Data centers reshape energy markets as AI power demand surges

Telborg · October 22, 2025 · ✓ verified

Dell Technologies outlines how data centers are reshaping energy markets due to AI-driven power demand.

  • Main announcement/action: Dell highlights DOE projections that data centers could consume up to 12% of the U.S. electricity grid by 2028, with total data center electricity usage increasing between 325 and 580 TWh by 2028; the piece cites industry estimates of ~70% of future server deployments optimized for high-density GPUs through 2030, a projected 6x server power consumption increase and 200% rise in storage power demand by 2028. The article promotes Dell innovations — ~65% performance-per-watt improvement over two product generations, a PowerCool eRDHx design claiming it can capture up to 100% of the heat and deliver up to a 60% reduction in cooling energy costs, and notes interest in Small Modular Reactors (SMRs) and renewables for consistent power.
  • Background and implementation details: The article describes Dell’s operational tools — Concept Astro (agentic AI, digital twins, automation) and iDRAC10 telemetry — to enable grid-aware workload scheduling by pulling real-time grid data and forecasting optimal execution windows; it references an NREL-produced map for the DOE Grid Deployment Office and third-party sources (Forbes, WRI, SPEC benchmarks) for technical and market context.
Keep reading
Investor Tour 2026 showcases Armenia's AI startup ecosystem Darpass · Nov 19 Vodafone Business and Cassava launch Egypt AI factory APO Group - Africa · Sep 08 ORNL algorithm weighs community factors for urban energy siting Oak Ridge National Laboratory · Sep 08 Gorilla Technology begins Yotta AI infrastructure delivery in India Gorilla Technology · Sep 08
Telborg · US Data Centers
Track the US data-center buildout — every day.

Real-time verified news and daily AI-written briefings, built from primary sources — power, grid, permits, land, financing. Start free.

Get Telborg Pro · $189/mo Get the daily briefing — free →

Every field traced to a primary source.