Today for AI
HOT RADAR
computeHEAT 11.8°

AI CLUSTERED EVENT · 10/13/2026

Bloomberg Intelligence Report: Building a 1GW AI Data Center Costs $30.7 Billion with Chips Over Half

1 reports archived1 independent sourcesupdated 10/13/2026, 12:04:16
Synthesis & Latest Updates
1 Sources Cross-Validated

A new Bloomberg Intelligence report estimates the total cost of building a 1GW AI data center in Abilene, Texas at $30.7 billion. Core chip and server hardware accounts for over half of this expenditure at $17.1 billion, followed by power infrastructure ($3 billion) and cooling systems ($1.66 billion). This breakdown quantifies the capital intensity of large-scale AI compute clusters, highlighting hardware procurement as the dominant cost driver.

LATEST/A new Bloomberg Intelligence report estimates the total cost of building a 1GW AI data center in Abilene, Texas at $30.7 billion. Core chip and server hardware accounts for over half of this expenditure at $17.1 billion, followed by power infrastructure ($3 billion) and cooling systems ($1.66 billion). This breakdown quantifies the capital intensity of large-scale AI compute clusters, highlighting hardware procurement as the dominant cost driver.

TIMELINECoverage timeline

Total 1 reports · Latest first
  1. IT之家 · 智能时代T2·68 pts
    • The total construction cost for a 1GW AI data center is estimated at $30.7 billion.
    • Chips and server hardware represent the largest single expense at $17.1 billion, accounting for over 55% of the total.
    • Power supply infrastructure ($3 billion) and cooling systems ($1.66 billion) constitute the primary non-computing hard costs.
    • Land acquisition and design/engineering costs are relatively minor compared to hardware and power infrastructure.