Nvidia's $200 Billion Bet on Agentic AI with Vera

Nvidia Bets on Agentic AI with Vera

Nvidia, under the leadership of Jensen Huang, has unveiled its latest processor, Vera, specifically designed for agentic AI. This strategic launch targets a potential market estimated at $200 billion. This initiative comes as cloud giants like Amazon, Google, and Microsoft develop their own AI chips, seeking to reduce their dependence on Nvidia, the current leader in GPUs.

By venturing beyond its traditional GPU domain, Nvidia hopes to capitalize on the growth potential offered by Vera. This processor is a direct response to the rise of AI chips developed by major competitors such as Amazon, Google, and Microsoft, who are looking to challenge Nvidia's hardware.

This announcement follows an exceptional quarter for Nvidia, with revenue reaching $81.6 billion and forecasts hitting $91 billion for the next quarter. The launch of Vera aligns with Nvidia's goal of controlling the infrastructure necessary for the operation of future autonomous agents.

Vera: A Processor for a Growing Market

First introduced in March, the Vera CPU is designed to meet the specific needs of agentic AI. Jensen Huang, CEO of Nvidia, claims that Vera is the first processor natively designed for this type of application. Until now, Nvidia primarily dominated the GPU sector, which is essential for training and executing AI models. However, Huang envisions a future where AI agents will not rely solely on GPUs.

In this vision, GPUs will continue to play a crucial role in reasoning models, while CPUs, such as Vera, will be essential for task execution, tool management, and the autonomous operations of agents. This evolution could generate massive demand for specialized processors.

Nvidia is thus positioning itself in an unexplored market: that of CPUs. The potential of this market is estimated at $200 billion, with revenue forecasts of $20 billion for this year. Nvidia aims to become the world's leading supplier of both CPUs and GPUs.

Traditional cloud architectures, optimized for multiple applications, are now being rethought to integrate autonomous agents. The Vera processor, with its ability to quickly process AI tokens, fits into this new direction.

The company has already achieved $20 billion in Vera processor sales this year, even though the product is still in its commercial infancy. Jensen Huang emphasizes that major cloud providers and system manufacturers are already collaborating with Nvidia on this architecture.

A Response to Growing Competition from Hyperscalers

Nvidia's push into the CPU market is a response to increasing pressure. Wall Street is questioning Nvidia's ability to maintain its dominant position in the face of hyperscaler initiatives.

Amazon Web Services recently signed an agreement with Meta for the use of millions of internally developed AI chips. Andy Jassy, CEO of Amazon, stated that AWS now offers AI processors capable of competing with those from Nvidia, both in GPUs and CPUs.

Microsoft, Google, and several startups are also developing their own AI accelerators to control their costs and strategic infrastructure. For Nvidia, the risk lies in seeing its major clients become competitors.

With Vera, Nvidia seeks to broaden its scope before the AI processor market becomes structured around alternative players. Jensen Huang aims to reassure investors about Nvidia's ability to sustain its growth despite the rise of proprietary chips.

Nvidia has a significant lead in GPUs, its CUDA software ecosystem, and large-scale AI infrastructures. However, the CPU market has historically been dominated by players like Intel and AMD. To establish itself sustainably, Nvidia will need to convince companies to rebuild part of their architecture around its processors.