NVIDIA is pushing its competitive frontier beyond chips and servers, venturing deeper into the land, power, and buildings that form the foundational layer of AI data centers.

On August 17, NVIDIA announced a strategic partnership with SB Energy, providing credit support for the PORTS-Pike technology campus in Ohio to transform it into a hyperscale data center dedicated exclusively to NVIDIA's AI computing infrastructure, with OpenAI serving as the sole tenant. Simultaneously, NVIDIA is investing $1.5 billion in SB Energy.

By directly entering the land, power, and shell (LPS) segment of underlying infrastructure, NVIDIA's move carries significance that extends well beyond a typical data center investment. The core shift lies in NVIDIA leveraging its own credit to underwrite AI infrastructure financing, in exchange for long-term, exclusive space to deploy its computing power. This marks the chip giant's evolution from a pure AI hardware supplier into a central participant and organizer of "AI factory" infrastructure.

The project's initial deployment capacity stands at 4.25 IT-GW, with NVIDIA holding an option to expand by an additional 3.75 IT-GW, bringing the campus's total planned capacity to 8 IT-GW. NVIDIA has disclosed that each generation of AI factory systems deployed at PORTS-Pike will incorporate approximately 1.5 million GPUs, corresponding to a revenue opportunity of roughly $150 billion to $200 billion for NVIDIA.

From Selling Chips to Securing the "AI Factory"

NVIDIA describes AI factories as the "defining infrastructure of the intelligence era," positioning LPS as the next class of strategic resource following chips, packaging, memory, and networking. The logic parallels traditional supply chain management: when customer demand shows high visibility, securing key inputs in advance becomes a priority.

PORTS-Pike will exclusively deploy NVIDIA's full-stack DSX AI factory platform, encompassing GPUs, CPUs, networking, and infrastructure software. For NVIDIA, once the campus is locked in, long-term assets such as land, power, and buildings can support multiple generations of GPU upgrades, rather than losing value when a single chip generation reaches the end of its lifecycle.

In other words, NVIDIA is attempting to transform AI infrastructure from a one-off exercise of "building data centers and buying GPUs" into a long-term asset that can continuously roll forward with each GPU generation.

OpenAI's Demand Serves as the Project's Anchor

Behind this collaboration lies the reality that computing demand from frontier AI companies is rapidly outstripping what traditional balance sheets can sustain.

NVIDIA has explicitly pointed out that the growth rate of computing demand from certain frontier AI labs has surpassed what their own balance sheets and long-term credit profiles can support, making it difficult for them to independently finance large-scale infrastructure.

The PORTS-Pike agreement effectively introduces a new financing model: NVIDIA uses its scale, credit strength, and long-term demand visibility to provide partial credit support for infrastructure financing, while OpenAI locks in computing capacity through long-term leases.

In terms of computing scale, OpenAI has already committed to deploying approximately 12 gigawatts (GW) of NVIDIA computing power by 2030. According to NVIDIA's disclosures, if the PORTS-Pike arrangement expands beyond the initial 4.25 GW phase, OpenAI's related deployment scale could rise to approximately 16 GW.

NVIDIA estimates that at this scale, the related computing opportunity would correspond to roughly $600 billion in NVIDIA compute business revenue by 2030. In essence, PORTS-Pike is not merely an infrastructure financing arrangement; it represents NVIDIA binding together compute procurement, infrastructure construction, and financing capability around core customer demand.

This also explains why NVIDIA is willing to step into infrastructure financing, an area that was not previously its core business: as long as customer demand is sufficiently certain, locking in LPS is equivalent to securing deployment space for multiple future GPU generations in advance.

How the 8GW Campus Takes Shape

PORTS-Pike is located in Pike County, Ohio, adjacent to the site of the decommissioned Portsmouth Gaseous Diffusion Plant. The campus spans both private and federal land, with development led by SB Energy in partnership with AEP Ohio, the U.S. Department of Energy, and the U.S. Department of Commerce.

The project will be built in phases, with the initial phase expected to come online between 2028 and 2030. SB Energy will provide data center construction, operation, and hosting services to OpenAI through a 20-year lease.

From a business model perspective, OpenAI handles long-term leasing and usage, SB Energy manages campus development and operations, and NVIDIA ensures sustained deployment of its computing power at the campus through investment and limited credit guarantees.

The project also includes substantial community commitments, including an $80 million initial community benefits fund and hundreds of millions of dollars in long-term community investment pledges, with expectations of creating tens of thousands of jobs.

NVIDIA's Guarantee Is Not a Blanket Backstop

The market should pay closer attention to exactly how much risk NVIDIA is taking on.

NVIDIA has made clear that its credit support does not cover the entire campus cost, nor does it backstop all of OpenAI's lease obligations. Instead, it is limited to specific portions of lease and electricity payments, along with one specific residual value commitment.

As OpenAI pays rent and data centers gradually come online, NVIDIA's contingent exposure will correspondingly decrease.

If OpenAI ceases to use a portion of the capacity in the future, NVIDIA believes the broad adaptability of the CUDA ecosystem would allow these computing facilities to be subleased to cloud service providers, enterprise customers, other AI labs, and startups. Therefore, the campus does not rely entirely on the long-term credit of a single tenant.

This is also why NVIDIA emphasizes the "strategic and disciplined" nature of its LPS approach: the company does not intend to fully absorb customers' data center construction costs, but rather to support only those core sites where demand is highly visible, sustained over a sufficient duration, and capable of supporting multiple generations of NVIDIA computing deployments.

AI Infrastructure Competition Enters Its Next Phase

Over the past few years, the core battleground in the AI industry has centered on GPUs, HBM, advanced packaging, and high-speed networking. As computing demand continues to surge, land, power, and data center construction cycles are emerging as the new bottlenecks.

The significance of PORTS-Pike therefore extends beyond a single 8GW data center project.

More importantly, NVIDIA is attempting to establish a complete "chips + networking + software + infrastructure" AI factory ecosystem. For customers, this means shifting from purchasing GPUs to directly obtaining an entire AI computing infrastructure that can be sustainably upgraded. For NVIDIA, it means further binding future GPU demand to its platform by securing LPS resources in advance.

If this model proves replicable, NVIDIA's role could undergo a significant transformation: it would no longer simply sell GPUs to data centers, but would begin shaping where those GPUs are deployed, who builds the facilities, and what power sources drive them.

This may be the most compelling aspect of the PORTS-Pike project worth watching closely.