Experts Explore Potential of Google AI Tools to Mitigate Energy Challenges for US Data Centers

Strategic Insights -

US researchers and scientists from national laboratories and Google teams joined forces to explore the potential of generative artificial intelligence (AI) tools in mitigating energy challenges for US data centers. The hackathon, hosted by the National Renewable Energy Laboratory (NREL) and Google, brought together experts from nine US Department of Energy (DOE) national laboratories to leverage Google's AI and large language model tools. Participants explored real-world problems in geospatial analytics, energy systems, data center optimization, and digital-twin development, demonstrating the potential of generative AI in accelerating research and automating complex tasks.

Key Takeaways:

  • The hackathon, held on June 17-18 in Washington, D.C., involved 50 top minds, including six NREL computational and data scientists, who engaged in hands-on experimentation with Google's AI tools, particularly Gemini.
  • Participants focused on geospatial analytics, leveraging Gemini to process and interpret spatial data, code-generation and debugging capabilities, and in-depth research and brainstorming.
  • Researchers explored specialized problems, such as predicting power grid behavior and interconnections in counties using the Population Dynamics Foundation model.
  • The event generated robust discussion, shared learning, and discoveries, identifying opportunities for future follow-on events and fostering innovation in the use of generative AI for scientific research.
  • NREL's computational science team, Google's public sector team, and the US DOE national laboratories collaborated to explore the challenges and opportunities of AI in data center energy optimization.
  • The hackathon provided a platform for experts to directly engage with Google's AI tools, particularly Gemini, and explore their applicability to a diverse range of scientific and engineering problems.

Statistics:

  • 50 top minds from national laboratories and Google teams participated in the two-day hackathon.
  • The hackathon involved experts from across nine US DOE national laboratories, including Argonne, Idaho, Jefferson, Lawrence Berkeley, National Energy Technology, Oakridge, Pacific Northwest, and Sandia.
  • Researchers utilized Google's AI tools, including Agentspace, Gemini, and Deep Research, to accelerate research and automate complex tasks.
  • The event generated approximately 50 research projects and demonstrations, focusing on geospatial analytics, energy systems, data center optimization, and digital-twin development.

Sources:

  • NREL and Google teamed up to host a hackathon to explore and leverage Google's generative AI and large language model tools to address critical challenges related to energy limitations for US data centers.
  • Google's AI platform includes several tools, such as Agentspace, Gemini, and Deep Research, that enable researchers to accelerate and expand their work overall.
  • Regiuel Days, account executive for federal science and research at Google, stated that the hackathon successfully demonstrated the potential of generative AI in accelerating research, automating complex tasks, and generating novel insights.
  • NREL's Gabriel Steenberg explored the Population Dynamics Foundation model to predict power grid behavior, feeding it county-level data to see if it could predict interconnections in other counties.
  • The event generated robust discussion, shared learning, and discoveries, identifying opportunities for future follow-on events and fostering innovation in the use of generative AI for scientific research.