NVIDIA CEO Sees $4 Trillion AI Infrastructure Boom as Demand Outpaces Supply

Key Points

• Jensen Huang projects $3 trillion to $4 trillion in global AI infrastructure spending by 2030, driven by generative AI demand and the end of Moore’s law. He said demand is growing faster than supply, with NVIDIA targeting 70% annual revenue growth while expanding capacity.
• NVIDIA is addressing infrastructure bottlenecks across chips, memory, packaging, power and data centers, while relying on regional cloud providers such as CoreWeave, Nebius and Nscale to secure capacity beyond major cloud companies.
• Huang identified cybersecurity, drug discovery and physical AI—including autonomous vehicles, warehouse robots and future AI-enabled networks—as major growth opportunities beyond coding.

NVIDIA NASDAQ: NVDA Chief Executive Officer Jensen Huang said the expansion of generative artificial intelligence, combined with the end of Moore’s law, is driving a major increase in demand for computing infrastructure and semiconductors.

Speaking at the Goldman Sachs Communacopia + Technology Conference, Huang reiterated his view that AI infrastructure spending could reach $3 trillion to $4 trillion by 2030. He described AI as a new computing layer designed to let users “ask anything” and receive generated answers, rather than simply retrieve pre-recorded information from storage systems.

“In this new world, where it is the end of Moore’s law and we need a lot more transistors,” Huang said, companies must pursue extensive hardware and software co-design to continue delivering substantial performance gains.

Huang said NVIDIA’s approach includes connecting chips through NVLink technology and designing systems that extend beyond individual processors. He contrasted the company’s current AI systems with its earlier graphics-chip products, saying the company now ships large-scale systems that integrate GPUs, networking and other components.

AI use cases beyond coding

Huang said coding has been an early and important AI use case, though he framed coding broadly to include recipes, business processes, supply-chain workflows and other repeatable instructions. He said cybersecurity could become the next major application area because AI systems capable of writing code should also be able to identify and patch software vulnerabilities.

He pointed to NVIDIA’s partnerships with CrowdStrike, Cisco and Palantir, describing efforts involving NVIDIA’s open Nemotron models for cybersecurity “red teaming” and “blue teaming.” Huang said Cisco would offer an NVIDIA AI factory platform, while Palantir would build on top of it for deployments to companies and countries.

Huang also argued that NVIDIA benefits from both closed and open AI models. He said the company’s computing platform runs models including Google’s Gemini, xAI’s Grok, Meta’s models, Anthropic’s models and OpenAI’s models.

Supply chain, data centers and neoclouds

On AI infrastructure constraints, Huang cited challenges involving packaging, DRAM, connectors, voltage regulators and wafers. But he said NVIDIA’s scale, longstanding supplier relationships and broad route to market provide advantages in managing the supply chain.

The company is also focused on downstream limitations involving land, electrical power and data-center buildings, which Huang called “land, power, and shell.” He said NVIDIA tracks available capacity globally through its relationships with cloud providers, original equipment manufacturers, AI-native companies and regional cloud providers.

Huang said NVIDIA remained confident it could grow revenue 70% year over year, while describing demand growth as exceeding 100% on an unconstrained basis. He said the company has a year to expand supply.

Regional cloud providers, or “neoclouds,” are an important part of the company’s strategy, according to Huang. He named CoreWeave, Nebius and Nscale, as well as Lambda and Firmus, as examples of companies helping secure infrastructure capacity outside of the largest cloud service providers.

• Huang said NVIDIA’s market includes cloud service providers, enterprises, regional clouds and sovereign AI deployments.
• He cited quantitative trading firms, including Jane Street and Hudson River, as enterprises adopting AI for prediction workloads.
• He also named Eli Lilly, Merck and Bristol Myers Squibb as companies using AI systems in drug-discovery workflows.

Huang said NVIDIA announced work in Australia involving 2 gigawatts of capacity for 2027, which he characterized as representing approximately $80 billion in infrastructure. He said Australia has excess energy but needs technology, ecosystem demand and capital to build AI infrastructure.

Compute as an investable asset

Huang said NVIDIA is working with the financial industry to position NVIDIA computing systems as “investable” and potentially asset-backed. He argued that the systems have durable value because older NVIDIA platforms, including Volta and Ampere, are still rented in the market.

Addressing concerns about circular financing, Huang said NVIDIA makes relatively small investments in some partners while seeing larger contracted demand through their customer pipelines. He said financing for these projects depends on contracted offtake and cited $100 billion in lined-up contracts, without specifying the companies or period covered.

“The returns are too great,” Huang said when asked why he does not view the approach as circular financing.

Physical AI opportunity

Huang said self-driving vehicles represent the first major “physical AI” application. He highlighted NVIDIA’s Alpamayo technology, which he described as a reasoning system for autonomous vehicles that can interpret unfamiliar environments.

He said meaningful progress in self-driving technology could occur over the next two to three years, citing Waymo, Tesla, NVIDIA’s partnership with Mercedes and its partnership with Uber. He also pointed to autonomous mobile robots, warehouse-navigation vehicles and logistics systems, including a partnership with Amazon.

More advanced robotic manipulation systems that can reason rather than follow pre-programmed instructions are likely a couple of years away, Huang said. Over a longer period, he said physical AI could extend to telecommunications, including AI-RAN systems being developed with Nokia for future 6G networks.

About NVIDIA (NASDAQ:NVDA)

NVIDIA Corporation is a technology company that designs accelerated computing platforms, graphics processors and related software. Its products are used for artificial intelligence, machine learning, high-performance computing, computer graphics, data-center applications and other workloads that benefit from parallel processing.

The company's offerings include GeForce graphics processing units (GPUs) and software for gaming and personal computers; data-center GPUs, systems and networking products; and professional visualization solutions for design, engineering, media and scientific applications.

This instant news alert was generated by narrative science technology and financial data from MarketBeat in order to provide readers with the fastest reporting and unbiased coverage. Please send any questions or comments about this story to contact@marketbeat.com.