OpenAI’s Wisconsin AI site moves into construction as its long-term compute plan turns physical
Original: RT @sk7037: Construction is underway at our site in Port Washington, Wisconsin. This is an important step in our long-term compute strategy… View original →
Construction starts to replace roadmap language
On March 6, 2026, OpenAI reposted a message from Sachin Katti stating that construction is underway at the company's site in Port Washington, Wisconsin. The post described the project as an important step in OpenAI's long-term compute strategy and thanked VantageDC and Oracle for helping bring the capacity online. It also linked the Wisconsin work to OpenAI's broader infrastructure push with NVIDIA systems and deeper ecosystem work with AWS.
That matters because the Wisconsin site was not a brand-new idea announced out of nowhere. In September 2025, OpenAI said in its OpenAI, Oracle, and SoftBank expand Stargate with five new AI data center sites update that the previously unnamed Midwest Stargate location would be in Wisconsin and developed by Oracle in partnership with Vantage. In the same announcement, OpenAI said the expanded Stargate footprint would bring planned capacity to nearly 7 gigawatts and more than $400 billion of investment over three years, keeping the company on a path toward a 10-gigawatt commitment.
The March 6 post is therefore important less as a financing announcement and more as an execution signal. It shows at least one of OpenAI's large compute bets has moved from partnership language and site selection into physical buildout. That distinction matters in AI infrastructure because much of the sector's public planning has focused on letters of intent, power availability, and partner capital rather than visible construction milestones.
Why Wisconsin matters in the wider OpenAI stack
The Wisconsin update also fits with OpenAI's wider compute portfolio. OpenAI said in its AWS partnership announcement that it would immediately start using AWS compute, with targeted deployment before the end of 2026 and room to expand beyond that. Separately, OpenAI and NVIDIA said they plan to deploy at least 10 gigawatts of NVIDIA systems for next-generation AI infrastructure, with the first gigawatt scheduled for the second half of 2026 on Vera Rubin.
Taken together, those announcements suggest OpenAI is not depending on a single provider or a single site. Instead, it is building a layered compute portfolio across Stargate facilities, hyperscaler capacity, and specialized system partnerships. The Port Washington construction update is significant because it turns that portfolio strategy into something tangible on the ground.
The March 6 X post does not disclose timelines for bringing the Wisconsin site into service, nor does it attach a fresh capacity number to Port Washington specifically. But it does offer a concrete signal that OpenAI's infrastructure strategy is entering an execution phase. Sources: the original X repost, OpenAI's Stargate site announcement, the AWS partnership post, and the NVIDIA systems partnership post: X, Stargate update, AWS partnership, NVIDIA partnership.
Related Articles
OpenAI announced $110B in new investment on February 27, 2026, alongside Amazon and NVIDIA partnerships aimed at compute scale. The company tied the move to 900M weekly ChatGPT users, 9M paying business users, and rising Codex demand.
NVIDIAAI says it is partnering with Thinking Machines to deploy at least 1 gigawatt of NVIDIA Vera Rubin systems for frontier AI training. Thinking Machines frames the partnership as infrastructure for both frontier model training and platforms for customizable AI.
OpenAI and Amazon announced a multi-year strategic partnership on February 27, 2026. The deal combines a new Amazon Bedrock-based Stateful Runtime Environment, exclusive third-party AWS distribution for OpenAI Frontier, approximately 2 gigawatts of Trainium capacity, and a $50 billion Amazon investment.
Comments (0)
No comments yet. Be the first to comment!