Pune – October 12, 2026 -- The global market for data center racks drawing more than 100 kW is forecast to grow from USD 37.2 billion in 2025 to USD 52.5 billion by 2034, a 3.5% CAGR, according to a new Credence Research Inc. report.
AI training and inference, accelerated computing and high-performance computing (HPC) workloads are driving demand. They require substantially more rack-level power and cooling than conventional enterprise infrastructure. Direct-to-chip liquid cooling, rear-door heat exchangers, immersion-ready enclosures and higher-capacity power distribution are becoming core design considerations for new deployments.
North America holds 36% of the market, ahead of Asia Pacific at 30%
North America's lead reflects hyperscale concentration, cloud investment and large AI infrastructure programs. Asia Pacific is supported by data center construction and sovereign-compute investment across China, Japan, South Korea and India. Europe holds 22%, with demand shaped by HPC, sovereign AI, energy efficiency and regulated digital infrastructure. Latin America and Middle East & Africa are smaller but benefit from cloud expansion and new data center projects.
The 100-200 kW segment dominates deployments while above-200 kW designs emerge
The 100-200 kW per rack range is the broadest deployment band and is increasingly practical for commercial AI infrastructure. Systems above 200 kW are gaining importance for advanced AI and HPC clusters and next-generation rack-scale systems. That shift raises demand for integrated rack, manifold, coolant distribution unit (CDU) and power architectures. Air cooling is becoming insufficient for the highest-density accelerator platforms.
Hyperscalers and cloud providers remain the largest end-user group. Colocation operators, enterprise AI and sovereign-compute programs are broadening demand. Operators weigh OCP Open Rack v3, EIA-310 19-inch and custom formats, balancing interoperability against workload-specific design.
Grid capacity and cooling-water supply delay deployments even when hardware is available
The report identifies facility limits, not rack space, as the largest constraint. Grid constraints, transformer capacity and cooling-water availability can hold up projects. Liquid cooling adds operational requirements around coolant quality, manifolds, CDUs, leak detection and service procedures. Operators must also coordinate rack architecture with facility systems and server warranties.
Credence expects high-density deployments to be sold increasingly as integrated infrastructure ecosystems rather than standalone racks. Factory-integrated modules combining liquid cooling, monitoring and power systems can reduce commissioning risk and shorten deployment timelines.
Vertiv and Schneider Electric expand liquid-cooling lineups for 100 kW-plus racks
The report names Vertiv, Schneider Electric, Eaton, Rittal, Dell Technologies, Supermicro, Hewlett Packard Enterprise, NVIDIA, nVent and CoolIT Systems as key players. Buyers increasingly weigh total lifecycle value and supplier support alongside headline pricing.
- January 2026: Vertiv introduced new MegaMod HDX configurations supporting rack densities from 50 kW to more than 100 kW and capacity up to 10 MW, combining direct-to-chip liquid cooling with air-cooled infrastructure.
- September 2025: Schneider Electric unveiled an end-to-end liquid-cooling portfolio with Motivair, including CDUs, rear-door heat exchangers and cold plates for 100 kW-plus AI racks.
- February 2026: Schneider Electric opened a Motivair liquid-cooling factory in Bengaluru, making India a manufacturing and export hub for high-density AI cooling systems.
- Rittal reported productive use of a water-based direct liquid cooling system at the Green IT Cube in Darmstadt and highlighted the shift toward rack densities above 150 kW.