In a bold new vision for the future of artificial intelligence infrastructure, OpenAI CEO Sam Altman announced that the company aims to build an AI factory capable of generating 1 gigawatt of compute power per week.
The announcement, made on October 29, 2025, on X (formerly Twitter), outlines OpenAI’s aggressive expansion strategy to meet the rising demands of large-scale AI models and global adoption. Altman emphasized that this capacity growth will come “at a greatly reduced cost relative to today.”
Currently, OpenAI has committed to approximately 30 gigawatts of compute, with an estimated total cost of ownership around $1.4 trillion. “We are comfortable with this given what we see on the horizon for model capability growth and revenue growth,” Altman said.
This announcement follows a significant corporate restructuring and a deepened partnership with Microsoft, positioning OpenAI for unprecedented scale and potentially paving the way for an eventual IPO.
Key Details of OpenAI’s Compute Expansion Plan
| Metric | Details (as of Oct. 2025) |
|---|---|
| Compute Expansion Goal | 1 gigawatt of new compute capacity per week |
| Current Commitment | 30 gigawatts of compute |
| Estimated Total Investment | $1.4 trillion |
| AI Factory Location Strategy | Multiple global sites under development |
| Partner Companies | Microsoft, Oracle |
| Objective | Reduce compute cost while expanding AI model capability |
| Target Mission | Ensure AGI benefits all humanity |
Altman’s statement reflects OpenAI’s confidence in the future of model capability growth — the rapidly advancing performance of generative and multimodal AI systems that require massive compute power for training and deployment.
OpenAI’s Corporate Restructuring: Foundation and Public Benefit Model
Just a day before Altman’s compute expansion post, OpenAI finalized its long-anticipated corporate restructuring, transforming its for-profit operations into a Public Benefit Corporation (PBC) under a nonprofit foundation.
Altman explained: “We have a non-profit called OpenAI Foundation that governs a Public Benefit Corporation called OpenAI Group. The foundation initially owns 26% of the PBC, but it can increase over time if the PBC does super well.”
This simpler structure, Altman said, will allow OpenAI to attract more investment and talent while maintaining its mission — to ensure that artificial general intelligence (AGI) benefits all humanity.
Under the new framework:
- Microsoft now holds a 27% stake in OpenAI Group, valued at approximately $135 billion.
- Microsoft gains exclusive intellectual property rights to OpenAI technology until 2032.
- OpenAI will purchase $250 billion worth of Microsoft’s cloud computing services.
- Microsoft’s right of first refusal to be OpenAI’s compute provider has ended, allowing OpenAI to diversify partnerships.
- Microsoft can now independently pursue AGI development with other partners.
Analysts believe this new structure may set the stage for an OpenAI IPO, as it enhances transparency and gives investors a clearer ownership model.
Building the Infrastructure for AGI: The “Stargate” Project
In line with its compute expansion plans, OpenAI and Oracle have partnered to build a massive data center campus called Stargate in Port Washington, Wisconsin.
This project, part of a broader $500 billion, 10-gigawatt initiative, is expected to deliver over 4.5 gigawatts of capacity once complete — making it one of the largest AI-focused infrastructure developments in the world.
| Project Name | Stargate Data Center |
|---|---|
| Location | Port Washington, Wisconsin |
| Partners | OpenAI & Oracle |
| Total Investment | $500 billion |
| Target Compute Capacity | 10 gigawatts total |
| Expected Output from Stargate Campus | 4.5 gigawatts |
| Purpose | Support AI model training, deployment, and AGI-scale infrastructure |
“The Stargate campus is a cornerstone in OpenAI’s long-term infrastructure strategy,” said Dr. Karen Meyers, Senior AI Systems Analyst at the Global Compute Research Institute. “It’s a physical manifestation of how computing capacity has become the new oil in the AI economy.”
Expert Commentary on OpenAI’s Ambitious Compute Goals
1. Sam Altman, CEO, OpenAI:
“Our goal is to build an AI factory that generates 1 gigawatt of compute per week — at lower cost and higher efficiency than anything that exists today.”
2. Dr. Karen Meyers, Global Compute Research Institute:
“This marks the dawn of industrial-scale AI infrastructure. Just as the industrial revolution was powered by steam, the AI revolution will be powered by compute.”
3. Michael Andrews, Tech Policy Expert at Stanford:
“OpenAI’s new structure gives it the flexibility to fund large infrastructure projects while still maintaining its social mission. It’s a blueprint for balancing innovation and responsibility.”
4. Ravi Shah, Investment Analyst at FutureTech Capital:
“Altman’s vision is as much about economic power as technological growth. Compute is now a key strategic asset, and OpenAI is ensuring it controls the supply chain that fuels its models.”
Why Compute Expansion Matters for the Future of AI?
AI systems are becoming exponentially more compute-hungry as they grow in size and sophistication. From training trillion-parameter models to deploying real-time AI agents at scale, compute capacity determines both innovation velocity and competitive edge.
Altman’s 1-gigawatt-per-week target would make OpenAI one of the largest consumers — and producers — of computing power globally, rivaling the data center footprints of hyperscalers like Google, Amazon, and Microsoft.
“Compute has become the backbone of AI progress,” said Meyers. “The companies that can scale compute affordably and sustainably will dominate the next era of artificial intelligence.”
Furthermore, OpenAI’s pursuit of reduced cost-per-compute signals a push for energy-efficient hardware and new chip architectures, likely in partnership with both established semiconductor players and new entrants in AI hardware design.
Economic and Environmental Considerations
OpenAI’s expansion plans also raise questions about sustainability and energy sourcing. With 30 gigawatts already committed and an additional 1 gigawatt planned weekly, energy demand from AI infrastructure is reaching record highs.
Experts predict that next-generation AI factories will rely heavily on renewable energy, advanced cooling systems, and carbon offset programs to minimize environmental impact.
“OpenAI’s challenge is not just building compute — it’s building it responsibly,” noted Michael Andrews. “The balance between innovation, energy use, and sustainability will define public sentiment toward AI infrastructure.”
Why This Matters?
OpenAI’s announcement represents more than a scaling plan — it’s a strategic declaration of AI industrialization. As the company evolves into a full-fledged infrastructure powerhouse, it is setting the pace for the entire sector.
With a simplified governance model, strengthened Microsoft partnership, and expanding data center network, OpenAI is positioning itself as the global leader in compute-driven AI innovation — paving the way for artificial general intelligence that is both powerful and publicly accountable.
Frequently Asked Questions
What does “1 gigawatt of compute per week” mean?
It refers to OpenAI’s goal to add new computing power equivalent to 1 gigawatt of energy capacity every week to support AI model training and operations.
How much compute power does OpenAI currently have?
OpenAI has committed to around 30 gigawatts of total compute capacity, with an estimated cost of $1.4 trillion.
What is the purpose of the Stargate project?
Stargate is a $500 billion data center campus co-developed with Oracle, expected to generate 4.5 gigawatts of compute capacity.
How does Microsoft fit into OpenAI’s new structure?
Microsoft holds a 27% stake in OpenAI’s Public Benefit Corporation and retains exclusive rights to OpenAI’s IP until 2032.
What is the mission of OpenAI under its new structure?
To ensure that artificial general intelligence benefits all humanity, while maintaining the flexibility to raise capital and innovate at scale.
Could OpenAI go public soon?
Many analysts believe that OpenAI’s new simplified corporate model could pave the way for a future IPO.