In a joint Bloomberg TV appearance at Dell Technologies World 2026 in Las Vegas, Nvidia CEO Jensen Huang and Dell Technologies CEO Michael Dell described a vision of distributed computing in which AI is executed at the point of context rather than relying solely on remote cloud servers.
Huang framed the transition as a move from traditional personal computers to what he labeled "personal AI." He argued that artificial intelligence must operate where information and actions intersect - on devices such as laptops, on manufacturing floors, or inside autonomous vehicles - instead of depending exclusively on cloud processing.
Dell emphasized that the PC remains central to knowledge-worker productivity and that customers are asking for machines with stronger performance to accommodate hybrid AI features. He noted that Dell is integrating the ability to run smaller, local AI models directly inside PCs to meet that demand.
Huang said the rise of agentic AI has altered computing architecture. According to his description, the approach involves constructing large language models to function as a "brain," and then adding a "harness" that enables the model to access memory, networks and tools so it can act as a digital agent.
Nvidia is introducing what Huang described as the highest performing CPU in the world, specifically engineered for agentic AI workloads. The company has collaborated with Dell to build a new form of long-term memory for agents that Dell calls the Dell AI Data Platform.
Dell conveyed that companies rethinking workflows around these agent-driven systems have observed productivity improvements measured in multiples - on the order of 10 to 30 times - instead of incremental gains of 10% to 30%.
Both executives said demand for AI infrastructure currently exceeds supply. They identified memory and advanced semiconductors as bottlenecks. Huang observed that, while the semiconductor supply chain is expanding, demand is growing faster.
Dell pointed out that agent frameworks deployed within enterprises require substantially more CPUs than prior computing models. He explained that agents rely on tools constantly and at high speed compared with human users, which drives up compute requirements inside companies.
Market markers in the session reflected near-term stock movement references: DELL -3.51% NVDA -2.28%.
Contextual note: The comments by both executives focused on architectural shifts and infrastructure constraints tied to running AI where decisions and data are created, and on technical partnerships intended to address agent memory and CPU performance for these emerging workloads.