Key Product
Liquid cooling, UPS, PDU systems
Bottleneck Status
18-month backlog on liquid cooling
Lead Time
18–24 months
Full briefing▼ Expand
Vertiv provides the power and thermal management infrastructure that makes AI data centers physically possible. As GPU power density climbs — an NVIDIA H100 server rack can consume 70–100kW, compared to 10–15kW for a traditional rack — conventional air cooling becomes inadequate, and liquid cooling becomes mandatory. Vertiv's direct liquid cooling (DLC) systems, uninterruptible power supplies (UPS), and power distribution units (PDU) are deployed in every major AI data center. The company has an 18-month backlog on liquid cooling systems, making it a genuine bottleneck in the AI infrastructure buildout. Vertiv's revenue has grown rapidly alongside the AI data center boom, and its order book gives it unusually high visibility into future data center investment.
Connected companies
Tap a chip to trace that company's chain.
Critical path — raw silicon to deployment
The tightest single-source dependencies, in order.
QWhat does Vertiv make?
Critical power and thermal management for AI data centers
Key products Liquid cooling, UPS, PDU systems
▲18-month backlog on liquid cooling