AI Datacenter Directory

The comprehensive resource for evaluating AI datacenter site selection across power, water, and edge readiness dimensions.

Why Site Selection Matters for AI Infrastructure

Selecting the right site for an AI datacenter is among the highest-stakes decisions in infrastructure investment today. A single bad site selection can expose operators to $50 million to $500 million in hidden costs over the project lifecycle — from interconnection queue delays and curtailment events to water scarcity shutdowns and inadequate edge connectivity.

The GLRI.io Directory provides broad infrastructure context for AI datacenter screening in the United States. We aggregate live data from six major grid operators (ERCOT, PJM, CAISO, MISO, SPP, and NYISO), water authority reports, and Internet Exchange Point (IXP) metrics to create a holistic readiness assessment for launch-market county and metro screening.

Our proprietary scoring methodology — IAIRS (Integrated AI Infrastructure Readiness Score) — evaluates sites across three critical dimensions: Power Readiness (based on Time-to-Power and Curtailment Stress Scores), Water Risk (composite drought and permit friction metrics), and Edge Connectivity (IXP proximity, fiber density, and latency profiles). Each dimension is independently scored and then weighted to produce a composite assessment that correlates strongly with project success rates.

The directory serves as your infrastructure context layer. Browse individual dimensions (Power, Water, Edge), use methodology-backed interpretation, and launch into tools that model specific scenarios.

Start with the dimension most critical to your project:

  • Power-constrained projects — Begin with /directory/power to analyze interconnection queues and TTPS scores
  • Water-sensitive deployments — Start at /directory/water to screen for drought risk and permit timelines
  • Latency-critical workloads — Explore /directory/edge for IXP proximity andcolo pricing