Elon Musk endorsed NVIDIA's eco-manifesto: the myth of the AI water crisis debunked
Major technology corporations are increasingly facing criticism over their colossal consumption of natural resources, particularly electricity and fresh water. However, NVIDIA's leadership firmly rejects accusations of inefficient water use. Company representatives claim that modern closed-loop liquid cooling systems have virtually eliminated the water evaporation characteristic of previous-generation infrastructure. Entrepreneur Elon Musk fully agrees with these arguments, taking the discussion about the environmental footprint of advanced technologies to a new level.
Real Water Consumption Figures in the AI Industry
In its blog, NVIDIA cites data from the Manhattan Institute for March 2026: according to their estimate, data centers in the United States consume only about 0.2% of the country's total fresh water, with the lion's share of this volume indirectly going toward generating electricity to power them. Modern liquid cooling systems, such as the NVIDIA DSX reference architecture, operate at a stable temperature of 45°C. Moreover, in northern latitudes, AI complexes can fully transition to dry radiators, eliminating the need for bulky evaporative cooling towers.
As early as 2025, NVIDIA stated that Blackwell systems are 300 times more water-efficient than traditional air cooling. Given that maintaining optimal temperature "consumes" up to 40% of a data center's total energy, this innovative approach fundamentally changes the rules of global competition in the AI market.
The Role of Elon Musk's Support in the Media Landscape
Musk, whose company xAI uses large clusters based on NVIDIA solutions, has repeatedly called NVIDIA's latest chips a key element of his artificial intelligence projects. His public endorsement significantly strengthens the position of partners. Developers aim to completely dispel the popular myth that neural networks deplete regional water supplies. As Ali Heydari, head of data center cooling and infrastructure at NVIDIA, emphasized: "The NVIDIA DSX reference architecture does not consume water. We have managed to minimize energy consumption and virtually eliminate water usage."
Limitations of New Technologies and Hidden Risks
However, optimistic national reports often mask real regional problems. Direct server cooling in the United States in 2023 required 17.4 billion gallons of water, while indirect needs (energy) amounted to 211 billion gallons. Indirect costs are rapidly rising amid the large-scale deployment of neural networks. Analysts at the Berkeley Lab predict that direct consumption will jump to 38-73 billion gallons by 2028.
The efficiency of dry radiators is directly tied to climate: in cooler states, they show excellent results, but in arid desert zones, their performance drops. The startup xAI also feels environmental pressure: the giant computing complex Memphis Colossus daily extracted up to 1.3 million gallons of clean drinking water from underground sources. Additionally, the company launched dozens of gas turbines before obtaining official permits, sparking outrage among local residents and leading to high-profile lawsuits.
Expert Opinion: The further development of the industry will entirely depend on the stringency of environmental oversight by authorities. While major players like NVIDIA and xAI demonstrate impressive technological breakthroughs in cooling, regional conflicts and hidden costs remain a serious challenge. Investors should closely monitor regulatory changes: they could become a decisive factor for asset values in the AI sector.