R&D/innovation

University of Houston Research Targets AI’s Power-Consumption Problem

Researchers at the University of Houston have developed a new ultrathin, carbon-based film that could make AI chips run faster, cooler, and more energy-efficient.

karim-thin-film-large-crop-02.jpg
The lightweight, 2D film, seen here, was designed to make AI faster and reduce power consumption.
Source: University of Houston

Researchers at the University of Houston (UH) have invented a new ultrathin material that could help AI computers run much faster while using far less electricity.

“Many AI data centers employ vast cooling systems that consume large amounts of electricity to keep the thousands of servers with integrated circuit chips running optimally at low temperatures to maintain high data-processing speed, have shorter response time, and extend chip lifetime,” said Alamgir Karim, Dow Chair and Welch Foundation Professor at the William A. Brookshire Department of Chemical and Biomolecular Engineering at UH.

The team created a lightweight, 2D film made mostly of carbon using a method developed by recent Nobel Prize–winning chemists. Because of the way the film is built with tiny, crystal-like layers full of small holes, it doesn’t trap much electricity or heat. This allows signals inside a chip to move faster without overheating.

The findings are a result of work by UH Professor Alamgir Karim, former doctoral student Maninderjeet Singh, civil engineering Professor Devin Shaffer, and doctoral student Erin Schroeder.

The research was funded by the American Chemical Society’s Petroleum Research Foundation New Direction program.

Learn more about the research here.