Overview
Speaking at the Morgan Stanley Technology, Media and Telecom conference on Wednesday, Nvidia CEO Jensen Huang addressed the company’s recent AI investments, its current demand profile, and a broader thesis about the centrality of compute in the modern economy.
OpenAI and Anthropic investments
Huang confirmed that Nvidia has completed a $30 billion investment in OpenAI. He said a previously discussed $100 billion arrangement is "not in the cards," explaining that a larger deal was not possible as OpenAI is preparing for an initial public offering later this year. On another front, Huang noted that Nvidia’s $10 billion commitment to Anthropic will likely be "the last one" the company makes to that particular firm.
Cloud capacity and partnerships
Huang described active expansion of OpenAI’s capacity across multiple cloud platforms, naming Microsoft Azure, Oracle Cloud Infrastructure, and Amazon Web Services specifically. He added that Nvidia is ramping up operations on AWS rapidly, while also working to increase Anthropic’s capacity on both AWS and Azure. These remarks underscore Nvidia’s role in scaling AI workloads across several major cloud environments.
Demand and positioning
On demand for Nvidia’s products, Huang said the company has seen a trajectory that moved "from very high to even higher than that." He positioned Nvidia at the leading edge of both physical AI and digital biology AI, framing the company’s products as central to these emerging domains.
Supply chain and scale
Huang emphasized that Nvidia has secured critical elements of its supply chain, listing components such as memories, wafers, CoWos, packaging, connectors, cables, copper, and multi-layer ceramic capacitors. He used a practical example to illustrate scalability, saying that when Satya Nadella requests Nvidia "to stand up a few gigawatts, the answer is no problem," signaling confidence in rapid deployment capacity.
Economic view on compute
Looking beyond hardware and investments, Huang articulated a broader economic perspective: "compute equals revenues," he said, arguing that every company will require compute and that, ultimately, compute will equal GDP. He added that the industry is unlikely to face a shortage of intelligence, but will face the challenge of providing sufficient compute to execute workloads.