NREL and Google Collaborate on AI Hackathon to Address Data Center Energy Challenges
In June 2025, the National Renewable Energy Laboratory (NREL) and Google hosted a two-day hackathon in Washington, D.C., gathering experts from nine U.S. Department of Energy (DOE) national laboratories. The event focused on applying Google's generative artificial intelligence (AI) and large language models to tackle energy constraints impacting the growth and scalability of U.S. data centers.
As data centers expand rapidly to support emerging technologies, their energy demands pose significant challenges. Ensuring energy reliability and cost-effectiveness is essential for sustainable scale-up. This hackathon aimed to bring together diverse expertise to explore practical AI-driven solutions for these issues.
Bringing AI Tools to Scientific Challenges
A hackathon, by definition, is an intense collaborative event where programmers, designers, and other specialists team up to develop solutions within a limited timeframe. This event stemmed from a shared interest between Google's public sector team and NREL’s computational science unit to expose researchers to advanced AI models with a clear application focus.
Ray Grout, director of NREL’s computational science center, highlighted that the growing energy demand for AI workloads and data centers made this topic a natural fit. The hackathon provided an opportunity to test Google's AI models against real-world scientific challenges.
Participants and Activities
About 50 experts, including six NREL computational and data scientists, participated in the event. They engaged hands-on with AI tools to address problems spanning geospatial analytics, energy systems, data center optimization, and digital-twin development.
Google introduced its AI platform tools, including Gemini, Agentspace, Idea Generation, and Deep Research. These tools are designed to accelerate research, support innovation, and improve operational efficiency. For example:
- Idea Generation: Combines AI with a tournament-style framework to generate and rank innovative ideas.
- Deep Research: Helps gather and analyze internal and external data efficiently.
- Geospatial reasoning and weather forecasting tools: Used to predict grid outages and support infrastructure resilience.
Key Outcomes and Insights
Participants applied Gemini's capabilities for spatial data interpretation, code generation, debugging, and brainstorming. Gabriel Steenberg from NREL leveraged a Population Dynamics Foundation model to predict power grid behavior by analyzing county-level data for interconnections.
Other DOE labs present included Argonne, Idaho National, Jefferson, Lawrence Berkeley, National Energy Technology, Oak Ridge, Pacific Northwest, and Sandia National Laboratories. These teams explored applications such as:
- Using Vertex AI and Google Earth Engine for data-center load balancing
- Real-time water data analysis
- Cybersecurity enhancement
The hackathon offered valuable feedback on how national labs might integrate Google’s AI tools into specialized scientific workflows, particularly in geospatial reasoning, digital twins, and autonomous engineering. It also fostered collaborative learning and identified opportunities for future joint initiatives.
Looking Ahead
Ray Grout emphasized the importance of uniting experts from both the national labs and Google to identify current capabilities and areas needing further development. Beth Hartman, Google’s industry executive for federal science and research, noted that the event deepened Google's understanding of the labs’ challenges and helped tailor AI tool access accordingly.
Plans are underway to host more hackathons with DOE labs, expanding participation to all 17 national laboratories. These collaborations aim to continue advancing AI applications that support energy security, grid resilience, and data center efficiency.
For professionals interested in enhancing their AI skills relevant to scientific research and energy applications, exploring specialized AI training courses can provide practical knowledge and hands-on experience. Resources such as Complete AI Training’s latest AI courses offer valuable learning paths.
Additional Resources
Your membership also unlocks:
