NSF Grant to Fund Data Center Fortification Against Energy Consumption and Cyber Threats

Strategic Insights -

Researchers at Penn State are leading a team to develop a novel fault detection and diagnostics (FDD) system to proactively identify potential issues in data center cooling, which accounts for over 4% of the world's electricity every year and is projected to increase to over 13% in the next three years due to the prevalence of artificial intelligence. The system will balance proactive prediction and real-time hardware monitoring to mitigate cascading failures and cyber threats, which can result in catastrophic overheating and equipment damage.

Key Takeaways:

  • The project, funded by a three-year, $340,000 grant from the U.S. National Science Foundation, aims to develop an FDD system that can detect and solve issues facing data center cooling systems, including cascading failures and cyber threats.
  • The current FDD methods use advanced statistical modeling and AI-powered machine learning algorithms, but have inefficiencies due to the need for field expertise and the difficulty in accurately predicting rare or unanticipated events.
  • The team's system will use both digital modeling tools and physical hardware testbeds to simulate the entire facility and identify potential failures, as well as offer real-time monitoring and replacement of physical equipment.
  • The FDD system is expected to improve data center protection across the United States by increasing repair efficiency and providing reliable cooling back-ups in the event of a cyberattack.
  • The project will build upon the existing work of Romulo Meira-Goes, assistant professor of electrical engineering, and Wangda Zuo, professor of architectural engineering, who have over 10 years of experience researching commercial data center cooling.

Statistics:

  • Data centers account for over 4% of the world's electricity every year and are projected to increase to over 13% in the next three years due to the prevalence of artificial intelligence.
  • The current FDD methods use advanced statistical modeling and AI-powered machine learning algorithms to detect potential failures, but have inefficiencies that require field expertise and can be difficult to accurately predict rare or unanticipated events.
  • The grant from the U.S. National Science Foundation is three years and $340,000.

Sources:

  • "Keeping hot data cool: NSF grant to fund data center fortification" (Penn State University)
  • Wangda Zuo, Professor of Architectural Engineering, Penn State University
  • Romulo Meira-Goes, Assistant Professor of Electrical Engineering, Penn State University
  • National Science Foundation (NSF)