NVIDIA's revenue has expanded rapidly over the past two years, but its customer base has simultaneously become increasingly concentrated.

According to a report from September 13, in the fiscal first half ending July, three customers each contributed more than 10% of total revenue, with their combined share reaching 44%. In the prior fiscal year, two customers accounted for 36%. Going back to FY2023, no single customer accounted for even 10% of NVIDIA's revenue.

This trend is driven by the explosive growth of NVIDIA's data center business, which surged from approximately $15 billion in FY2023 to $193.7 billion last fiscal year, with expectations of another doubling this year. As the business scales, its reliance on top-tier customers deepens accordingly.

Who Are the Three Major Customers?

NVIDIA has not named these three customers in its public filings, but the analysis suggests they may include Dell or Hon Hai Technology, both of which integrate NVIDIA chips into servers for resale to other enterprises.

Dell recently disclosed that its "AI-optimized servers" revenue doubled year-over-year to $16.4 billion in the second quarter, driving overall revenue growth of 58%. Hon Hai also reported a 35% year-over-year increase in first-half revenue, with AI equipment sales serving as the primary growth engine.

Additionally, NVIDIA CFO Colette Kress stated in February that the five major cloud providers and "hyperscale" customers, a category that may include Meta and SpaceX, collectively contributed more than 50% of data center revenue.

Microsoft, Meta, and SpaceX are all considered significant buyers of NVIDIA, with each aggressively building out large-scale AI data centers. Google and Amazon also purchase NVIDIA chips in large volumes, but their extensive use of in-house AI chips makes them less likely to rank among NVIDIA's largest customers.

The Concern of In-House Chips

Customer concentration itself poses a risk, but the fact that these major customers are also developing their own chips complicates the situation further.

Microsoft, Meta, and SpaceX are all advancing in-house AI chip initiatives. Once these chips mature, their procurement demand for NVIDIA could gradually decline.

This explains why NVIDIA CEO Jensen Huang has continued to invest in emerging cloud providers (neocloud) and AI enterprises in recent years, with CoreWeave and Nebius being both investment targets and key buyers of NVIDIA chips.

Legendary short-seller Michael Burry, renowned for accurately predicting the 2008 subprime crisis, has also publicly identified NVIDIA's rising customer concentration as a significant risk in recent months.

Soaring Accounts Receivable and Extended Payment Terms

Customer concentration is also reflected in another dimension of NVIDIA's financial data.

In July, NVIDIA disclosed that 70% of its accounts receivable balance came from five customers. By comparison, in January, three customers accounted for 56%, and at the end of FY2025, two customers represented 33%.

The accounts receivable balance itself is also climbing rapidly. At the end of August, NVIDIA reported that first-half accounts receivable jumped 64% to $63 billion, attributing the increase to "extended payment terms for large multi-quarter agreements with certain investment-grade customers."

NVIDIA has already extended payment terms from 45 days to 60 days in the current quarter. However, in its securities filings, the company indicated it may further extend these terms in the future:

"Customer payments are typically due shortly after product delivery. In certain cases, we have provided and may continue to provide longer payment terms, ranging from 90 days up to one year, for purchases by investment-grade customers to support their large-scale data center builds."

Extended payment terms mean NVIDIA receives less cash in any given quarter, while also tying its financial health more closely to the creditworthiness of its largest customers.

Expanding to Overseas and Emerging Customers

Under pressure from rising concentration, NVIDIA is accelerating its efforts to diversify its customer base.

This week, NVIDIA added several Australian enterprises to its "AI Factory" initiative, providing chips and networking equipment to companies building and operating data centers. Earlier this year, NVIDIA also advanced similar projects in India and Armenia.

These moves are part of NVIDIA's broader strategy to cultivate sovereign customers and emerging cloud providers beyond its core U.S. buyer base.