AI Data Center Engineering StudioPower, cooling, and network design for GPU-scale AI infrastructure

Calculate Power Usage Effectiveness from IT/GPU load, cooling power, distribution losses, and auxiliary loads — with DCiE and a benchmark rating against typical AI/GPU facility PUE.

Estimate kW/rack density from GPU server presets (H100, GB200 NVL, A100) or custom nameplate wattage, and check the load against PDU/circuit capacity per NEC continuous-load rules.

Convert IT/GPU electrical load into total cooling tons and BTU/hr, split between liquid and air cooling, and estimate chilled-water flow rate for CDU/CRAH sizing.

Why a 10 kW GPU server can't simply be cooled by more fans — air runs out of heat-carrying capacity long before GPU racks do, which is why liquid cooling went from a niche HPC technique to the default AI infrastructure architecture.

A data center can post a great PUE while consuming millions of gallons of water a year — Power Usage Effectiveness and Water Usage Effectiveness measure two different resources, and optimizing one can work against the other.