Nvidia just told the chip industry it’s coming for everything. The company is projecting a $200 billion total addressable market for CPUs, a segment it has never seriously competed in, and CEO Jensen Huang confirmed that forecast includes anticipated demand from China.
That last part is particularly interesting given the US government’s ongoing export restrictions on advanced AI chips to China. Nvidia is essentially saying: we see the geopolitical headwinds, and we’re still counting that revenue in our projections.
The Vera play: purpose-built for agentic AI
At the center of this push is Vera, Nvidia’s new CPU that was unveiled in March 2026. The company describes it as the “world’s first CPU, purpose-built for agentic AI.”
Nvidia is projecting nearly $20 billion in standalone CPU revenue visibility for the current fiscal year. That’s not a long-term aspiration. That’s money the company expects to see rolling in now.
Initial Vera deliveries have already gone out to some of the biggest names in AI: Anthropic, OpenAI, and Oracle.
The earnings backdrop: $81.6 billion and counting
Nvidia posted $81.6 billion in revenue for the quarter, an 85% increase year-over-year. Data center sales accounted for $75.2 billion of that total.
Nvidia expects Q2 revenue to reach $91 billion. That guidance includes no China data center compute revenue.
The $200 billion CPU TAM forecast, however, does include China. Huang confirmed this on May 23, 2026, drawing a clear distinction between the company’s near-term revenue guidance and its longer-term market sizing for the CPU opportunity.
Why this matters for the competitive landscape
For the past several decades, the CPU market has been a two-horse race between Intel and AMD. Nvidia dominated GPUs and was content to let others handle the general-purpose processing side of the equation.
Nvidia’s move into CPUs isn’t a diversification play in the traditional sense. It’s a vertical integration strategy. The company already controls the GPU layer of AI infrastructure. By adding a purpose-built CPU, it can offer a complete AI stack: the GPU for training and inference, and now the CPU for orchestrating the AI agents that run on top of those models.
The China dimension adds a layer of complexity. Including Chinese demand in a $200 billion TAM forecast while simultaneously excluding China from near-term revenue guidance suggests Nvidia is betting that export restrictions will eventually ease, or that it can develop compliant products that still serve that market.