Jensen Huang, the CEO of Nvidia, has made a striking claim about the evolution of agentic AI, stating that it has moved beyond mere technology demonstrations to become a source of actual revenue across multiple industries. This advancement allows systems to autonomously plan, reason, and execute tasks without needing constant human oversight, a development that is accelerating at a pace that demands attention from investors and developers alike.
Huang's assertion is not just a prediction; he argues that the software industry is on the verge of transformation. He believes that soon, every software application will integrate agentic capabilities, fundamentally reshaping the industry. This change is significant, suggesting a future where software does not merely provide suggestions but actively performs tasks.
Understanding the difference between traditional AI and agentic AI is crucial. Traditional AI functions like a highly intelligent intern, capable of answering questions but reliant on direct instructions. In contrast, agentic AI resembles a junior employee who can autonomously take a goal, break it down into actionable steps, utilize various tools, make decisions throughout the process, and deliver completed tasks. This marks a profound leap in AI capabilities, enabling multi-step decision-making and independent tool usage.
For example, while a chatbot can summarize emails, an agentic AI could take that further by reading emails, identifying actionable items, scheduling meetings, drafting responses, and managing calendar conflicts—all without user intervention. This evolution in product offerings is significant, as Huang emphasizes.
Nvidia has demonstrated its commitment to this vision by launching the Nemotron model family, designed for complex agentic AI tasks. These models are equipped to handle the intricate reasoning and tool usage that agentic workflows require, establishing a solid foundation for developers to create autonomous systems within Nvidia’s ecosystem.
The rise of agentic AI is not just about enhancing intelligence; it fundamentally grants agency to AI systems. Unlike previous AI iterations that relied on prompts, agentic AI operates with a level of autonomy that boosts productivity in enterprise environments marked by repetitive, complex, and time-sensitive tasks. As these systems become more common, the demand for computational resources is surging, particularly for GPUs and high-bandwidth memory capable of supporting these intensive workloads.
Nvidia, already a dominant player in the data center GPU market, stands to gain significantly as demand for agentic AI solutions grows. The company is well-positioned to capture a substantial portion of this spending, following a familiar pattern in the tech industry: when new AI capabilities emerge, the need for increased computational power follows, leading to growing order books for Nvidia.
However, Huang is not content to limit agentic AI to data centers alone. Nvidia is working to extend these capabilities to consumer-grade PCs, particularly those running Windows. The goal is to make agentic AI accessible beyond enterprise settings, greatly broadening the potential market for Nvidia’s hardware across the computing spectrum.
This push into consumer computing could have significant implications. If agentic AI becomes standard in personal computers, it could transform PCs from tools that users manage into systems that operate autonomously on their behalf. Such a shift promises not only a new hardware upgrade cycle but also positions Nvidia as a key player in the evolving landscape of personal computing.
The stories that move AI & crypto markets — before the market reacts.
Free. 7am ET. Five stories. 62,400 readers.
