Intelligent Design Automation for Heterogeneous Electronic-Photonic Integrated Systems

Abstract: The exponential growth of AI and data-intensive computing has fundamentally outpaced the scaling capabilities of conventional hardware reliant solely on electronic transistors. To bridge this gap, heterogeneous electronic–photonic integrated circuits (HeteroEPICs) have emerged as a promising solution, leveraging the unique advantages of photonics. However, the practical adoption of HeteroEPICs is currently impeded by three significant challenges: simulation software, hardware robustness, and system efficiency.

In this talk, I present my research on intelligent design automation for HeteroEPICs, structured to directly address these three challenges. First, to tackle the simulation software bottleneck, I introduce SPIPE, a differentiable electronic–photonic co-simulator (attracted interest from Cadence). Second, addressing hardware robustness, I demonstrate LightSim, a generic emulator designed to evaluate the reliability of foundation models (e.g., LLMs) running on HeteroEPIC accelerators. Third, to maximize system efficiency, I propose KirchhoffNet, a novel computing paradigm for generative AI inspired by the physics of electronic-photonic computing (recognized with an MLSys 2024 Rising Star Award and used in UCSD CSE 291E). I conclude by highlighting the critical role of AI in post-Moore design automation, exemplified by my research in LLM post-training for digital electronic design. Finally, I propose an exciting future agenda for large-scale heterogeneous electronic-photonic integration.

Zhengqi Gao is a final-year Ph.D. student in the Department of Electrical Engineering and Computer Science at the Massachusetts Institute of Technology. He received his B.E. and M.S. degrees from the School of Microelectronics at Fudan University in 2018 and 2021, respectively. His research focuses on design automation for heterogeneous electronic-photonic integrated circuits, and machine-learning-driven hardware/software co-design. His first-author publications span electronic design automation (EDA), photonics, and machine learning venues, such as ICCAD, DAC, IEEE Journal of Lightwave Technology, Optica Photonics Research, NeurIPS, ICLR, and ICML. His interdisciplinary research has been recognized across these communities: in EDA, earning First Place at the ACM/IEEE DAC Ph.D. Forum (2025); in photonics, earning Editor’s Highlights in Optica Express (2024) and Photonics Research (2023); and in machine learning, being named an ML and Systems Rising Star (2024). He also interned with NVIDIA Research on semiconductor lithography and with Apple on Vision Pro hardware/software design.