NVIDIA Secures Land, Power, and Shell to Accelerate AI Factory Infrastructure Supply
NVIDIA has launched its LPS (Land, Power, Shell) strategy to directly secure the physical infrastructure needed for large-scale AI data centers. Under this model, NVIDIA is partnering with SB Energy to establish a dedicated computing facility in Ohio, with OpenAI serving as the primary tenant.
NVIDIA is expanding its supply chain management beyond silicon and networking hardware by introducing the Land, Power, and Shell (LPS) strategy. This initiative aims to directly secure the physical real estate, power grids, and building shells required to build the massive data centers NVIDIA refers to as AI Factories. By controlling these foundational elements, the company seeks to mitigate physical infrastructure bottlenecks that currently constrain AI deployment. As the first major execution of this strategy, NVIDIA announced a partnership with SB Energy to secure exclusive capacity at the PORTS-Pike Technology Campus in Portsmouth, Ohio. This site will host NVIDIA-managed computing resources, with OpenAI serving as the initial anchor tenant. This shifts the traditional cloud procurement model, where individual hyperscalers independently source their own facilities, into a hardware-vendor-driven physical infrastructure model. Building efficient AI Factories requires end-to-end integration across advanced silicon, high-bandwidth memory, networking fabrics, and physical facility designs. Directly controlling the physical layer allows NVIDIA to optimize power distribution and cooling systems tailored specifically to its dense architectures. For enterprises relying on massive computing scales, this strategy promises to remove supply uncertainties, though it also deepens industry reliance on NVIDIA's vertically integrated ecosystem.
Primary source
Open NVIDIA ↗This article summarizes the source. Check the original for full details and conditions.
Comparison
| Aspect | Before / Alternative | After / This |
|---|---|---|
| Infrastructure Sourcing | Cloud providers and tenants independently negotiate and source land, power, and facilities | NVIDIA directly secures and packages physical sites (LPS) for its customers |
| Hardware-Facility Integration | High-density AI hardware is retrofitted into existing general-purpose data centers | Facilities are optimized from the ground up for dense liquid-cooled AI architectures |
| Supply Chain Focus | Primarily focused on securing GPUs, high-bandwidth memory, and networking components | Covers the entire pipeline from silicon to physical grid power and building structures |
