Nvidia said two unnamed buyers made up 39% of its Q2 revenue

Nvidia disclosed on Wednesday that just two unnamed customers were responsible for 39% of its total revenue in the second quarter of its fiscal year, a detail buried inside a regulatory filing submitted to the U.S. Securities and Exchange Commission.

The company listed the buyers simply as “Customer A” and “Customer B,” with the first accounting for 23% and the second for 16% of Nvidia’s sales during the three-month period ending in July. Combined, they nearly controlled $6 billion of the chipmaker’s Q2 topline.

That level of concentration is significantly higher than the same quarter last year, when Nvidia’s two biggest customers made up 14% and 11%.

The spike is now fueling deeper scrutiny into who exactly is behind the massive surge in AI chip spending, and what that means for Nvidia’s revenue stability going forward.

Despite repeated speculation that cloud heavyweights like Amazon, Microsoft, Google, or Oracle might be behind the numbers, Nvidia declined to name the clients.

Nvidia keeps mystery buyers hidden behind layers of supply chain

In the filing, Nvidia described Customer A and Customer B as “direct customers.” That doesn’t mean they’re using the chips themselves.

These direct customers are firms that purchase Nvidia’s hardware to assemble complete systems or boards, which are then sold to the actual end users; like cloud companies, government agencies, and corporate enterprises.

The list of potential intermediaries includes original design manufacturers and equipment builders such as Foxconn, Quanta, and big system integrators like Dell.

Nvidia also acknowledged having indirect customers, the companies who eventually use the systems but don’t buy chips directly from Nvidia.

These are the cloud infrastructure players, tech firms, and large organizations building internal AI platforms. The company said it can only estimate how much of its revenue comes from indirect buyers, using purchase orders and internal sales records.

The mystery deepens with a second nugget from the filing. Nvidia said two indirect customers each represented over 10% of total revenue, and both were served through either Customer A or B.

That detail has triggered more guesswork around whether the indirect buyers are the usual cloud suspects, or perhaps new players scaling up fast in AI.

CFO Colette Kress said during an earnings call that about 50% of Nvidia’s data center revenue came from large cloud service providers. That’s notable because data center sales made up 88% of the company’s total revenue in Q2. Kress told analysts:

“We have experienced periods where we receive a significant amount of our revenue from a limited number of customers, and this trend may continue.”

Analysts watch 2026 cloud capex as growth barometer

The growing reliance on large, unnamed clients has caught Wall Street’s attention. Frank Lee, an analyst at HSBC, wrote in a note Thursday that the market is unlikely to see “further earnings upside revision or share price catalyst in the near-term unless we have increasing clarity over upside in 2026 [cloud service provider] capex expectations.” He currently has a hold rating on Nvidia stock.

Meanwhile, Nvidia added that an “AI research and development company” also brought in a “meaningful” amount of revenue through both direct and indirect purchases. No name was attached to that customer either.

The company told investors that demand isn’t just coming from public cloud providers. Nvidia pointed to a broader mix of buyers, including enterprises building in-house AI systems, foreign governments, and a new category it calls “neoclouds.” These are newer infrastructure providers aiming to challenge the big four with platforms optimized for AI workloads.

CEO Jensen Huang told investors that Nvidia’s long-term forecast sees AI infrastructure growing to $3 to $4 trillion by 2030. He added that for every $50 billion spent on an AI-focused data center, Nvidia could capture about 70% of the cost, not only from selling GPUs but also networking, accelerators, and software stacks.

Jensen also said the current wave of spending is unprecedented. “As you know, the capex of just the top four hyperscalers has doubled in two years as the AI revolution went into full steam,” he told analysts, referring to Amazon, Microsoft, Google, and Oracle.

Don’t just read crypto news. Understand it. Subscribe to our newsletter. It's free.