Hometown: Gilroy, California, United States
Graduation date: Spring 2023
FURI | Fall 2022
Characterization of Synaptic Electronic Devices for Brain-Inspired Computing Systems
The project, a continuation from Spring 2022, seeks to answer the research question: Is hexagonal boron nitride (h-BN), a layered two-dimensional (2D) material, a promising candidate to implement brain-inspired (neuromorphic) computing devices and circuits? Research this semester will extend beyond testing the standalone behavior of individual memristor devices to investigating their collective ability to implement machine learning algorithms in a memristor array. The focus will be on performing dot-product operations, an action fundamental to nearly all neural network models. Comprehensive electrical measurements and statistical analysis will be used to demonstrate pulsed programming of a memristor array and its ability to carry out dot-product operations as part of executing a logistic regression classification task.
Mentor: Ivan Sanchez Esqueda
Featured project | Fall 2022
Electrical engineering senior Sritharini Radhakrishnan is working to improve brain-inspired, or neuromorphic, computing through her FURI project with Ivan Sanchez Esqueda, an assistant professor of electrical engineering. Her work focuses on testing a certain type of memristor, a new type of electric circuit component that retains memory even without power. This was Radhakrishnan’s first engineering research experience, and she cites it as a valuable one.
What made you want to get involved in FURI?
I wanted to participate in FURI for the unique opportunity it provides students to explore state-of-the-art engineering research under the guidance of distinguished faculty members. Before being part of this program, my idea of research was primarily informed by the wet-lab chemistry experiments I had performed in my life sciences classes. As a third-year electrical engineering student, I thought it was long overdue to figure out what research could be in the context of engineering.
Why did you choose the project you’re working on with Assistant Professor Ivan Sanchez Esqueda?
I chose to work with Dr. Ivan Sanchez Esqueda because the project he offered was truly cutting-edge. The field of neuromorphic computing is of ever-increasing interest because it promises to support artificial intelligence, cloud computing and Internet of Things applications energy-efficiently even as the amount of data produced outgrows the computing power of conventional circuits.
Memristor devices are crucial to bringing this possibility to fruition because they act as synaptic devices to enable a neuromorphic computing scheme. Dr. Sanchez Esqueda’s research explores 2D memristors to dramatically increase the performance and efficiency of the devices while making them scalable to use in next-generation integrated electronic systems. By working with him, especially as an undergrad, I felt I could get a head start in gaining the technical know-how and experience to significantly contribute to any engineering research in my career.
What was it like getting started in FURI?
After I reached out to Dr. Sanchez Esqueda, he explained his ongoing projects very clearly and answered all my questions with a lot of patience. After our first meeting, he sent me helpful research reviews to read and answered even more of my questions via email. He worked with me to write my FURI application and come up with milestones I felt confident in achieving. During the lab tour he set up with [Graduate Research Associate] Sahra Afshari, a brilliant and hardworking PhD student, and my graduate research mentor, I realized how inclusive and nurturing his lab environment was. I am beyond grateful for starting my engineering research career with his group.
How will your engineering research project impact the world?
This project will make a difference in the field by providing more substantial evidence of 2D memristor array capabilities and the potential of neuromorphic computing systems.
Everyone in the field aims to advance neuromorphic computing to the point that any device using the current von Neumann computing architecture can be replaced with a neuromorphic one. To achieve such a goal, it is vital to show that the neuromorphic computing scheme is competitive with the von Neumann one by demonstrating that neuromorphic circuits can carry out complex operations faster and with greater efficiency.
My FURI project will show how an array of 2D hexagonal boron nitride (h-BN) memristor devices can perform the dot-product operation, a function common to all machine learning algorithms. To further the demonstration, the project will show that the h-BN memristor array can perform linear and logistic regression.
Did you have a particular “aha!” moment during your project?
My “aha!” moment happened recently. I was reading a paper about new advances in neuromorphic computing and was surprised to see that I followed along without having to backtrack and read the text twice. I noticed that graphs of device characteristics made sense at first glance and that I was asking myself good follow-up questions as I read along. I have to say that all the time spent in the lab working with Sahra and Dr. Sanchez Esqueda has really paid off! Without their constant support and involvement in my professional growth, I wouldn’t be where I am.
How do you see this experience helping with your career goals?
This experience has opened my mind to pursuing a graduate degree in engineering. Both industry and academic research are great places to practice clever thinking in the face of complex challenges. But I think the latter is better suited to me because of the space I’ll have to figure out how all the electrical engineering tools I’ve gained so far can help me become a better problem-solver. I learn best through trial and error. As there’s flexibility and room for mistakes in academic research, I look forward to that freedom!
Having a great faculty mentor like Dr. Sanchez Esqueda and a graduate research mentor like Sahra has also proven to me that academia will always go the extra mile to prioritize your learning.
The knowledge I gained working with h-BN memristors as a resistive random-access memory device helped me to land a memory validation internship with Intel and a research aide position working with memristor technology developed by Sandia National Laboratories with electrical engineering Associate Professor Matthew Marinella.
What is the best advice you’ve gotten from your mentor?
Even in the face of a device physics issue that we couldn’t do much to change, Dr. Sanchez Esqueda made the most of the situation by explaining the reason behind it and brainstorming ways to create demonstrations with what we had so we could highlight important trends in memristor devices.
The lesson I took away is that although the majority of research is dealing with things that don’t go as planned, it is important to find value in the result either way. In seeing how Dr. Sanchez Esqueda handles the unexpected in his lab, I see that having a long history of “failures” can lead to inventing alternative, equally enlightening paths to answering a research question.
Why should other students get involved in FURI?
I think all engineering students should participate in FURI because it’s a great way to witness and participate in the top-tier research happening across the university. Being in the program is another way to broaden your network within the Fulton Schools. FURI students can develop critical technical skills to utilize in their next class or job while building long-lasting relationships with professors and graduate students.