NVIDIA is expanding its city AI partnerships around shared computing rather than separate departmental infrastructure, according to its November 2 update, which positions NVIDIA smart city AI around public services, infrastructure planning, and operational coordination. The article states that physical AI, systems that interpret or model real-world conditions, connects the strategy to streets, buildings, and infrastructure, while agentic AI, software that uses models and tools to carry out tasks, adds a path from observation to action. Microsoft’s Azure Digital Twins already provides a reference point for modeling connected environments, and NVIDIA’s update emphasizes partners bringing models, simulation, and computing into local services. For city buyers, the article says the consequential question is whether those pieces produce a reusable operating foundation or another collection of difficult integrations, and that the shift concerns how cities organize AI capacity as much as which applications they deploy. NVIDIA presents shared infrastructure as a way to support departments, neighboring municipalities, and local developers, but the article notes that an application and a shared service create different obligations, with a common platform needing rules for access, maintenance, priorities, and responsibility across organizational boundaries. It adds that shared computing does not automatically create shared definitions of an incident, a completed job, or an acceptable response, and those agreements remain municipal work.