Congress finds the following:
(1)
The 2024 United States Data Center Energy Usage Report published by Lawrence Berkeley National Laboratory indicates that data centers accounted for 4.4 percent of total United States electricity consumption in 2023, up from 1.9 percent in 2018. The report further projects that data centers could represent between 6.7 percent and 12.8 percent of total electricity consumption by 2028, due to the rapid growth of AI, cloud computing, and other digital technologies.
(2)
Traditional air-cooled systems are reaching limits to effectively remove heat from AI chips and hardware, and liquid cooling enhanced thermal performance is increasingly becoming a necessity for high-density AI servers and data centers due to the growing power consumption and heat generation of AI workloads.
(3)
Liquid‑cooling technologies, including direct‑to‑chip (DTC) liquid cooling and single‑phase or two‑phase immersion cooling, can improve thermal performance, enable higher densities, and reduce cooling system load when properly engineered and maintained.
(4)
Effective liquid‑cooling deployments require interoperable components and engineered subsystems, including coolant distribution units, secondary loops, manifolds, hoses, quick‑disconnects, valves, pumps, filters, leak detection and containment, corrosion control, and appropriate instrumentation and controls.
(5)
Interfaces for heat‑reuse are integral to liquid systems and can reduce thermal load on heat‑rejection equipment by transferring heat through plate heat exchangers or other devices to beneficial secondary uses where technically appropriate.
(6)
Federal agencies, including the Department of Energy, are considering the deployment of AI systems across government-owned facilities.
(7)
A comprehensive, independent assessment of emerging data center architectures and cooling technologies is essential to inform efficient and cost-effective deployment decisions across the Federal Government.