OpenAI has abandoned its plans to rent compute capacity from the Stargate Norway data center, instead opting to work directly with Microsoft for its AI infrastructure needs.
The Shift in AI Compute Strategy
OpenAI has significantly adjusted its plans for accessing compute power in Norway, confirming the withdrawal from a deal with the Stargate facility. This move follows the company's recent decision to pause a similar infrastructure project in the United Kingdom, citing concerns over energy costs and regulatory environments.
Microsoft is now set to take over the compute capacity that was previously earmarked for OpenAI at the planned 230MW "Stargate Norway" facility in Narvik. According to an OpenAI spokesperson, the company is now in discussions to rent capacity directly from Microsoft.
Rationale for Partnering with Microsoft
OpenAI stated that shifting its focus to Microsoft is a more financially sound decision for the company, as it aligns with existing contracted spending agreements. The company emphasized that Microsoft is a crucial partner in its global network.
Key points regarding the partnership include:
- Financial Alignment: The direct rental arrangement with Microsoft is deemed more financially sensible for OpenAI.
- Existing Contracts: The partnership falls under existing contracted spending, referencing OpenAI's prior agreement to purchase $250 billion in services from Microsoft's Azure cloud-computing division.
- Commitment: An OpenAI spokesperson confirmed, "We are moving ahead with our plans in Norway... we will work with them to access compute in Norway just as we already do in other parts of the world."
Microsoft's Expansion in Norway
Microsoft is actively expanding its presence at the Narvik campus, solidifying its role as a major AI infrastructure provider. Nscale, the developer of the Stargate facility, announced that Microsoft is expanding its agreement by deploying over 30,000 Nvidia Rubin GPUs.
Jon Tinter, president of business development and ventures at Microsoft, stated that this expansion helps ensure that Microsoft customers have access to the advanced AI infrastructure needed as demand grows across Europe.
Broader Context of OpenAI's Spending
OpenAI's strategic adjustments reflect a broader effort to manage expectations regarding its massive spending plans, particularly as the company prepares for a potential Initial Public Offering (IPO) later this year.
In recent months, the company has signaled immense future investment in AI infrastructure:
- Future Spending Targets: OpenAI previously told investors it was targeting roughly $600 billion in total compute spend by 2030.
- Historical Commitments: This follows earlier comments from CEO Sam Altman suggesting the company would hit $1.4 trillion in infrastructure commitments over the next eight years.
- Recent Pauses: The company also confirmed halting its U.K. Stargate project and previously shuttered its video generation service, Sora, in March.
