Nvidia CEO Proposes Modular Infrastructure and Robotic Innovation to Drive Europe's AI Sovereignty
October 26, 2026
During a significant presentation at the Tempodrom in Berlin, Nvidia leader Jensen Huang detailed a strategic roadmap for Europe to transition from a mere user of artificial intelligence to a primary manufacturer of the technology. Speaking to an audience of several thousand at the GTC conference, Huang emphasized that the region possesses the industrial foundation necessary to lead in the emerging era of physical AI. This vision represents a shift from previous company summits that focused primarily on processing efficiency, moving instead toward the deployment of autonomous agents and industrial robotics.
A centerpiece of the announcement was the introduction of the DSX Pod, a prefabricated 'AI factory' designed to bypass the bureaucratic and logistical hurdles often associated with European infrastructure development. These modular units arrive pre-contained in metal shipping structures, outfitted with advanced Vera Rubin computing hardware. By delivering a fully assembled system, Nvidia claims that data centers which typically take years to build can now be operational within two weeks. This rapid deployment model is intended to minimize the time hardware remains inactive and accelerate the region’s computational capacity.
Huang further projected that robotics would become a dominant global industry, highlighting Europe’s history of manufacturing as a competitive advantage. To support this, Nvidia is launching a comprehensive software stack for robotics and a prototype machine named Agile One, scheduled for release in early 2025. The integration of AI extends into telecommunications and transportation, with the CEO confirming collaborations with major German automakers like Mercedes and Volkswagen to advance autonomous driving systems. Even mobile network providers are reportedly retrofitting cell towers with specialized graphics processing units to handle modern data demands.
Beyond physical machines, the presentation touched on the evolving landscape of software. Huang suggested that future enterprises will likely utilize a hybrid of proprietary and open-source models, the latter of which can be tailored using a company's internal data to reach high performance levels. He argued that the rise of AI agents—programs capable of reasoning and executing tasks—will lead to a surge in software utilization, subsequently driving a massive need for more computing power. Despite a brief interruption by environmental activists, the event concluded on a lighter note with an AI-produced song in the style of Eurovision, summarizing the company's multi-layered approach to hardware, software, and autonomous systems.
Read original at Wired AI.