Nvidia has unveiled its latest AI model, the Nemotron 3 Ultra, at Computex 2026 in Taipei. The model, featuring between 500 and 550 billion parameters, is designed for advanced reasoning and planning, capable of executing complex tasks with minimal human intervention. This launch positions Nvidia as a comprehensive AI platform provider, beyond its traditional role as a chipmaker. The Nemotron 3 Ultra is part of a three-tier family, with the Nano variant for lighter tasks and the Super model, released in March 2026, for mid-range applications. Utilizing latent mixture-of-experts techniques, the Ultra model offers up to five times higher throughput than previous versions, enhancing efficiency and reducing costs for enterprise AI applications. Despite the focus on AI, the keynote did not address blockchain or cryptocurrency initiatives, indicating a strategic emphasis on AI and enterprise computing.