Nvidia CEO Jensen Huang Touts Full-Stack AI Ambitions at GTC Berlin as Ken Griffin Warns of AI Bubble

Nvidia CEO Jensen Huang took the stage at GTC Berlin on October 21 to outline the company's expanding footprint in Europe and position Nvidia as more than a chipmaker, calling it a full-stack AI factory. Huang cited more than 12,000 companies in Nvidia's European ecosystem, over $10 billion invested in the region over the past year, and 14 new startup investments announced that week, while singling out Nebius as performing strongly.

Huang detailed the breadth of Nvidia's CUDA architecture, which now spans cloud providers including AWS, Microsoft Azure, Google Cloud, Oracle Cloud, and Meta; AI labs such as OpenAI, Anthropic, Google Gemini, and SpaceX; neocloud operators including CoreWeave, Crusoe, Lambda, and IREN; hardware partners Dell, HPE, Lenovo, Supermicro, Foxconn, ASUS, and Gigabyte; and more than 10,000 AI startups ranging from Perplexity and Cursor to Mistral, Figure, Harvey, and Runway. The company also unveiled a new DSX POD design aimed at accelerating AI infrastructure rollouts and launched an open-shell reference architecture for AI agents.

The bullish Nvidia showcase arrived alongside a pointed counterpoint from Citadel founder Ken Griffin, who said the world is almost certainly in an AI bubble. On the enterprise security side, JFrog used its swampUP Europe 2026 conference to introduce AI-SEC, a suite covering local shadow AI detection and bring-your-own LLM key scanning, aimed at giving organizations control over AI agents, skills, and models from development through production.