NVIDIA has introduced its Rubin AI servers that utilize a 45°C water-glycol coolant system to efficiently cool chips and networking components, a significant advancement over traditional air cooling methods. This innovation allows for a dramatic decrease in water usage for cooling, potentially reducing consumption from about 2.6 million gallons per megawatt per year to nearly zero in optimal climates. As the data center industry increasingly shifts towards efficient cooling options, these liquid cooling methods not only enhance energy efficiency but also facilitate heat reuse, contributing to a more sustainable operational model.

NVIDIA: NVIDIA is a leading technology company specializing in graphics processing units and AI accelerators for data centers and high-performance computing. The company recently highlighted advancements in liquid cooling for its Rubin AI servers that support warmer coolant temperatures and closed-loop systems. This development is central to the news as it addresses water usage in AI infrastructure through direct-to-chip cooling.
Manhattan Institute: The Manhattan Institute is a public policy think tank that conducts research on urban issues, economics, and technology impacts. It has examined water consumption patterns in U.S. data centers and noted shifts toward more efficient cooling approaches. The institute’s analysis provides supporting context for the liquid cooling advancements discussed in the NVIDIA announcement.

`json
{
“Industry Shift”: “AI data center operators are increasingly adopting liquid cooling systems that reduce water use and enhance energy efficiency.”,
“Cooling Innovation”: “Liquid cooling methods for AI servers are advancing the development of closed-loop systems that significantly lower the dependency on conventional cooling methods.”,
“Environmental Focus”: “Water usage in data centers is a critical issue, driving the adoption of innovative cooling technologies aimed at minimizing water consumption.”
}
`