One real facility — a 2 MW GPU training hall for a mid-size AI infrastructure operator — carried through all 6 stages of data center engineering, from cooling topology selection to a completed cybersecurity review.
6 steps. Steps 1-2 are free. Steps 3-6 unlock with a one-time purchase.
Choosing the air/liquid cooling split and CDU architecture for a 2 MW GPU hall.
Utility service through PDU to rack — voltage levels, busway vs. whip, and distribution architecture.
Sizing UPS capacity and standby generation for the facility's critical load and redundancy target.
N+1 vs. 2N power and cooling redundancy decisions and their cost/availability tradeoff for this facility.
DCIM monitoring points, BMS integration, and alarm response for power and cooling systems.
Securing the facility OT network — BMS/DCIM segmentation, remote access, and vendor risk.
Educational project reference. This project walks through a realistic scenario using the site's real calculators and written guides — it is a learning and reference resource, not a permit-ready construction document. Final system design, code compliance, and installation require a licensed engineer or contractor. See our full disclaimer.