OpenAI Weighs Nvidia-Backed Lease for 10 GW Ohio Data Center Campus

Why It Matters

The deal could lock in massive compute capacity for OpenAI and cement Nvidia’s role as both supplier and financial backer, reshaping AI infrastructure economics and raising strategic considerations for enterprise buyers.

Key Takeaways

• OpenAI may lease 10‑GW Ohio data center, backed by Nvidia.
• Project cost estimated at $500 billion, with 20‑year lease terms.
• Nvidia would supply hardware and guarantee lease and financing.
• First phase slated for 2028; full build could take a decade.
• CIOs urged to diversify compute sources beyond a single vendor.

Pulse Analysis

The reported Ohio lease marks a watershed moment in AI infrastructure financing. By tying a $500 billion, 10‑gigawatt campus to a 20‑year lease, OpenAI secures a dedicated compute pool while offloading upfront capital risk to Nvidia, which will also act as a guarantor. This model mirrors the broader shift toward vertically integrated ecosystems where chipmakers, cloud providers and energy firms co‑invest in capacity to meet exploding demand for large language models. For investors, the arrangement signals confidence in sustained AI growth but also introduces concentration risk around a single hardware platform.

Strategically, the partnership deepens the symbiosis between OpenAI and Nvidia that began with a 2025 agreement promising up to $100 billion of investment per gigawatt deployed. By guaranteeing financing, Nvidia moves beyond a vendor role to become a sponsor, aligning its silicon roadmap with OpenAI’s service roadmap. The Ohio site, slated to start operations in 2028, will likely rely on natural‑gas‑driven power from the DOE‑backed redevelopment of the former Portsmouth plant, underscoring the importance of energy security in AI compute economics. However, the decade‑long build timeline and regulatory hurdles add execution risk that could affect cost forecasts and availability.

Enterprises should watch this development closely. CIOs are cautioned to avoid lock‑in to a single compute ecosystem, negotiating contracts that preserve flexibility to shift workloads to alternative clouds such as AWS or Google Cloud. The scale of the Ohio campus promises lower marginal costs for OpenAI’s services, but the financing structure may impose minimum usage commitments that could affect pricing for downstream customers. Diversifying compute sources and monitoring the rollout timeline will be critical to managing both cost and supply‑chain risk in an era of ever‑growing AI demand.