NVIDIA States AI Data Centers Use 0.2% of Daily U.S. Water, Significantly Reduced Due to Liquid Cooling Technology
After adopting 45°C liquid cooling technology, AI factories can use dry coolers instead of traditional cooling tower systems under suitable climate conditions, reducing cooling water usage from about 2.6 million gallons per MW per year to nearly zero.
Liquid cooling allows AI factories to achieve both water and energy savings while providing opportunities for heat reuse to local communities, becoming assets to the energy grid.
Source: Public Information
ABAB AI Insight
NVIDIA has previously promoted liquid cooling solutions, and this data emphasizes that the environmental footprint of AI data centers is manageable, continuing its strategy of expanding from GPU hardware to a full-stack green infrastructure.
On the capital front, liquid cooling technology reduces regulatory and community resistance, directing funding towards the construction of AI factories that support efficient cooling, while heat reuse provides a new pathway for energy assetization.
Similar to the water conservation commitments of Google and Microsoft data centers, the current AI infrastructure is at a critical juncture of transitioning from power bottlenecks to dual optimization of water resources and sustainability, with 45°C liquid cooling becoming an industry benchmark.
Essentially, this involves technological substitution and capital concentration, where liquid cooling innovations alleviate environmental controversies, shifting pricing power from traditional cooling towers to efficient liquid cooling systems, accelerating the positioning of AI factories as sustainable energy assets.
ABAB News · Cognitive Law
AI computing power manufacturing demands, liquid cooling technology resolves controversies, and efficiency improvements yield environmental dividends.
Traditional cooling towers represent an old paradigm, while dry + heat reuse is the new direction; the greening of infrastructure determines social acceptance.
Data centers are not energy-consuming islands but energy assets, with pricing power determined by technologies that balance computing power and sustainability.