Liquid Cooling & the AI Thermal LeadNarrow moat

The single most powerful force behind Vertiv's transformation is the shift to liquid cooling for artificial intelligence, and it is here that the company's specialization pays off most directly. The chips that train modern AI — the dense clusters of GPUs — consume and dissipate so much power that traditional air cooling simply cannot keep up. A rack of AI accelerators can draw many times the power of a conventional server rack, and the only practical way to carry away that much concentrated heat is with liquid, piped directly to the chips or immersing them entirely. This is not an incremental change; it is a generational shift in how data centers are cooled, and it plays squarely to Vertiv's greatest engineering strength.

Vertiv is a leader in the data-center liquid-cooling market — at or near the top of the field — and its cooling revenue has been growing explosively as AI density climbs, with liquid-cooling sales at times more than doubling year over year and the segment expected to compound at very high rates for years. The company offers the full range of liquid approaches: direct-to-chip cold plates that bring coolant to the hottest components, the coolant-distribution units that manage the liquid loops, and immersion systems that submerge the hardware. Crucially, Vertiv has worked to get itself designed into the reference architectures that the chipmakers and system builders publish — so that when a customer builds an AI cluster to a standard blueprint, Vertiv's cooling is part of the recommended design.

Why is this a source of moat and not just a source of growth? Because thermal management for AI is genuinely hard, high-stakes engineering, and getting it wrong is catastrophic — inadequate cooling throttles or destroys millions of dollars of chips. Customers want a proven partner with the engineering depth, the reference-design credibility, and above all the ability to deliver and install cooling capacity at scale and at speed, which is the industry's binding constraint. Vertiv's lead in the technology, combined with its manufacturing scale and its ability to pair cooling with the rest of the stack, is a real advantage that is compounding as the installed base of liquid-cooled AI infrastructure grows.

The honest reasons this is a narrow rather than a wide moat run through the accompanying threats. Cooling is a fragmented, fiercely contested market — Schneider Electric, specialist thermal firms, and a rush of new entrants all want this business, and the leader's share is measured in the low double digits, not in dominance. And the chipmakers, by defining the reference cooling designs, hold real power to standardize and potentially commoditize parts of the stack over time. Vertiv's liquid-cooling lead is a genuine, valuable, front-of-the-pack position in the most important growth market in infrastructure — but it is a lead to be defended by relentless engineering and execution, not a locked-in monopoly, and that is exactly why it anchors a strong narrow moat rather than a wide one — even as it powers guidance of ~$14B for FY2026¹.