Nvidia CEO: 40% of revenue now from non-cloud customers, not just hyperscalers

Jensen Huang said something that should reframe how you think about Nvidia’s growth story.

It did for me.

“We gained share the number of customers outside of the clouds. The top five cloud service providers has now grown to 40% of our company’s business. Most people think that most of our business is in the cloud, but in fact, the fast, one of the fastest growing segments is outside of the enterprise.”

In other words, the market has been getting Nvidia wrong. The assumption that hyperscalers drive the whole story is outdated. An analyst on set who holds the stock called Huang’s explanation “as clear and understandable” a case as a CEO could make to investors on why the market has been mispricing the company.

The Numbers Behind the Shift

Nvidia (NASDAQ:NVDA | NVDA Price Prediction) just reported a Q4 FY2026 total revenue of $68.13 billion, up 73.2% year over year. The Data Center segment alone generated $62.31 billion, up 75% year over year. Inside that number, Data Center Networking surged 263% year over year to $10.98 billion, driven by the NVLink fabric ramp for GB200 and GB300 systems.

Throughout FY2026, large cloud service providers represented approximately 50% of Data Center revenue each quarter. That means the other half, and by Huang’s telling one of the fastest-growing portions, is coming from enterprise, industrial, sovereign AI, and infrastructure partners.

The Marvell angle is a concrete example. Nvidia’s ecosystem is spreading into enterprise and industrial through investments including a deal with Marvell. Marvell Technology (NASDAQ:MRVL) reported Q3 FY2026 Data Center revenue of $1.519 billion, up 38% year over year, with data center now representing 73% of total sales. The NVLink Fusion partnership deepens that relationship and extends Nvidia’s reach into semi-custom AI infrastructure that enterprises and governments are building themselves.

Why This Matters for Investors

Huang’s point is essentially that Nvidia has already become the toll bridge for AI infrastructure, and the traffic lanes are multiplying. Sovereign AI projects in Saudi Arabia, the UAE, Taiwan, Sweden, and a dozen other countries are building national AI platforms on Nvidia hardware. Industrial customers like Foxconn are seeing thermal simulations accelerate by 150x using Omniverse.

The stock is trading at $174.40, down 6.48% year to date but up 60.95% over the past year. The consensus analyst target sits at $268.22, with 48 buy ratings and 12 strong buys against just one sell. If you believe Huang’s thesis that the enterprise and industrial wave is just beginning, the YTD pullback looks like a setup rather than a warning sign. Investors who focus solely on hyperscaler spending are working from an incomplete map.