Nvidia's latest financial disclosures reveal a significant and increasing reliance on a small number of undisclosed customers, raising questions about the sustainability of its rapid growth amidst the booming artificial intelligence sector.
Technology Sector Faces Customer Concentration Scrutiny
U.S. technology equities are navigating a complex landscape, with Nvidia Corp. (NVDA), a leading designer of graphics processing units (GPUs) and a bellwether for the artificial intelligence (AI) market, drawing investor attention for its increasingly concentrated customer base. While the company continues to report robust revenue growth, recent disclosures highlight that a substantial portion of its sales is attributed to just a few large, unnamed clients, a dynamic that introduces a notable layer of risk.
The Event in Detail: Concentrated Revenue Streams
Nvidia's financial filing for its fiscal second quarter, ending July 27, 2025, revealed that two undisclosed direct customers, referred to as "Customer A" and "Customer B," collectively accounted for 39% of the company's total revenue. This marks a significant increase from the 30% these same two customers represented just three months prior, in the fiscal 2026 first quarter, where Customer A contributed 16% and Customer B 14% of total revenue.
In the latest quarter, Customer A's contribution rose to 23% and Customer B's to 16% of Nvidia's total revenue, which stood at $46.74 billion. This concentration is underscored by the fact that without the revenue from Customer A and Customer B, Nvidia's quarter-over-quarter revenue would have declined by $2.36 billion. The company reported overall revenue growth of $2.64 billion quarter-over-quarter. Furthermore, a deeper analysis reveals that an astonishing 85% of Nvidia's total sales in the recent quarter originated from just six direct customers, four of whom individually contributed 10% or more to revenue.
Nvidia's direct customers typically include add-in board manufacturers, original equipment manufacturers (OEMs), system integrators, and distributors. Indirect customers, such as major cloud service providers (Amazon Web Services, Microsoft Azure, Google Cloud, Oracle), often acquire Nvidia's chips through these direct channels, and are widely believed to be the primary drivers behind the sales to Customer A and Customer B.
Analysis of Market Reaction: Vulnerability Amidst Growth
The increasing reliance on a handful of major clients presents a dual-edged sword for Nvidia. While these large customers fuel immense short-term revenue, they also introduce significant long-term risk. Market analysts suggest that if even one of these dominant customers scales back orders or shifts spending, Nvidia's revenue could experience a sharp decline within a single quarter. This concern is amplified by the fact that these large tech giants are actively developing their own in-house AI chips, such as Google's Tensor Processing Units (TPU) and Microsoft's Maia, to reduce their dependence on external vendors like Nvidia. This strategic move by key clients represents a tangible competitive threat to Nvidia's long-term market dominance.
This phenomenon is not unique to Nvidia within the semiconductor industry. Broadcom Inc. (AVGO), another significant player in the AI semiconductor space, also faces similar customer concentration challenges. Broadcom's management has forecasted that its AI semiconductor sales will comprise over 32% of its revenue in the fiscal third quarter, with a large fraction stemming from a few hyperscalers. Broadcom is reportedly working to diversify its core customer base beyond its established contributors, aiming to convert new prospects into revenue-generating clients by 2027.
Broader Context and Implications: The Hyperscaler Spending Cycle
The AI investing theme, and chipmakers like Nvidia and Broadcom in particular, face a broader risk tied to the capital expenditure (capex) trends of hyperscalers. Currently, capex as a percentage of revenue for major cloud companies, including Amazon, Microsoft, Alphabet, Meta Platforms, and Oracle, is at a five-year high, indicative of a significant expansion phase. However, this cycle may eventually shift as these companies prioritize generating free cash flow over accelerated growth, which would directly impact their spending on AI infrastructure and, by extension, Nvidia's revenue.
Industry benchmarks typically suggest that revenue concentration becomes a high-risk factor when more than 10% of revenue comes from a single client, or 25% from a group of five prominent clients. Nvidia's current figures, with 39% from two customers and 85% from six, significantly exceed these benchmarks, signaling a heightened risk profile. Despite these concerns, the overall demand for AI infrastructure is projected to grow substantially, with some forecasts suggesting the market could reach between $3 trillion and $4 trillion by the end of the decade. Nvidia's technological leadership, particularly with its Blackwell architecture and CUDA ecosystem, positions it to capture a large share of this expanding market.
Looking Ahead: Diversification and Sustainable Growth
Nvidia is actively addressing these market challenges through strategic initiatives aimed at diversifying its customer base and strengthening its competitive position. The company is investing heavily in research and development, exploring alternative markets, and focusing on its software ecosystem. These efforts are crucial for mitigating the risks associated with customer concentration and ensuring sustainable growth.
While analysts generally maintain an optimistic outlook for Nvidia, projecting sustained double-digit growth in both revenue and earnings per share through 2028, investor sentiment towards chip stocks remains "brittle." This suggests that any deviation from growth expectations or a slowdown in hyperscaler capital expenditures could lead to significant market reactions. The ongoing interplay between robust AI demand, the evolving capex strategies of hyperscalers, and Nvidia's strategic responses will be key factors for investors to monitor in the coming quarters.