HR 5332119th CongressIntroduced

Liquid Cooling for AI Act of 2025

Introduced: Sep 11, 2025

Technology & Innovation

Standard Summary

Comprehensive overview in 1-2 paragraphs

The Liquid Cooling for AI Act of 2025 directs the Comptroller General (GAO) to perform a technology assessment focused on liquid-cooling systems for artificial intelligence compute clusters and high-performance computing facilities. The goal is to evaluate research needs, costs and benefits, and market conditions, and to develop federal government-wide best-practice guidance for designing, operating, and deploying liquid cooling in data centers. The bill also requires the Department of Energy (DOE) to designate a liquid cooling advisory organization to help shape the review and to deliver a report with findings and recommendations to Congress. Within 90 days of enactment, GAO must provide a report; within 180 days, DOE must assess the GAO report and offer considerations for Congress about maintaining U.S. leadership in AI and recommendations for R&D on liquid cooling and heat reuse. The act emphasizes potential energy savings, higher compute density, and better resilience, while addressing safety, cybersecurity, and interoperability concerns. In short, the bill aims to accelerate, standardize, and inform the federal government’s use of liquid cooling for AI and HPC workloads, with a focus on energy efficiency, performance, and the development of federal best practices and reference architectures.

Key Points

1GAO technology assessment timeline and scope
2- Initiate review within 30 days of enactment.
3- Examine research and development needs, costs/benefits, energy impacts, compute capacity gains, performance and security effects, market trends, and comparisons between direct-to-chip cooling and immersion cooling across density bands.
4Comprehensive review elements
5- Include coolant options (water, water-glycol, engineered fluids), materials compatibility, corrosion control, filtration, degasification, monitoring, safety, failure modes, and lifecycle costs.
6- Assess heat reuse opportunities, reference architectures (rack/row/room layouts), and failure mitigation strategies (e.g., pump failures, leaks in shared facilities).
7Advisory organization and stakeholder engagement
8- The Secretary of Energy and GAO must designate a liquid cooling advisory organization to consult with federal and non-federal stakeholders (industry, academia, national laboratories, manufacturers, operators) and promote best practices.
9Reports and congressional follow-up
10- GAO to report findings within 90 days of enactment to the Secretary of Energy and relevant committees.
11- DOE to evaluate the GAO report within 180 days and provide congressional-facing assessments, including considerations for U.S. leadership in AI and recommendations on R&D and heat reuse.
12Definitions and scope
13- Defines liquid cooling, direct-to-chip cooling, immersion cooling, heat reuse, AI as per established federal AI definitions, and National Laboratories for consistency.
14- Clarifies which congressional committees are involved in oversight.

Impact Areas

Primary group/area affected- Federal data centers, AI compute clusters, and high-performance computing facilities that may adopt liquid cooling; guidance could shape federal cooling procurement, design standards, and deployment decisions.Secondary group/area affected- DOE, National Laboratories, federal agencies operating data centers; hardware and data-center equipment vendors; AI facility developers; and researchers focusing on cooling technologies, energy efficiency, and resilience.Additional impacts- Potential energy-use reductions and improved compute density, with implications for electricity infrastructure planning and potential waste-heat reuse opportunities.- Enhanced emphasis on security, reliability, and interoperability in cooling systems, plus a driver for education and workforce training on best practices.- Could influence industry standards and the development of reference architectures for data-center liquid cooling across density bands.Direct-to-chip (DTC) liquid cooling: Liquid coolant is circulated in direct contact with heat-generating components (like CPUs/GPUs) to remove heat.Immersion cooling: Electronics are submerged in a dielectric (non-conductive) liquid to remove heat.Heat reuse: Capturing waste heat from data centers for beneficial secondary uses (e.g., heating buildings or other processes).Liquid cooling for data centers may involve components such as coolant distribution units, secondary loops, manifolds, hoses, valves, pumps, leak detection, corrosion control, and instrumentation to ensure safe, reliable operation.

Generated by gpt-5-nano on Oct 2, 2025