Nvidia Secures 72MW Infrastructure Commitment Through Sharon AI

Nvidia is tightening its grip on the physical layer of AI deployment. Through a new six-year agreement for AI infrastructure, Sharon AI has committed 72mw of capacity.

This deal moves beyond software and chips into the fundamental requirement of the AI era: power and space. By securing a multi-year commitment for 72mw, Sharon AI is locking in the compute density required for large-scale model training and inference.

The significance lies in the duration and the scale. A six-year term provides the stability necessary for the massive capital expenditures inherent in AI workloads. For Nvidia, these types of infrastructure agreements ensure that the demand for their hardware remains tetherable to available power and data center capacity.

As the industry faces constraints in power availability, the ability to formalize capacity through long-term agreements becomes a primary competitive advantage. The bottleneck for AI growth is shifting from chip availability to the physical ability to power and cool the hardware.

Watch the intersection of energy procurement and compute capacity; the next phase of AI scaling will be defined by who controls the megawatts.