Hajiesmaili Receives NSF Grant for Project on Machine Learning, Data and Energy

Mohammad Hajiesmaili
Mohammad Hajiesmaili

Mohammad Hajiesmaili, research assistant professor in the College of Information and Computer Sciences (CICS), has received a $500,000 grant from the National Science Foundation (NSF) to use machine learning to make data storage more energy efficient. It’s the second grant for the project, following the Google Research Award he received in April along with co-investigators Ramesh Sitaraman, professor at CICS, and David Irwin, professor of electrical and computer engineering.

“These two grants for Hajiesmaili’s research are a recognition of how increasingly important it is to address energy use in computing,” said Prashant Shenoy, professor and associate dean at CICS, and director of the Center for Smart and Connected Society. “The application of machine learning to data center energy procurement is a promising approach for building greener and more reliable distributed systems.”

Data centers are using increasingly massive amounts of power, consuming by some estimates up to 3% of the world’s power supply and doubling their appetite every four years. While companies such as Google and Amazon offset the energy consumption of their data centers with on-site wind and solar farms, they continue to face the challenge of how to most efficiently gather energy from multiple sources, such as renewables, batteries, and the public power grid.

Hajiesmaili’s project aims to follow up on proven successes in using machine learning to improve the energy efficiency of data centers—for example, at one of Google’s centers, the DeepMind AI reduced the cooling bill by 40%—by using machine learning to optimize the energy procurement process. By modeling how to more efficiently use renewable sources and energy storage systems, this research has ramifications beyond data centers.

“A broad goal of my research is to ease the reliable and efficient incorporation of renewable energy into IT infrastructure and the electric grid,” says Hajiesmaili. “This research is a significant step towards reducing the energy cost of data centers and thereby reducing the overall cost of internet services. More broadly, by providing a robust way to incorporate local renewable sources to the energy portfolio of large energy customers, it facilitates the integration of renewables into the electric grid, leading to a more green and sustainable world.”