Is Nvidia transforming into an AI systems company with Vera Rubin in full production?

Jensen Huang, CEO of Nvidia, delivered a keynote speech at the GTC Taipei 2026 conference, revealing that Nvidia is transitioning from a GPU company to an AI systems company. The company’s next-generation AI accelerator, Vera Rubin, is now in full production, marking a shift from generative AI to agent AI, which focuses on AI assistants. Huang emphasized the need for a new computing architecture for agents, integrating models, tools, runtime environments, and memory systems. Nvidia’s collaboration with Korean partners like Samsung and SK hynix was highlighted, showcasing the use of advanced technologies like HBM4 memory and TSMC’s 3-nanometer process. The conference drew over 6,000 attendees, underscoring the growing interest in AI infrastructure. Huang predicted that AI agents will lead the next decade, with Nvidia focusing on building AI factories and optimizing computing performance for profitability. The event also introduced Nvidia’s new CPU “Vera,” designed specifically for agents, and featured partnerships with global companies, including Naver Cloud and Hyundai Motor Group. Nvidia unveiled its large language model “Nemotron 3 Ultra” and outlined future plans for AI accelerators and software development, positioning itself as a leader in AI systems.