Creating Smarter, More Efficient Computing Systems
Tergel Molom-Ochir '23
Electrical and Computer Engineering
What drew you to this field of study?
As a middle schooler, I used to take apart electronics, such as cell phones, remote controllers, and tablets, and then put them back together. It fascinated me how tiny yet computationally capable each electronic component was. This is when I first became interested in engineering and computing.
My curiosity about mobile edge computing—the practice of capturing, storing, processing, and analyzing data near where the data is generated—was sparked when I questioned whether a system could be designed to perform well and be energy efficient at the same time. In 2019, as a freshman, I joined Professor Prashant Shenoy’s Laboratory for Advanced Software (LASS) in the Department of Electrical and Computer Engineering at UMass Amherst. Intrigued by efficient computing and inference-making, I continued working in the lab each year. As I learned more about the research process, I became independent with my day-to-day work. I then delved into work with PhD student Walid Hanafy on energy-aware and energy-efficient machine intelligence through algorithm and hardware selection.
How do you conduct your research?
I start by considering multiple questions or directions that can help the research move forward. Even if I head down a “wrong” path, valuable data is yielded that eventually sets me in the right direction. In conducting research, I usually find myself drawing on a whiteboard, writing down pieces and parts of information, zooming out, and connecting the dots. Once I am done brainstorming, I begin writing programs and do some computational work on computers or cluster computers.
What do you see as the impact—or potential impact—of your work?
With recent advances in software and hardware pushing forward the limits of computations, our society faces a complex set of questions when it comes to energy efficiency and performance constraints. This matters because with the help of mobile edge computing, we can reduce operational costs and be smart and efficient with energy usage, all the while improving performance of the device and the application. Through my research, I draw attention to the selection of an energy-efficient hardware accelerator under cost, power, and performance constraints.
With research, sometimes you can reach a dead end and get stuck, but these moments can also be exciting as they provide feedback on your approach.
How does your faculty mentor support your research?
Professor Shenoy gives valuable advice on research direction and allows me to figure things out by myself, which allows me to take creative approaches and learn a lot. We discuss what to do, and then the specifics of how to do it are generally up to me and my creativity. His support and approach allow me to focus on a problem and have fun.
What do you find most exciting about conducting research?
The universe is like a puzzle, and solving parts of the puzzle can be exciting. With research, sometimes you can reach a dead end and get stuck, but these moments can also be exciting as they provide feedback on your approach.
What are you most proud of?
I am proud of the work we do at LASS; the questions we ask and the problems we solve are quite challenging and serious. I am proud to be part of a team that investigates important systems issues for distributed systems. In terms of specific accomplishments, I would mention my co-authored papers, which were presented at the ACM (Association for Computing Machinery) e-Energy Conference, ACM SenSys, and IEEE MIT URTC (Undergraduate Research Technology Conference).
How has your research enhanced your overall educational experience at UMass?
The research that I have done during my undergraduate career has helped broaden my understanding of computation and energy efficiency. It has also presented valuable learning experiences, rewarding collaborations, and insight into graduate-level research. Moreover, the challenging problems we solved and the questions we answered through our work have motivated me to start thinking about energy efficiency, latency, and accuracy from the hardware design side as well. This has influenced my choice of engineering classes in the field of nanodevices and neuromorphic engineering.
What are your plans for the future?
Working in the lab under Professor Shenoy has inspired me to pursue graduate school. In the future, I intend to complete a PhD and continue my involvement in research.
Why would you recommend UMass to a friend?
I would recommend UMass because of its extensive facilities, impressive selection of classes, and other resources available, which enable undergraduates to contribute to cutting-edge research. In any field you can think of, UMass Amherst most likely offers classes or has laboratories undertaking research in that area.