Updated
Updated · Yahoo Finance · Jul 29
Jensen Huang Says AI Needs 1,000x More Power as US Data Center Demand Hits 66 GW by 2027
Updated
Updated · Yahoo Finance · Jul 29

Jensen Huang Says AI Needs 1,000x More Power as US Data Center Demand Hits 66 GW by 2027

3 articles · Updated · Yahoo Finance · Jul 29

Summary

  • Jensen Huang told a Stanford class in May that AI computing will likely need 1,000 times more power than is available now, framing the scale of future energy demand.
  • Goldman Sachs projects U.S. data-center electricity demand will more than double to 66 gigawatts in 2027 from 31 GW in 2025, lifting the sector's share of national power use above 8% from just over 4%.
  • Bloom Energy estimates AI could consume as much as 12% of U.S. electricity by 2030, suggesting the current buildout would still be an early stage.
  • GE Vernova is positioned to benefit because its natural-gas turbines can power data centers on-site, avoiding long grid-connection delays and matching operators' need for scalable generation.

Insights

Will AI’s power hunger make GE Vernova the real winner of the next computing boom, or will grid delays and emissions concerns slow it down?
As AI electricity demand surges, which matters more now for hyperscalers: access to chips, or access to reliable megawatts?
If data centers can’t wait for the grid, are on-site gas turbines the only practical answer—or just a costly stopgap?