Research and Publications

I have a combined background in computational science and domain sciences of physical waves and electronics. My current primary research interest is in high-fidelity computational algorithms for microelectronics applications. I work closely with material scientists, device designers, circuit architects and machine learning experts, to codesign workflow that benefits the development of future electronics.


My github profile 

Scroll down for my full publication list or visit my Google Scholar


Beyond Moore's Law

As Moore’s Law first predicted in 1975, CMOS silicon chips are approaching limits in miniaturization and performance.  It is critical to explore new physical phenomena that will lead to significantly higher energy efficiency in microelectronics. Ferroelectric, spintronic, and multiferroic materials have become leading contenders for future electronics. Orders of magnitude improvement in energy efficiency are possible by exploiting correlations (electronic charge/spin and dipolar). Thus, we aim to design and manipulate this energy barrier to specifically reduce the operating voltage substantially below what is achievable by today's CMOS technology.

Ferroelectric materials have enabled a wide portfolio of innovative microelectronics devices due to their switchable polarization in response to applied electric fields. The remnant polarization in the ferroelectric material at zero applied electric field allows for nonvolatile retention in these devices. The unique physics of ferroelectric field effect transistors (FeFETs) has been instrumental in the design of other new technologies including nonvolatile memories, logic-in-memory (LiM) architectures, oscillators and negative capacitance field effect transistors (NCFETs). We have built a massively parallel, 3D phase-field simulation framework for modeling ferroelectric materials-based scalable logic devices. The charge (Q) v.s. voltage (V) responses for these 3D structures clearly indicate stabilized negative capacitance with multidomain formation, which is corroborated by amplification of the voltage at the interface between the ferroelectric and dielectric layers.

Spintronics has emerged as one of the leading options for low-energy computing due to an intriguing combination of non-volatility, higher logic efficiency, and the potential for logic-in-memory function. Spintronics aims to enhance and replace standard charge-based electronics by harnessing the electron’s spin. Materials that have more than one ferroic ordering at the same time are known as ‘multiferroics’. In particular, strong magnetoelectric (ME) coupling in multiferroic materials could enable ultra-low voltage switching to convert charge to spin via ME actuation at <100 mV. We are actively developing highly efficient logic-in-memory operations with robustly switched spins by electric fields. 

Related publications:


Prabhat Kumar, Andy Nonaka, Revathi Jambunathan, Girish Pahwa, Sayeef Salahuddin, and Zhi Yao, Computer Physics Communications 290 (2023): 108757.

accompanying code repository 


Sajid Husain, Isaac Harris, Guanhui Gao, Xinyan Li, Peter Meisenheimer, Chuqiao Shi, Pravin Kavle, Chi Hun Choi, Tae Yeon Kim, Deokyoung Kang, Piush Behera, Didier Perrodin, Hua Guo, James Tour, Yimo Han, Lane Martin, Zhi Yao, and Ramamoorthy Ramesh, Nature Communications, 15, 479, 2024.


Xiaoxi Huang, Xianzhe Chen, John Mangeri, ... , Zhi Yao, Ramamoorthy Ramesh, under review, Link to Arxiv.


Exascale Modeling

As the microelectronics community continues to explore new materials and technologies, the demand for modeling tools has exceeded current capabilities. Emerging post-CMOS technologies often rely on trial-and-error development strategies due to the lack of adequate simulation tools. There is an ever-increasing need for higher-fidelity simulations via higher spatiotemporal resolution and/or improved coupling that can seamlessly incorporate new physics into algorithms for widely-used, standard models.

We address the need for enhanced modeling for more realistic devices by developing an algorithmically flexible capability that is performant on manycore/GPU-based supercomputers. The main product of this research is the ARTEMIS package. ARTEMIS is able to effectively capture the multiphysics aspect of emerging microelectronics, with increased spatial resolutions. This allows for GPU simulations of various devices including multiferroic logic, ferroelectric capacitors and transistors, magnetic RF devices, high-frequency circuits, etc. 

Code Packages:

Click here for direct access to the Microelectronics project on GitHub.

Radio-frequency Devices

RF components that rely on coupled physical mechanisms have gained prominence. Devices like acoustic wave resonators and magnetically induced antennas provide excellent performance for radio frequency front-ends. This will enable broader exploitation of new mechanisms for everything from radar systems to tiny, implantable health-monitoring devices.

Related publications:


Rui-Fu Xu, Louis-Charles Ippet-Letembet, Sidhant Tiwari, Zhi Yao, Shih-Ming Huang, Rob N Candler, Shih-Yuan Chen, Applied Physics Letters 123, no. 16, 2023.


Zhi Yao, Sidhant Tiwari, Joseph Schneider, Robert N. Candler, Greg P. Carman, and Yuanxun Ethan Wang, IEEE Journal on Multiscale and Multiphysics Computational Techniques, 6, pp.249-258, Dec. 2021.


Zhi Yao, Yuanxun Ethan Wang; Scott Keller; Gregory P. Carman, IEEE Transactions on Antennas and Propagation, vol. 63, pp. 3335-3344, Aug. 2015.

Student Project: Hybrid Quantum

Hybrid quantum systems offer a new paradigm of combining quantum modules and well-defined dynamic physics in the classical regime. They have found applications in coherent information processing, as the coherence of information carried in dynamic excitations can be maintained while being transduced between modules.  

Full Publication List: