Nvidia CFO reports sovereign AI business doubles YoY, rises 35% QoQ

Governments are no longer content to rent AI capacity from Silicon Valley hyperscalers. They want to own it. And Nvidia is the company they keep calling.

Nvidia CFO Colette Kress confirmed during the company’s most recent earnings call that the sovereign AI segment more than doubled year-over-year and grew 35% quarter-over-quarter. Those are not the kind of numbers that get buried in footnotes.

The scale of the sovereign AI buildout

The company’s AI Clouds, Industrial and Enterprise category, which houses sovereign AI, posted revenue growth of 138% year-over-year and 25% quarter-over-quarter in fiscal Q2 2027, the period ending July 26, 2026.

Sovereign AI revenue in Nvidia’s fiscal year 2026 had already tripled compared to the prior year, crossing $30B, with major contributions from clients in Canada, France, the Netherlands, Singapore, and the UK. By the first quarter of fiscal 2027, that year-over-year growth rate was still running above 80%.

Nvidia’s total revenue for the quarter hit $96.2B, with Data Center revenue reaching $89B, up 117% year-over-year.

Why nations are writing billion-dollar checks to Nvidia

Nearly 40 countries now have Nvidia-powered AI infrastructure in operation, and those nations collectively represent roughly $50 trillion in GDP. Nvidia management has framed sovereign AI investment as structurally tied to GDP growth, meaning the larger a country’s economy, the more it will eventually spend to keep its AI stack independent.

Korea offers one of the more concrete recent examples. Nvidia is helping build a gigawatt-scale AI factory ecosystem there through partnerships with NAVER and Brookfield.

What the acceleration means for the competitive landscape

The growth trajectory from fiscal 2026 to fiscal 2027 shows no sign of decelerating. Sovereign AI revenue tripled in fiscal 2026, grew more than 80% year-over-year in fiscal Q1 2027, and then the broader ACIE category posted 138% year-over-year growth in Q2.

The Blackwell Ultra platform, Nvidia’s current flagship compute architecture for AI workloads, is the product most sovereign clients are deploying at scale.