In 2026, five US data centers will each cross the 1-gigawatt threshold—consuming as much power as a nuclear reactor. This has never happened before in computing history.
The facilities represent a new species of infrastructure:
Anthropic-Amazon (January 2026) — New Carlisle, Indiana. Already operational with 500,000 AWS Trainium 2 chips, doubling to 1 million by year-end. Computing equivalence: 300,000 H100 GPUs.
xAI Colossus 2 (February 2026) — Memphis, Tennessee. Elon Musk’s facility will deliver computing power equivalent to 1.4 million H100s. The original Colossus became the world’s largest AI training cluster just four months after construction began.
Microsoft Fayetteville (March 2026) — Georgia. Part of Microsoft’s aggressive expansion. Their Wisconsin facility could exceed 3GW by 2028—among the largest computing installations ever constructed.
Meta Prometheus (May 2026) — Ohio. Includes 200MW of behind-the-meter generation—power produced onsite that bypasses the grid entirely.
OpenAI Stargate (July 2026) — Abilene, Texas. Features 10 natural gas turbines for 361MW backup power. OpenAI has signed infrastructure deals totaling $1.4 trillion over eight years.
Combined: 5+ gigawatts, $145 billion invested, 3 million+ GPUs, all coming online within 7 months.
For the full breakdown of Big Tech’s gigawatt-scale AI infrastructure race, read The State of AI Data Centers on The Business Engineer.
Gennaro is the creator of FourWeekMBA, which reached about four million business people, comprising C-level executives, investors, analysts, product managers, and aspiring digital entrepreneurs in 2022 alone | He is also Director of Sales for a high-tech scaleup in the AI Industry | In 2012, Gennaro earned an International MBA with emphasis on Corporate Finance and Business Strategy.
