Georgia Tech School of Electrical and Computer Engineering (ECE) Ph.D. student Prasanna Venkatesan won a Student Best Poster Award at the Superior Energy-efficient Materials and Devices (SUPREME) Annual Review, a Joint University Microelectronics Program (JUMP) 2.0 research center focused on microelectronics, held at Cornell University.

He was recognized for his poster, “Direct Observation of Nanoscale Polarization Switching in HZO Ferroelectrics Using STEM-EBIC,” which explored atomic-scale, multi-bit ferroelectric memory, a new ultra-fast, low-power, and high-reliability non-volatile memory technology.

The work addressed two of the central challenges facing the future of data storage: how small a memory cell can shrink, and how many bits it can reliably hold.

Hafnium-oxide (HfO₂) based ferroelectrics, including the hafnium-zirconium-oxide (HZO) that Venkatesan studied, are among the most promising materials for dense, energy-efficient, non-volatile memory. They retain their polarization down to films just a few nanometers thick and are compatible with existing semiconductor manufacturing. Yet the microscopic physics of these materials’ polarization switching— the reversal of the internal electrical charge in ferroelectric materials—has remained poorly understood.

The reversal of a device's polarization state is important for this technology as it enables non-volatile data storage for next-generation low-power and high-speed computing systems.

Using scanning transmission electron microscopy with electron-beam-induced current performed in Professor Chris Regan’s group at the University of California, Los Angeles (UCLA), Venkatesan and his collaborators were able to image polarization switching in HZO down to 0.475 × 0.475 nanometers (nm)—the smallest ferroelectric unit ever measured—demonstrating that well-defined, reliable switching persists all the way to atomic dimensions.

This revealed for the first time that ferroelectrics exhibit sharp switching with no intermediate steps at the atomic and domain-scale. The team also identified that the switching characteristics exhibit wide variation even across 100 nm. 

By demonstrating robust switching down to atomic dimensions, the team has paved the way to sub-nm scale memory while also demonstrating the possibility of robust multibit operation by carefully engineering the statistics of switching characteristics across a ferroelectric film.

The work was also presented at the 2025 IEEE International Electron Devices Meeting (IEDM), the premier international forum for microelectronics research. The work was completed in collaboration with UCLA, the University of Virginia, NanoElectronic Imaging Inc., Applied Materials, Samsung Electronics, and additional academic and industry partners.

Venkatesan’s research focuses on semiconductors, with over six years of experience in memory technologies and AI-assisted device characterization and modeling.

He currently leads collaborations with industry partners to integrate AI for accelerated semiconductor characterization, device modeling, and process optimization, including leaders like Samsung and Intel.

Additional awards and honors include best paper awards at International Reliability Physics Symposium 2025, International Integrated Reliability Workshop 2024, CRNCH Ph.D. Fellowship, IEEE Electronics Device Society Masters’ Fellowship 2020, and selections to the S. N. Bose Scholars Program 2018 and IUSSTF-Viterbi Program 2018.