Nvidia just told the world it sees a $200 billion total addressable market for CPUs. And yes, that number includes China.
CEO Jensen Huang confirmed the figure during media interactions in Taipei on May 23, framing it as the long-term opportunity for Nvidia’s newly unveiled Vera CPU architecture.
The Vera CPU and a $20 billion 2026 target
The Vera CPU architecture was introduced during Nvidia’s Q1 fiscal year 2027 earnings call earlier that week. It represents Nvidia’s first genuine attempt to capture meaningful share in the CPU market, a segment historically owned by Intel and AMD.
Nvidia expects nearly $20 billion in standalone CPU revenue for 2026, fueled by partnerships with major hyperscalers and system manufacturers.
The earnings report that preceded Huang’s Taipei comments painted the picture of a company firing on all cylinders. Nvidia posted record Q1 FY27 revenue of $81.6 billion, an 85% jump year-over-year. Data center revenue alone hit $75.2 billion, accounting for the vast majority of total sales.
The China question
US export controls currently restrict shipments of advanced AI chips to China, and Nvidia’s own near-term revenue guidance excludes certain China-related AI GPU sales because of those restrictions. The company acknowledged this directly in its earnings materials.
Huang has expressed optimism about the potential for future sales of the H200 chip to Chinese customers if licenses are granted and restrictions are eventually eased.
Including China in the long-term CPU TAM follows the same logic. Huang isn’t saying Nvidia will sell Vera chips to Chinese customers next quarter. He’s saying that over a multi-year horizon, China represents a meaningful portion of global CPU demand, and Nvidia intends to be positioned for that market if and when access opens up.
What this means for investors and the competitive landscape
Nvidia’s push into CPUs fundamentally changes the competitive dynamics of the semiconductor industry. Intel has been the dominant CPU maker for decades, though it has faced growing pressure from AMD and, more recently, from Arm-based designs used by companies like Amazon and Apple.
The nearly $20 billion in expected 2026 CPU revenue, if achieved, would make Nvidia a top-tier CPU vendor essentially overnight. That figure alone would represent a significant fraction of the total server CPU market, and it’s coming from a product line that didn’t exist a year ago.
The China dimension adds a layer of optionality that’s difficult to price. If export controls ease, Nvidia could see a significant revenue uplift from a market it’s currently largely locked out of. If controls tighten further, the $200 billion TAM figure becomes more theoretical than practical, at least for the foreseeable future.