NVIDIA is no longer just selling chips; it's now guaranteeing the physical infrastructure that powers artificial intelligence at scale. The company announced a sweeping partnership with OpenAI and SoftBank to develop an 8-gigawatt AI data center campus in Pike County, Ohio, with NVIDIA investing $1.5 billion directly into the project and securing exclusive control over the computing hardware deployed there.

This deal represents a fundamental shift in how the AI industry builds its foundation. Rather than waiting for customers to find their own land and power, NVIDIA is taking on the role of infrastructure guarantor, positioning itself as a stakeholder in the long-term physical assets that underpin AI development. The move signals that in the AI era, controlling access to computing capacity means controlling access to land, electricity, and the buildings themselves.

What Makes This Ohio Campus Different From Other Data Centers?

The PORTS-Pike Technology Campus, being developed on and around the decommissioned Portsmouth Gaseous Diffusion Plant in Appalachian Ohio, is designed to operate at unprecedented scale. OpenAI will be the sole customer for the full 8 IT-gigawatt (IT-GW) capacity, operating under a 20-year lease with SB Energy, the developer that will build, own, and operate the facilities.

The project is structured in phases, with capacity expected to come online beginning in 2028. NVIDIA will initially secure 4.25 IT-GW of the campus for its DSX AI factory platform, with an option to extend into the remaining 3.75 IT-GW. This architecture includes GPUs, CPUs, and networking hardware designed to be upgraded with each new generation of chips, allowing customers to repeatedly refresh their computing power without rebuilding facilities from scratch.

To power this massive deployment, SB Energy and SoftBank have committed to building at least 10 gigawatts of new energy generation capacity. The partners will also invest at least $4.2 billion in new regional grid infrastructure through a partnership with AEP Ohio, designed to protect existing ratepayers from bearing the cost of AI infrastructure buildout.

How Are Tech Giants Reshaping the Data Center Market?

The Ohio deal is part of a broader trend where hyperscalers and AI companies are moving beyond simply renting computing capacity and instead securing long-term control over the physical assets that support their operations. NVIDIA founder and CEO Jensen Huang framed the investment as essential to the company's future, stating that land, power, and shell capacity have become vital assets in the AI era.

This strategy extends beyond NVIDIA. Neocloud providers like Iren, which both sell to and compete with major cloud companies, are also expanding aggressively. Iren recently announced it had delivered the first of four 50-megawatt clusters in Childress, Texas, to Microsoft under a $9.7 billion contract. The company is building 4.35 gigawatts of combined capacity in Texas and Oklahoma over the next three years, compared to just 160 megawatts across its three current Canadian data centers.

Iren's shift from Canada to the United States reflects a critical challenge facing AI infrastructure development: the need for abundant, affordable power and permitting speed. Kent Draper, Iren's chief commercial officer, explained that the expansion potential in British Columbia is limited by the smaller power grid, and Canada's regulatory environment is more challenging than across the border.

Steps to Understanding AI Infrastructure Investment Priorities

• Power Availability: Companies prioritize regions with access to large amounts of cheap electricity, whether from renewable sources, natural gas, or nuclear facilities, since data centers consume enormous amounts of energy to run and cool computing hardware.
• Permitting Speed: Jurisdictions with streamlined approval processes and fewer zoning restrictions attract investment faster; many unincorporated communities in Oklahoma and Texas lack zoning rules, accelerating project timelines.
• Long-Term Stability: Investors seek regulatory certainty and stable government relationships, as AI data centers require 20 to 30-year commitments to justify the billions spent on construction and infrastructure.
• Grid Capacity: New facilities must connect to power grids with sufficient capacity to handle their demand, or developers must invest in building new generation and transmission infrastructure alongside the data center itself.

The Ohio project includes a significant community investment component. OpenAI has agreed to contribute $40 million to a community benefits fund, bringing the total to $80 million alongside SB Energy's initial pledge. The fund is designated to support affordable energy, job creation, workforce development, and broader economic development in Pike County.

Why Are Investors Concerned About AI Infrastructure Valuations?

While companies like NVIDIA and OpenAI are betting billions on AI infrastructure, some of the world's largest investors are sounding alarms about whether these investments will deliver promised returns. Nicolai Tangen, CEO of Norway's Government Pension Fund Global, the world's largest sovereign wealth fund managing $2.4 trillion, warned that AI stock valuations now pose a serious risk to global markets.

Tangen cautioned that in an extreme market collapse, a massive loss to the fund's portfolio is "not completely improbable." The fund, created to invest Norway's vast oil and gas revenues, delivered a record profit of 186 billion dollars in the first six months of 2026, with much of those gains driven by AI-chip stocks. However, the CEO warned that a sharp correction in AI valuations could potentially erase much of the massive wealth built up over the past 30 years.

Tangen

"Few managers are inclined to take profits when such a fundamental technology buildout, fueled by literally unprecedented levels of capital spending, shows little evidence of brittleness," said Bill Megginson, a leading researcher on sovereign wealth funds.

Bill Megginson, Finance Professor at the University of Oklahoma

The Bank for International Settlements warned in June that AI "exuberance" risks ending in a bust if returns fall short of expectations. Major technology companies are expected to invest more than $1 trillion in AI-related infrastructure like chips, data centers, and power infrastructure in the race to match or beat human intelligence.

Norway's fund is particularly vulnerable because it follows a passive, benchmark-based investment strategy that tracks global markets, with technology accounting for roughly one-third of its stock investments. Unlike other sovereign wealth funds that can hedge against downturns by holding cash or buying protective options, Norway's strict government mandate prevents such defensive positions. Stress tests conducted by Norges Bank Investment Management, the unit that manages the fund, estimate that an AI-driven market correction could reduce the fund's value by around 18 percent, equivalent to roughly $432 billion.

Despite these warnings, most institutional investors continue to fund AI infrastructure buildout. The reasoning is straightforward: the fundamental technology buildout is real, capital spending is at unprecedented levels, and there is little evidence the infrastructure itself is fragile. The question investors face is whether the financial returns will eventually justify the enormous upfront costs.

The Ohio campus represents one answer to that question. By securing long-term customer commitments, guaranteeing power supply, and positioning itself as a stakeholder in the physical infrastructure, NVIDIA is betting that the demand for AI computing power will remain strong enough to justify the investment. Whether that bet pays off will help determine whether AI infrastructure represents a genuine economic transformation or a speculative bubble waiting to burst.