← Back to News
Technology

The Stargate Project: Inside the $500 Billion AI Infrastructure Initiative

In what has been called the largest AI infrastructure project in history, the Stargate Project aims to invest $500 billion in AI data centers across the United States over four years. Announced at the White House on January 21, 2025, just one day after President Trump's inauguration, this joint venture between OpenAI, SoftBank, Oracle, and Abu Dhabi's MGX represents an unprecedented commitment to American AI leadership.

Stargate at a Glance

  • Total investment: $500 billion over 4 years
  • Initial deployment: $100 billion
  • Planned capacity: Nearly 7 GW across 5+ U.S. sites
  • NVIDIA commitment: $100 billion in AI processors
  • Jobs created: 100,000+ American positions
  • Global expansion: UAE, Norway, Argentina

The Vision and Structure

SoftBank CEO Masayoshi Son serves as chairman of Stargate LLC, with SoftBank handling financial responsibility and OpenAI managing operational execution. Key technology partners include Microsoft, NVIDIA, Oracle, Arm, and OpenAI itself.

The project's stated goal is securing American leadership in AI while creating hundreds of thousands of jobs and generating massive economic benefits. President Trump said the initiative would create over 100,000 American jobs almost immediately, while Sam Altman projected hundreds of thousands of positions across the supply chain.

Rapid Expansion

By September 2025, Stargate had exceeded expectations. The flagship campus in Abilene, Texas—internally known as 'Project Ludicrous'—was already operational on Oracle Cloud Infrastructure with NVIDIA GB200 racks. Oracle began delivering the first GB200 racks in June, with plans to install over 64,000 chips by March 2026.

Five new U.S. sites were announced:

  • Shackelford County, Texas
  • Doña Ana County, New Mexico
  • Lordstown, Ohio
  • Milam County, Texas
  • Wisconsin (Oracle/Vantage partnership)
  • Michigan's Saline Township (gigawatt-scale campus)

Global Reach

Stargate UAE: 200 MW by 2026 with potential 1 GW scale. Stargate Norway: Launched July 2025 in Narvik with 230 MW and 100,000 NVIDIA GPUs expected by end of 2026. Stargate Argentina: $25 billion investment with 500 MW capacity—the largest data center in Latin America.

"No one in the history of man built data centers this fast." — Sarah Friar, OpenAI CFO

Hardware and Supply Chain

The scale of hardware commitments is staggering: NVIDIA announced a $100 billion investment to supply AI processors; AMD agreed to provide 6 GW worth of Instinct GPUs; Broadcom will supply 10 GW of custom hardware. Samsung and SK hynix joined in October 2025 to scale advanced memory production.

Comments

Be the first to comment!