Daily newsHot NewsHot TopicReleases

OpenAI, Oracle Expand Stargate AI Project to 5GW

OpenAI Stargate Data Center Infrastructure
Massive AI infrastructure powering future technology. Source: Tom’s Hardware

Key Takeaways

  • OpenAI Stargate Expansion: Adding 4.5 GW through Oracle partnership, bringing total capacity to over 5 GW
  • Musk’s xAI Ambition: Plans to deploy 50 million H100-equivalent units within 5 years
  • Massive Scale: Over 1 million GPUs will be operational by year-end, targeting 100x increase
  • Real-World Challenges: $500 billion Stargate project faces delays and internal disagreements

OpenAI Expands AI Infrastructure With Oracle

OpenAI has officially announced plans to expand an additional 4.5 gigawatts (GW) through a strategic partnership with Oracle to accelerate the future development of artificial intelligence.

This agreement is part of OpenAI’s long-term vision to deploy 10 gigawatts of computing capacity across the United States. The project will supplement the existing Stargate I facility in Abilene, Texas, bringing the total scale far beyond the initial commitment announced at the White House in January.

Oracle’s advanced cloud computing infrastructure. Source: Dgtl Infra

OpenAI CEO Sam Altman described this as “a massive infrastructure project.” He confirmed that over one million GPUs will be operational by the end of this year, and set an ambitious goal of increasing that number by 100 times.

The additional agreement with Oracle will bring Stargate’s total capacity to over 5 GW, enough to power more than two million AI chips. Altman emphasized: “We are planning to significantly expand Stargate’s ambitions, far beyond the $500 billion commitment we announced.”

Elon Musk’s Bold Plans for xAI

In response to OpenAI’s announcement, Elon Musk revealed an ambitious plan for his AI company xAI. “The goal of xAI is to bring 50 million H100-equivalent AI computing units online within 5 years,” the billionaire declared.

According to estimates, this number would represent 500 times the computing power of the world’s most powerful AI supercomputer just one year ago.

NVIDIA H100 AI Computing Chips
H100 GPUs driving the AI revolution forward. Source: NVIDIA

xAI’s Colossus 2 supercomputer, expected to come online soon, will use 550,000 GB200 chips, equivalent to approximately 5.5 million H100 chips. If Musk’s plan becomes reality, this scale would increase nearly 10-fold from current numbers.

Challenges Facing the $500 Billion Stargate Project

Earlier this year, US President Donald Trump officially launched Stargate – a $500 billion AI infrastructure initiative led by the private sector. The project is backed by OpenAI, SoftBank, and Oracle, with the goal of building AI data centers across the United States and creating over 100,000 jobs.

However, according to recent reports, this initiative has faced significant delays and internal disagreements between key partners SoftBank and OpenAI.

Despite initial promises to deploy $100 billion immediately, the project has had to scale back its near-term goals to building a single data center by year-end. This highlights the real-world challenges in deploying large-scale AI projects, particularly in terms of financing and coordination between stakeholders.

You have not selected any currencies to display