Author ORCID Identifier
https://orcid.org/0000-0001-9569-1817
Defense Date
2026
Document Type
Dissertation
Degree Name
Doctor of Philosophy
Department
Electrical & Computer Engineering
First Advisor
Prof. Supriyo Bandyopadhyay
Second Advisor
Prof. Vitaliy Avrutin
Third Advisor
Prof. Nibir Dhar
Fourth Advisor
Prof. Jayasimha Atulasimha
Fifth Advisor
Prof. Avik Ghosh
Abstract
Nanomagnetic devices are of great interest in digital hardware because of their non-volatility and dynamic ability to change magnetization but suffer from high switching error rates and temperature sensitivity. Magnetostrictive nanomagnets that utilize strain to switch between stable magnetization states encoding bit information are of interest since they are extremely energy efficient as piezoelectric layers can be used to rotate magnetization that have switching energies in the range of attojoules. Their stochasticity can also be useful in probabilistic, analog, neuromorphic, and collective computing systems, where occasional switching errors are not devastating. The dissertation extends spintronics beyond conventional computing schemes by using strain to reshape energy barriers and utilize stochasticity for various applications. It first discusses a strain operated Magnetic Tunnel Junction (MTJ) based matrix multiplier device that enables multiplication, accumulation and allows nonvolatile storage with substantially fewer magnetic devices than conventional crossbar approaches. Strain is explored further to reconfigure low-barrier nanomagnets from binary stochastic neurons to analog stochastic neurons by progressively making anisotropic energy barrier flatter, while simultaneously tuning fluctuation rate, correlation time, bandwidth, and noise spectrum. The same energy-landscape engineering principle is further applied to a ternary stochastic neuron with aid of both strain and spin polarized current. Independent studies examine Weyl-semimetal based low energy Spin Field Effect transistors, change in topological insulator transport using ferromagnetic nanomagnetic, and the sensitivity of stochastic-neuron performance to material and dimensional parameters and spin inertia. Together, these investigations connect device physics with hardware concepts to establish straintronics as a platform for unconventional computing.
Rights
© The Author
Is Part Of
VCU University Archives
Is Part Of
VCU Theses and Dissertations
Date of Submission
8-4-2026