Nvidia CEO Jensen Huang highlighted the transformative impact of agentic AI on software development and the global economy, emphasizing Taiwan’s vital role in Nvidia’s AI supply chain and the company’s evolution into a comprehensive AI systems provider. He showcased Nvidia’s advanced AI infrastructure, like the Vera Rubin system, underscoring the growing demand for AI compute power as a key driver of revenue and profitability in the AI era.
Nvidia CEO Jensen Huang delivered a keynote speech at Asia’s largest tech summit in Taipei, emphasizing the transformative impact of AI, particularly agentic AI, on the technology landscape and the global economy. Huang highlighted Taiwan’s critical role in Nvidia’s supply chain and ecosystem, praising the island’s tech industry and its projected 10% GDP growth. He explained that AI is not a threat to software jobs but rather a multiplier that increases productivity and demand for software engineers, as AI tools enable developers to produce significantly more output.
Huang introduced the concept of agentic AI, describing it as a new computing model where AI agents, powered by large language models, orchestrate tasks by using tools and managing memory in a distributed and heterogeneous computing environment. He illustrated how these agents can perform complex functions, such as generating code or creating CAD files for 3D printing, marking a shift from traditional software applications to AI-driven workflows. This new paradigm requires software to be designed in a way that AI agents can effectively utilize, opening vast opportunities for software companies.
A significant portion of Huang’s speech focused on Nvidia’s advanced AI infrastructure, particularly the Vera Rubin system, which is designed to handle the demanding compute requirements of agentic AI. Vera Rubin, now in full production, represents a leap in AI hardware with features like liquid cooling, integrated CPUs, and revolutionary networking capabilities. Huang emphasized the importance of performance per watt, reliability, and system longevity in AI data centers, noting that compute demand directly translates to revenue and profitability in the AI era.
Huang also discussed Nvidia’s evolution from a GPU company to a comprehensive systems company, building entire AI factories that include chips, racks, networks, power, and cooling systems. He stressed the complexity and scale of these AI factories, which can cost up to $100 billion per gigawatt of power capacity, and the necessity for these systems to work flawlessly from day one. Nvidia’s integrated approach, including software libraries and ecosystem partnerships, positions it as a key enabler for companies worldwide to deploy AI infrastructure efficiently and effectively.
Market analysts echoed Huang’s optimism, noting strong growth in Taiwan and Korea driven by AI demand, with software companies benefiting from AI-driven business process automation and new revenue streams. The market appears to agree with Huang’s assertion that compute equals revenue, as token consumption and pricing power in AI continue to rise. Overall, Huang’s keynote painted a picture of AI as a powerful economic driver, with Nvidia at the forefront of enabling the next generation of AI technologies and infrastructure globally.