Minjie Chen: Technology to boost energy efficiency in data centers

Sept. 2, 2019

Minjie Chen and his team are building a family of devices to dramatically reduce power consumption at the gigantic data centers that serve as the backbone of internet services and cloud computing. These centers, each holding racks of computer servers, consume 90 billion kilowatt-hours of electricity each year in the United States.

[embed]https://youtu.be/NOOcAL57Waw?rel=0[/embed]

The team’s technology restructures the way power is converted from the 480-volt alternating current of the electricity grid down to the 5-volt-or-lower direct current needed for central processing units and hard drives. In today’s data centers, this process happens at each computer, sapping about 40% of the original energy. Chen, an assistant professor of electrical engineering, and his team are building a new energy processor that reduces the voltage in a central unit, then simultaneously supplies power to a large number of computing devices. 

Instead of lots of cascaded power conversion stages, they aggregate power conversion into one unit, and then distribute that power. They estimate they can increase the energy efficiency of the power-delivery system from about 60% to 88%. The technology works with solar farms and battery storage systems as well as data centers.

In February, the team won first place at Princeton’s 14th Annual Innovation Forum, a campus-wide pitch competition hosted by the Keller Center for Innovation in Engineering Education and the Office of Technology Licensing. 

Team members: Graduate students Ping Wang and Youssef Elasser; postdoctoral research fellow Yenan Chen

Collaborator: Robert Pilawa-Podgurski, associate professor in electrical engineering and computer science at the University of California-Berkeley

Development status: Patent protection is pending. Princeton is seeking outside interest for the further development of this opportunity.

Funding: National Science Foundation, U.S. Department of Energy ARPA-E, Siebel Energy Institute

princeton.edu/powerlab