GW

GLOBAL AI INFRASTRUCTURE FORENSIC RESEARCH PORTAL

Cutoff: Nov 14, 2026

36 Physical Campuses (49.45 GW Unique / 71.44 GW Gross) • 21 Companies • NVIDIA 10-Q Note 10 Audit • 2027–2030 Stress Engine

Active Scenario:

Page 1 — Executive Dashboard: Global AI Infrastructure Outlook (2026–2030)

Comprehensive reconciliation of gross announcements vs. unique physical GW, scenario-adjusted completions, capex funding gaps, and three independent demand measures.

Quick Drilldown:

Annual New Capacity Additions (2027–2030, GW/yr)

Planned Unique Announcements vs. Active Scenario Completions, Delays & Cancellations

Baseline

Cumulative Operational AI Capacity vs. 3 Demand Measures (GW)

Separates Cumulative Supply from Contracted, Utilized & Independently Monetized Demand

Companies Accounting for Largest 2027–2030 GW Reductions (Active Scenario vs. Announced)

Click any company row to inspect its project-by-project pipeline, debt wall, and SEC citations

Company Category 2026 Live '27–'30 Plan Scenario Add GW Reduction Unfunded Capex Action

Four-Quadrant Risk Migration (36 Physical Projects)

Project count and unbuilt 2027–2030 GW across Baseline, NV Withdrawal & Bear

Quadrant Baseline NV Withdraw Severe Bear Active Modeled
Key Structural Finding: Under the Severe Neocloud Collapse scenario, 19.99 GW of unbuilt capacity (20 projects) migrates into Quadrant 4 (Weak Financing / Weak Demand), while 3.97 GW in Quadrant 2 undergoes financial restructuring or hyperscaler step-in rescue.