Nvidia CEO Jensen Huang believes the artificial intelligence boom is evolving into something much larger than just a semiconductor cycle. Speaking to CNBC after Nvidia’s earnings, Huang said the company is helping build what he described as a “brand new industry” that stretches across energy, chips, infrastructure, AI models and applications.
“We’re inventing a brand new industry,” Huang said, adding that “computing is essential at every single one of the layers.”
Huang described this ecosystem as a “five layer cake”, where Nvidia is investing and partnering across every part of the AI value chain rather than operating only as a chipmaker.
Here’s how Huang broke down the five layers:
| layers | What it means | Examples |
|---|---|---|
| Energy | AI boom starts with energy because data centers require enormous amounts of power. As AI infrastructure scales up globally, investment in electricity generation and energy infrastructure will also need to rise. | Energy infrastructure and power investments |
| Chips & hardware | Nvidia’s role goes beyond its core business of GPUs and now includes networking, photonics, memory and large-scale computing architecture needed to run AI systems efficiently. Vera Rubin is one of the most complex computing systems ever built. | Semiconductor chips, Silicon photonics, networking, copper scale-up technologies |
| infrastructure | This layer includes companies building AI cloud infrastructure and compute capacity. Huang said Nvidia is supplying full systems including CPUs, GPUs, networking and software stacks to hyperscalers, sovereign AI projects and AI-native cloud providers. | CoreWeave, EBS, cloud infrastructure providers |
| AI models | AI has moved from generative AI to “agentic AI”, where models can reason, plan tasks and use tools autonomously. These companies are building the foundational models powering that shift. | LLMs like OpenAI, Anthropic, xAI and others |
| Applications | This is the layer where AI eventually impacts businesses and society through real-world software and products built on top of AI infrastructure and models. | Voice booking, retail search, security healthcare, (QureAI) etc. |
Huang said Nvidia is investing and partnering across all five layers rather than focusing only on semiconductor sales. The company’s strategy, according to him, is to support the broader ecosystem as AI adoption accelerates globally.
Nvidia, the most valuable company in the world, reported better-than-expected first quarter earnings on May 20.
First Published: May 21, 2026 1:11 PM ist
