Shah Lab Advances Ferroelectric-Based Neuromorphic Hardware for Energy-Efficient Edge Intelligence

The University of Maryland’s Department of Electrical and Computer Engineering (ECE) is advancing research in energy-efficient edge intelligence through a $450,000 award from the U.S. Army Combat Capabilities Development Command Army Research Laboratory (DEVCOM ARL) and research fellowships for two doctoral students in the Shah Lab.

The award supports the project “Ferroelectric-based Mixed-Signal Neuromorphic SoC for Edge Intelligence.” Research in the lab, led by Professor Sahil Shah, will explore ferroelectric-based devices and mixed-signal neuromorphic architectures to enable energy-efficient processing in a system-on-chip (SoC) for edge applications.

In addition, ECE Ph.D. students Irem Didin and Sayma Nowshin Chowdhury have received research fellowships through Oak Ridge Associated Universities (ORAU) in conjunction with DEVCOM ARL. Both students are advised by Shah and serve as DEVCOM Army Research Laboratory Research Associates.

These efforts address the growing need to process large volumes of sensor data directly on edge devices. Performing inference locally can improve energy efficiency and support perception, navigation, and decision-making when communications bandwidth is limited or connectivity is disrupted. Such capabilities are relevant to remote monitoring, persistent sensing, and autonomous ground and aerial systems operating in unknown or contested environments.

Together, the award and fellowships support the Shah Lab’s work toward intelligent edge systems capable of timely analysis and autonomous responses under demanding power and communications constraints.

This research builds upon previous work on energy-efficient neuromorphic systems (see IEEE publication), which explores low-power circuit designs and non-volatile memory integration to accelerate machine learning workloads at the edge.

Chowdhury received a B.Sc. in Electrical and Electronic Engineering and an M.Sc. in Communications and Signal Processing from the University of Dhaka in Bangladesh, as well as an M.S. in Electrical and Computer Engineering from the University of Maryland. Her research interests include energy-efficient hardware for machine learning and novel architectures for efficient on-chip learning.

Didin earned her undergraduate degree in Electrical and Electronics Engineering from Middle East Technical University (METU), Türkiye. Her research focuses on energy-efficient neuromorphic systems through the 3D integration of silicon-based artificial neurons and synapses, along with low-power solid-state circuits and architectures for on-chip learning in edge devices.

DEVCOM ARL connects the military, academia, and industry to develop innovative technologies that address Army research needs.

Published October 5, 2026