Nvidia's top two mystery customers made up 39% of the chipmaker's Q2 revenue

Nvidia recently disclosed that two of its customers accounted for a staggering 39% of its revenue during the second quarter, as outlined in a financial report submitted to the Securities and Exchange Commission. This revelation has sparked concerns regarding the heavy reliance on a small number of clients for the chipmaker's financial stability. In more detail, "Customer A" contributed 23% of the total revenue, while "Customer B" represented 16%. This marks a significant increase from the previous year's figures, where the top two clients accounted for 14% and 11% of sales, respectively. The company typically shares insights about its major customers each quarter, but this latest disclosure has intensified discussions about whether Nvidia's rapid growth is primarily coming from a select few large cloud service providers, including Microsoft, Amazon, Google, and Oracle. Nvidia's Chief Financial Officer, Colette Kress, noted that large cloud service providers contribute approximately 50% of the company's revenue from data centers, which itself constituted 88% of Nvidia's overall earnings in the second quarter. The company acknowledged that they have experienced fluctuations where a significant portion of revenue comes from a limited customer base, a trend that may persist in the future. Market analysts are increasingly focusing on cloud capital expenditure commitments as indicators for Nvidia's future growth prospects. HSBC analyst Frank Lee commented that unless there is greater clarity regarding forecasts for cloud spending in 2026, there may be little potential for earnings growth or stock price momentum in the near term. Interestingly, the identities of Customer A and Customer B remain undisclosed, as Nvidia officials opted not to reveal them. The company categorizes its clientele into "direct" and "indirect" customers. The direct customers, like Customer A and Customer B, are businesses that purchase Nvidia chips to integrate into systems or circuit boards for resale to data centers and cloud providers, rather than being the end-users themselves. While some direct customers may use the chips for their own internal systems, Nvidia has also indicated that two indirect customers accounted for over 10% of total revenue by purchasing systems through Customers A and B. Adding to the complexity, an "AI research and development company" was also identified as a significant revenue contributor through both direct and indirect sales channels. The demand for Nvidia's AI systems continues to surge, not only from major cloud providers but also from various enterprises and emerging companies that are innovating in the AI sector. Nvidia anticipates recording $20 billion in revenue this year from "sovereign AI" sales to foreign governments, underscoring the diverse sources fueling its growth. Nvidia's CEO, Jensen Huang, has set ambitious projections, estimating that the AI infrastructure market could reach $3 to $4 trillion by the end of the decade. He believes Nvidia could capture a substantial portion of costs related to AI-focused data centers. As AI investments from hyperscalers rise, with commitments reaching $600 billion this year, Huang asserts that Nvidia is well-positioned to capitalize on this transformative technological shift.