Securing the Infrastructure of Intelligence

AI factories represent the core infrastructure of the current AI age, where compute converts energy and data into intelligence that drives businesses, industries, and nations. These factories demand a comprehensive set of essential resources: cutting-edge chips, packaging, memory, networking, land, power, and shell. Companies like NVIDIA leverage their scale, long-term foresight, and supply-chain alliances to secure crucial semiconductor resources, mirroring this approach to secure LPS capacity exclusively for NVIDIA AI factories.

Today, a strategic partnership has been established between NVIDIA and SB Energy to secure LPS capacity at the PORTS-Pike Technology Campus in Portsmouth, Ohio, which will host NVIDIA compute exclusively for OpenAI. OpenAI will assume the role of the tenant.

For the majority of NVIDIA's clients, securing LPS has been an integral part of their infrastructure strategy. The world's leading cloud service providers and investment-grade enterprises possess the financial stability, infrastructure expertise, and long-term contracts to independently secure LPS. They construct and manage AI factories using NVIDIA-accelerated computing, networking, systems, and software.

However, frontier AI labs face a different challenge. These labs demand substantial quantities of training and inference compute but often encounter difficulties in supporting their balance sheets and long-term credit ratings, leading to constraints on securing AI factory infrastructure independently. As these companies grow, their growth is increasingly limited by the availability of compute.

More compute translates to enhanced intelligence, expanded product offerings, increased users, and amplified revenue. NVIDIA is instrumental in providing the necessary infrastructure to fuel this positive cycle.

PORTS-Pike is a prime location for multiple generations of NVIDIA compute. OpenAI will develop and operate a world-class AI factory at this site, employing NVIDIA's comprehensive DSX AI factory platform, comprising GPUs, CPUs, networking, and infrastructure software. The initial deployment is projected to generate 4.25 gigawatts of AI factory capacity.

Each subsequent generation of NVIDIA AI factory systems installed at PORTS-Pike could represent roughly 1.5 million NVIDIA GPUs or $150 billion to $200 billion in NVIDIA revenue. Over a 20-year period, the site can accommodate multiple upgrade cycles. This economic insight underscores that the LPS commitment secures a long-term AI factory site, while the NVIDIA compute inside can be upgraded repeatedly, delivering greater production, greater intelligence, and improved economics.

NVIDIA may choose to extend the arrangement at PORTS-Pike beyond the initial 4.25 gigawatts, securing the remaining 3.75 gigawatts. OpenAI and NVIDIA are expanding the potential for significant NVIDIA compute deployments through 2030.