Jensen Huang Targets $200 Billion Market with Nvidia’s Agentic AI-Focused Vera CPU

What happened

Nvidia founder and CEO Jensen Huang announced that the company's new Vera CPU has opened a brand new $200 billion total addressable market (TAM). During an earnings call following record-breaking quarterly revenue of $81.6 billion, Huang positioned the Vera processor as the center of a global transition toward agentic and robotic physical AI. He argued that while GPUs handle the 'thinking' aspects of AI models, the emerging ecosystem of autonomous agents will require massive amounts of specialized CPU power.

What's new in this update

Huang disclosed that Nvidia has already generated $20 billion in revenue from standalone Vera CPU sales this year. This rapid adoption suggests that Nvidia's entry into the CPU market—a sector historically dominated by Intel and AMD—is gaining significant traction. The company is forecasting $91 billion in revenue for the next quarter, driven in part by the bundling of Vera with the upcoming Rubin GPU architecture and its appeal to major hyperscalers.

Key details

The Vera CPU differs from classic cloud architecture by focusing on the speed of token processing rather than the ability to run multiple simultaneous app instances via traditional cores. Huang explained that because AI agents will use digital tools similar to how humans use PCs, the demand for dedicated CPUs will scale with the number of agents. He predicted a future with billions of AI agents, each requiring the specific processing capabilities that the Vera chip was purpose-built to provide.

Background and context

While Nvidia is the undisputed leader in GPUs, the CPU market has remained a competitive frontier. Major cloud providers like Amazon Web Services (AWS) have been developing their own homegrown AI CPUs, such as the chips recently selected by Meta, to reduce dependence on external vendors. Nvidia’s introduction of Vera in March represents a direct challenge to both these internal projects and the traditional X86 architecture offered by Intel and AMD.

What to watch next

The primary indicator of success will be the continued partnership of major hyperscalers and system makers in deploying Vera-based systems. Industry observers will also watch if Nvidia can maintain its momentum as AWS CEO Andy Jassy and other cloud leaders continue to pitch their own proprietary silicon as superior alternatives for high-scale AI workloads.