Data Center Startup Achieves Rapid AI Deployment Using Brownfield Strategy

Brownfield Approach Accelerates AI Infrastructure Deployment

In the race to establish artificial intelligence computing capacity, one data center startup has demonstrated how brownfield development can dramatically reduce deployment timelines, according to industry reports. Sources indicate that Patmos successfully established a functional AI data center within just 90 days by repurposing an existing industrial facility rather than constructing from scratch.

Brownfield Approach Accelerates AI Infrastructure Deployment
The Time Advantage of Existing Infrastructure
Advanced Cooling for High-Density AI Workloads
Sophisticated Fluid Management System
Heat Exchange Architecture

The Time Advantage of Existing Infrastructure

The brownfield versus greenfield debate in data center development has gained new urgency as demand for AI computing surges, analysts suggest. Company representatives reportedly emphasized that existing buildings offer significant time savings because they already contain critical infrastructure components. “On a brownfield, I can go into a building that’s already got power coming into the building,” Morgan of Patmos stated in the report. “Sometimes they’ve already got chiller plants, like what we’ve got with the building I’m in right now.”

The company specifically selected a former printing press building that provided established utility connections and structural elements that could support high-density computing requirements, the report states. This approach reportedly allowed the company to bypass months of construction and permitting processes typically associated with greenfield development.

Advanced Cooling for High-Density AI Workloads

To support the intensive thermal demands of AI processors, the facility implemented sophisticated liquid cooling systems capable of handling unprecedented power densities, according to technical documentation. The data center reportedly supports between 50 to over 140 kilowatts per cabinet, far exceeding the 1-2 kW densities common in traditional data centers.

The system specifically accommodates Nvidia’s Grace Blackwell processors, which sources indicate generate exceptional heat output. To manage this thermal load, engineers designed a multi-loop liquid cooling system that separates water sources into distinct, closed loops. This architecture ensures that municipal water never directly contacts sensitive IT equipment, the report confirms.

Sophisticated Fluid Management System

“We have five different, completely separated water loops in this building,” Morgan stated in the report. The cooling tower reportedly uses city water exclusively for evaporation, while separate closed loops serve the data hall with specialized coolant. This segregation prevents contamination and corrosion while maximizing thermal efficiency.

The facility taps into Kansas City’s district chilled water supply for primary cooling, sourced from a nearby utility plant. Within the data center, a dedicated loop circulates a glycol-based fluid that is filtered to extremely low micron levels and formulated to be electronically safe, according to technical specifications.

Heat Exchange Architecture

Heat exchangers transfer thermal energy from the data hall fluid to the district chilled water while maintaining complete separation between the two systems. The implementation includes both liquid-to-chip and rear-door heat exchangers for immediate heat removal at the source, the report states. This approach reportedly allows the facility to maintain optimal operating temperatures for the high-performance computing hardware despite the extraordinary heat generation.

Industry analysts suggest that such cooling innovations will become increasingly critical as AI computing densities continue to escalate beyond traditional air cooling capabilities. The successful implementation in a converted industrial space demonstrates that existing structures can be adapted to support even the most demanding contemporary computing requirements when approached with creative engineering solutions.

Reference: Brownfield land development involves reusing existing structures or previously developed sites, while data centers house computer systems and associated components. Advanced computer cooling techniques are essential for managing heat from high-performance processors.