Graduate Student Highlight: Bharadwaj Madabhushi
Content
Bharadwaj Madabhushi, a sixth-year Ph.D. candidate in Electrical and Computer Engineering at the University of Massachusetts Amherst, is building a career at the intersection of academic research and industrial impact. His work in hardware security blends deep fundamental understanding with a focus on real-world vulnerabilities.
Before entering the Ph.D. program at UMass Amherst, Madabhushi spent five years at tech giant AMD, where he worked as a senior design engineer. His responsibilities included developing detailed verification strategies and test plans for complex SoCs (System-on-Chip) with a strong focus on validating security architecture and system functionality across multiple product lines. His technical rigor and effectiveness earned him multiple promotions and internal recognition, including the Spot Executive Award.
Madabhushi’s transition to graduate research was not a pivot away from industry but rather a deepening of his focus. At UMass Amherst, under the supervision of his advisors, Professors Sandip Kundu and Daniel Holcomb, he has been investigating critical vulnerabilities in FPGA-based cloud systems—FPGAs, or field-programmable gate arrays, are a type of configurable integrated circuit that can be repeatedly programmed after manufacturing.
Madabhushi’s research has shown how remnants of data within the memory of FPGAs, particularly in multi-tenant environments, can be extracted after a user process terminates, thus revealing a significant privacy risk. These findings, detailed in IEEE publications including DATE 2024 and HOST 2024, have contributed to increasing awareness of data persistence issues in reconfigurable hardware platforms.
In addition to identifying vulnerabilities, Madabhushi has worked to develop secure and efficient hardware solutions. In collaboration with ECE professor and department head Russell Tessier, he has explored the use of FPGAs as edge cloud accelerators in environments where centralized cloud infrastructure may be unreliable or unavailable. Work coauthored with Zhehang Zhang, a doctoral student of Professor Tessier, was published in IEEE ASAP 2025, presenting a system where computationally heavy tasks such as machine learning inference from drones are offloaded to nearby FPGAs. The study highlights how this approach enables real time decision making in critical scenarios such as emergency response and surveillance by utilizing the low latency and reconfigurable nature of edge hardware.
Madabhushi has a secondary research focus: satellite image processing techniques. His research on onboard image processing—which was featured in conferences such as IEEE IGARSS 2025 and OCEANS 2024—has demonstrated that embedded processors and TPUs (Tensor Processor Units) can be used for real-time object detection in synthetic aperture radar (SAR), which is the most widely used type of radar for remote sensing and imaging applications. Madabhushi’s innovation reduces the need for raw data transmission and has the potential to improve maritime surveillance and environmental monitoring, especially in satellite systems with limited bandwidth.
Alongside his technical achievements, Madabhushi has demonstrated a commitment to teaching and mentoring during his time at UMass Amherst—he has served as a teaching assistant for several core undergraduate courses in programming, digital systems, and hardware design, and even received the Best TA Award for his work in ECE 331
Throughout his doctoral journey, Madabhushi has remained connected to industry through internships at Intel and AMD, contributing to projects in hardware security, design verification, and automation. With a defense planned for December 2025, Madabhushi intends to return to industry, continuing to work on secure hardware systems while maintaining his connections to academia. His trajectory reflects a blend of academic insight, engineering discipline, and practical impact—a vital profile in the evolving landscape of hardware security.