AI workloads are introducing unprecedented demands on data center infrastructure. Rack densities exceeding 40–100 kW, the adoption of liquid cooling, and the need for continuous cooling and advanced resiliency are pushing traditional data center design approaches beyond their limits.

To help organizations address these challenges, Uptime Institute has published AI in Practice Paper 1: Design Development and Review. This paper is the first in Uptime Institute’s five‑part series examining the design, delivery, commissioning, and operation of AI‑ready data centers.

Based on Uptime Institute’s global field experience, the paper provides practical advice for owners, operators, designers, and developers planning new AI facilities or adapting existing data centers for AI workloads.

The paper outlines the core design principles required to support high‑density AI infrastructure reliably and at scale, including:

• Power and cooling strategies for extreme rack densities
• Integration of liquid and hybrid cooling solutions
• Design considerations for resiliency, scalability, and maintainability
• Aligning facility design decisions with long‑term operational requirements

The guidance in the paper is backed by AI Infrastructure Advisory — Uptime Institute's independent, vendor-neutral practice. Through design reviews, technical validation, and project lifecycle support, we help organizations apply principles outlined in this paper to projects, reducing execution risk and grounding AI infrastructure decisions on unparalleled insight experience and expertise.

The paper is available now and can be downloaded from Uptime Institute’s website.

Download the AI in Practice Paper 1: Design Development and Review paper

More information about AI Infrastructure Advisory and the full set of the five‑paper series is also available through the same page.