The artificial intelligence boom is generating a significant shift in data center management, with liquid cooling technologies emerging as a key investment opportunity. As AI systems grow more powerful, the heat they produce necessitates innovative solutions to prevent energy inefficiencies linked to traditional air-cooling methods.
The transition to liquid cooling is rapidly gaining traction. Analysts from Goldman Sachs have noted it as one of the fastest-growing themes associated with AI advancements. Shawn Tuteja, who leads the ETF and custom baskets volatility trading team, mentioned, "Our baskets team sees the next spin-off or the next iteration of the AI equipment trade being within liquid cooling." This shift is grounded in a growing consensus among investors and technology companies that energy-efficient cooling systems will be critical as AI applications expand.
Conventional cooling systems consume vast amounts of electricity, often using nearly as much energy as the computing hardware itself. Liquid cooling, however, offers a potential reduction in energy consumption—up to 10 times more efficient—by transferring heat more effectively. "With liquid cooling technology, you can significantly reduce by as much as 10X the amount of electricity that goes towards cooling," emphasized Tuteja, suggesting that this efficiency will allow more resources to be directed towards powering AI operations.
The demand for liquid cooling solutions has already seen substantial growth, with a reported increase of about 30% year to date. As data centers adjust to the energy requirements of next-generation AI models, advanced thermal management solutions are becoming increasingly essential. Dauvin Peterson, an analyst at 22V Research, highlighted that effective thermal management is key for maximizing AI token production and making sure strong returns on investment for hyperscalers.
The upcoming rollout of new AI server architectures from NVIDIA is expected to further accelerate this trend. The Vera Rubin configurations, set to ship in 2026, will feature integrated liquid cooling systems, underscoring the technology's importance in large-scale AI infrastructure. Peterson remarked, "As a standard for AI factories, it presents a clear and credible demand pull."
Prominent companies are already seizing this emerging market. Carrier, a major HVAC manufacturer, reported a remarkable 500% increase in data-center-related orders, with CEO David Gitlin stating that the company’s backlog fully meets its $1.5 billion sales target for the segment this year. Trane Technologies is also witnessing a significant rise in commercial HVAC demand linked to AI infrastructure, with a backlog increase of $2.7 billion following its acquisition of data-center cooling specialist Stellar Energy.
Infrastructure providers like Vertiv are also reporting impressive earnings. The company recently announced adjusted earnings of $1.17 per share, surpassing analysts' expectations, along with a 30% revenue increase to $2.7 billion. Vertiv’s shares have surged 84% this year, and its price target has been raised by TD Cowen, reflecting strong investor confidence in the company's growth trajectory.
Eaton is experiencing positive results as well, with quarterly earnings boosted by increasing AI-related demand from data center customers. Meanwhile, Ecolab is expanding its presence in the market with a planned $4.8 billion acquisition of liquid-cooling hardware provider CoolIT, signaling broader industry recognition of the importance of efficient cooling solutions.
Several other companies stand to benefit from this trend, including Madison Air, Dover, Chemours, and Solstice Advanced Materials, which provide refrigerants and thermal management systems essential for next-generation cooling infrastructure.
As the AI sector evolves, the emphasis on liquid cooling technologies marks a critical shift in how companies tackle the challenges of power consumption and efficiency. The growth of this sector not only underscores the need for innovative cooling solutions but also indicates a broader change in investment strategies within the AI infrastructure landscape. The effects of this transition will likely reverberate across the industry, influencing how data centers operate in the years ahead.
