The chipmaker is also investing $1.5 billion in developer SB Energy, securing exclusive access to an enormous data-center project that could generate as much as $200 billion in Nvidia revenue.
MARKET INSIDER — Nvidia has agreed to provide up to $105 billion in guarantees supporting OpenAI’s 20-year lease of a giant artificial-intelligence data center in Ohio, significantly expanding the chipmaker’s financial involvement in infrastructure built to use its processors.
The commitment does not represent an immediate $105 billion cash investment. Instead, Nvidia is providing conditional support for portions of the project’s lease and power obligations and guaranteeing a minimum residual value for the site.
Nvidia will also invest $1.5 billion in developer SB Energy and become the exclusive AI-computing infrastructure provider for the facility.
Key Highlights
- Nvidia will provide up to $105 billion in conditional credit support and invest $1.5 billion in SoftBank-owned SB Energy.
- OpenAI has agreed to lease as much as eight gigawatts of computing capacity at Ohio’s PORTS-Pike campus for 20 years, with the first 800 megawatts expected in 2028.
- The structure could generate up to $200 billion in Nvidia revenue from the initial project but increases its exposure to OpenAI and the broader AI-investment cycle.
How the $105 billion guarantee works
SB Energy will build, own and operate the PORTS-Pike Technology Campus in Pike County, Ohio. OpenAI will lease the data-center capacity, while Nvidia will supply the chips, networking equipment and supporting computing systems.
Nvidia’s guarantee covers parts of the land, power and building-shell obligations associated with the initial 4.25 gigawatts of computing capacity. The company also has an option covering the remaining 3.75 gigawatts, bringing OpenAI’s potential total to approximately eight gigawatts.
The guarantee is not equivalent to Nvidia funding the entire construction cost or assuming all of OpenAI’s lease obligations.
OpenAI is responsible for making rental payments. If it defaults, Nvidia would cover the difference between the project’s guaranteed minimum value and the amount SB Energy could recover by selling the property or leasing it to another customer.
This structure allows SB Energy to seek financing with a stronger credit backstop. The project is expected to combine equity with project-finance loans and potentially bonds, although the final capital structure has not been disclosed, according to Reuters
Nvidia secures long-term demand for its chips
Nvidia will be the exclusive computing-infrastructure provider at PORTS-Pike. The facility will use its DSX AI factory platform, including Nvidia GPUs, CPUs and networking products.
The commercial logic is clear: Nvidia is helping make the project financeable while securing a huge potential customer for multiple generations of its hardware.
CEO Jensen Huang said the initial 4.25-gigawatt deployment could generate as much as $200 billion in Nvidia revenue. Including the proposed Ohio expansion and other OpenAI infrastructure, Nvidia estimates it could generate as much as $600 billion from supplying OpenAI with 16 gigawatts of computing capacity by 2030.
Those figures represent potential revenue rather than firm profit forecasts. Actual sales will depend on construction schedules, OpenAI’s demand, financing availability and the pace at which new generations of Nvidia systems replace earlier equipment.
Nvidia argues that securing long-lived sites allows processors and networking hardware to be upgraded repeatedly without having to obtain new land, grid connections and planning approvals for every product cycle.
Circular-financing concerns remain
The arrangement will nevertheless intensify investor scrutiny of financial relationships within the AI industry.
Nvidia is investing in the developer, guaranteeing obligations connected to the tenant and supplying the equipment that the tenant will install. Supporting the infrastructure therefore helps create demand for Nvidia’s own products.
Huang rejected the description of the transaction as circular financing, arguing that Nvidia is using its financial strength and visibility into future computing demand to unlock scarce infrastructure.
The distinction will ultimately depend on the project’s economics.
If OpenAI generates sufficient revenue from AI services to meet its lease and computing obligations, Nvidia’s guarantee may never result in a meaningful cash payment. It would instead function as credit support enabling construction and future hardware sales.
If OpenAI cannot meet those obligations or demand for AI computing disappoints, Nvidia could face losses under the guarantee while also experiencing weaker chip sales. The same economic risk could therefore affect both sides of its involvement.
Nvidia recently joined BlackRock and other financial institutions in launching financing platforms targeting more than $500 billion in third-party capital for AI infrastructure. The strategy reflects both the scale of demand and the difficulty of funding data centers through technology companies’ balance sheets alone.
Power is becoming the principal constraint
The project’s scale demonstrates that electricity, land and grid access are becoming as strategically important as processors.
OpenAI has agreed to secure approximately eight gigawatts of IT capacity at PORTS-Pike. The first 800 megawatts are expected to begin operating in 2028, with the wider campus built out through 2032.
SB Energy and SoftBank plan to construct at least 10 gigawatts of new power generation. They will also invest $4.2 billion in regional grid infrastructure through a partnership with AEP Ohio.
OpenAI said the project would pay its own energy and infrastructure costs rather than shifting them to local electricity customers. The data center will use closed-loop, air-cooled systems designed to limit continuing water consumption, according to OpenAI’s project announcement
These commitments address two increasingly important obstacles to U.S. data-center development: concerns that AI facilities will raise household electricity prices and opposition to their water consumption.
The project includes federal land previously associated with the Portsmouth uranium-enrichment complex and involves the U.S. departments of Energy and Commerce. Existing infrastructure and government support may reduce the risk of local permitting delays.
Economic opportunity for Ohio
OpenAI expects construction to create approximately 35,000 jobs during the six-year development period through 2032. The completed campus is projected to support around 2,500 long-term operating positions.
OpenAI and SB Energy have each committed $40 million to community programs, creating an $80 million benefits fund covering areas such as education, healthcare, utilities, housing and workforce training.
OpenAI also announced $84 million in Codex credits through ChatGPT for Ohio college students.
The project could revitalize an area previously associated with energy-intensive industrial activity. However, the distinction between temporary construction positions and permanent employment remains important: only a fraction of the headline job total will continue after the campus is completed.
What investors should watch
The first test is whether the initial 800-megawatt phase begins operating on schedule in 2028. Large data centers can be delayed by transmission constraints, equipment shortages and construction complexity even when community support is relatively strong.
Investors should also examine Nvidia’s future disclosures for the accounting treatment, duration and activation conditions of the guarantee. The $105 billion headline represents maximum potential exposure, not an expected cash outlay, but it is too large to ignore.
OpenAI’s financial performance will become increasingly relevant to Nvidia. The chipmaker is no longer exposed only through hardware sales; it is also supporting infrastructure whose economics depend on OpenAI’s ability to commercialize that computing capacity.
The transaction may give Nvidia exceptional visibility into future demand. It also concentrates more financial risk among a small group of interdependent companies.
The decisive question is whether the enormous capital committed to AI infrastructure produces enough recurring revenue to justify its cost. The answer will determine whether Nvidia’s guarantee proves to be an efficient use of its balance sheet—or an expensive mechanism for supporting its largest customers.
