Elon Musk has backed NVIDIA’s claim that water use in data centers is much lower than critics suggest.
Data centers face increasing scrutiny over the amount of water and electricity they consume. NVIDIA says its latest cooling systems can nearly wipe out the water that older installations lose to evaporation.
What NVIDIA claims about data center water use
Nvidia mail Cited March 2026 appreciation From the Manhattan Institute that data centers use approximately 0.2% of fresh water in the United States, most of it indirectly through power generation.
The company said 45°C liquid cooling allows AI factories in cold climates to run on dry chillers instead of evaporation towers.
The shift could reduce the facility’s cooling water from about 2.6 million gallons per megawatt per year to nearly zero.
NVIDIA filed the same case in 2025, claiming… Blackwell The systems were 300 times more efficient in using water than air cooling.
Since cooling can take up to 40% of a data center’s electricity, the design also reduces energy costs. Now similar trade-offs are shaping the world AI data center race.
Why is Elon Musk’s endorsement important?
Musk, who runs large NVIDIA-powered clusters through xAI, has repeatedly done so He praised the latest NVIDIA chips As the backbone of his AI projects.
for him support Supports NVIDIA’s campaign against the view that the growth of artificial intelligence is depleting local water supplies. The company describes its system as a closed loop that recirculates coolant instead of consuming fresh water.
“The NVIDIA DSX reference design for AI factories consumes no water…We have eliminated huge amounts of energy use and almost all water use,” Ali Heidari, director of data center cooling and infrastructure at Nvidia
Follow us on XTo get the latest news as it happens
Caution issue
The national number hides the details. NVIDIA’s near-zero claim covers direct cooling, which is the smallest part of the footprint.
US data centers used about 17.4 billion gallons of water directly in 2023, according to Berkeley Lab. a report They were found, along with another 211 billion gallons indirectly through the electricity they consumed.
This indirect share is rising as AI expands, and direct use alone is expected to reach 38 to 73 billion gallons by 2028.
Dry coolers also depend on the climate, performing better in cold areas and less well in hot, dry conditions.
The tension can be seen in Musk’s xAI device. Its Memphis Colossus site extracted approximately 1.3 million gallons of drinking water per day from the local aquifer and ran dozens of gas turbines before obtaining permits, creating Data center pollution lawsuit and community appeals.
How regulators and water-shortage regions respond may determine whether efficiency gains keep pace with the expansion of the industry on a broader scale Artificial Intelligence Capital Competition.
this post Elon Musk backs NVIDIA after AI’s biggest environmental criticism takes a hit appeared first on BeInCrypto.
