Nvidia founder and CEO Jensen Huang is perhaps one of the biggest enterprise users of all time when it comes to his company. He may even surpass Salesforce’s Marc Benioff in relentless optimism about his company’s future and revenue.
Even so, it lives up to the hype, quarter after quarter.
Rather than caution you to view the proclamation that it found a “brand new $200 billion TAM for Nvidia” with skepticism, I’d argue that it earned some confidence.
Huang laid this huge new market at the feet of Nvidia’s new CPU product, Vera, which was introduced in March. Speaking on Wednesday’s earnings call — after Nvidia posted another record quarter with $81.6 billion in revenue and forecast $91 billion for the next — Huang touted Vera as a potentially transformative product. And one that already has promising sales figures.
But no matter how well Nvidia performs, Wall Street worries about what will pull Nvidia off its feet.
Lately, such fears have focused on the CPU. Nvidia is the GPU king, while the CPU markets have historically belonged to companies like Intel and AMD. (Nvidia has made CPUs in the past, of course, but that’s not its core business.)
For example, last month, Amazon Web Services talked about a giant contract it signed with Meta for millions of Amazon’s domestic AI processors. Amazon CEO Andy Jassy has been clear that he believes AWS can make AI chips, both GPUs and CPUs, at least as well, and possibly better, than Nvidia.
But now, with the Vera CPU, which is sold on its own and comes with its Rubin GPU, Huang believes he has unlocked “a major new growth driver” for his company, because Vera is, he believes, “the world’s first CPU built specifically for agent AI,” Huang said on the call.
“Vera opens up a brand new $200 billion TAM for Nvidia, a market we’ve never faced before, and every major hyperscaler and system builder is working with us to develop it. The world is rebuilding computing for agency AI and robotic natural AI. Nvidia is at the center of these transitions,” Huang said.
He explained that while the “thinking” part of an AI model uses GPUs, agents run primarily on CPUs. They use CPUs to do their assigned tasks and, he predicts, will run their own form of CPU-based computing.
Vera is for agents because it is specifically designed to process tokens as quickly as possible. This contrasts with classic cloud architecture CPUs that are designed with “cores” or the ability to run multiple instances of applications as quickly as possible.
That sounds reasonable, but with major cloud providers as well as startups pursuing AI chip development, what makes him think Nvidia will be the go-to source for agent CPUs?
Because, Huang says, Nvidia has already sold $20 billion worth of standalone Vera processors this year, and we’re just getting started.
“The world has a billion users, human users. My view is that the world will have billions of agents, not today. I mean, we’ll evolve into that, but we’ll have billions of agents, and those billions of agents will all be using tools. And those tools will be like computers, like we humans use computers today,” he said.
“We’re going to need a lot more processors,” he explained.
When you purchase through links in our articles, we may earn a small commission. This does not affect our editorial independence.
