Nvidia CEO Jensen Huang pitches “AI factories” that can be deployed faster, with new DGX Pod aimed at financing hurdles

Nvidia CEO Jensen Huang said the company is working toward an approach to AI infrastructure that businesses can stand up in weeks rather than months, framing the goal around what he called “AI factories.” Speaking in an interview during Nvidia’s GTC Berlin event, Huang linked the push to practical barriers companies face when they try to deploy large-scale AI systems, particularly around how quickly and how easily those deployments can move from planning to delivery.

At the center of Huang’s pitch is DGX Pod, a newly highlighted offering designed to accelerate the deployment of AI factories. While the specific technical scope and commercial terms were not detailed in the post that circulated the interview, the stated intent was clear: reduce the time required to bring AI infrastructure online and make those deployments easier to pay for and manage.

DGX Pod is presented as a packaging concept rather than a single chip or standalone server. In Huang’s framing, the value comes from assembling a complete factory-scale compute setup into a unit meant to streamline procurement and implementation, which can be slowed by the complexity of configuring and financing many separate components.

Huang’s comments also emphasized speed as an operational objective, not just a performance target. The interview’s headline claim, that Nvidia wants AI factories built in “weeks, not months,” positions DGX Pod as a response to real-world deployment timelines that can be stretched by ordering, integration, and capital planning cycles.

The discussion took place at GTC Berlin, one of Nvidia’s major industry conferences focused on AI, accelerated computing, and data center systems. Nvidia typically uses these events to outline product roadmaps for developers and enterprises, especially around its data center platforms and related software ecosystems.

Industry demand for AI infrastructure has increasingly shifted from isolated proofs of concept to more continuous, factory-style capacity for training and deploying models. In that context, packaged systems like DGX Pod are aimed at making it easier for customers to translate demand into installed compute capacity, potentially compressing the time between a business decision and a usable AI environment.

Even with Huang’s emphasis on faster deployment, key details were not disclosed in the interview post that was circulated via the business feed. The post did not provide additional specifics on DGX Pod’s configuration, target customer segments, pricing, or the precise financing mechanisms it is meant to facilitate. It also did not quantify how much timeline reduction customers could expect relative to other deployment paths.