Customer-concentration disclosure is the requirement that a company state when a small number of customers account for a large share of its revenue, because that dependence is a material risk to the business. For semiconductor companies selling AI accelerators and memory, a handful of hyperscale buyers can represent an outsized portion of sales, so concentration disclosure is one of the most-read parts of a chip 10-K. The disclosure typically appears in the risk factors, the business description, and the financial-statement footnotes, and the common reporting threshold is a customer representing 10% or more of total revenue.

The disclosure is grounded in Item 101 of Regulation S-K, which requires a description of any dependence on a few customers, and in Item 105's risk-factor requirement. Item 101 directs registrants to describe their revenue-generating activities and any dependence on key products, services, product families, or customers, including governmental customers.

"Revenue-generating activities, products and/or services, and any dependence on revenue-generating activities, key products, services, product families or customers, including governmental customers"— 17 CFR 229.101 (Item 101), eCFR, source

Direct versus indirect concentration

Semiconductor concentration is complicated by indirect sales: a chip designer may sell through original equipment manufacturers, original design manufacturers, system integrators, and distributors, so the named buyer on the invoice is not the company driving the demand. NVIDIA (NVDA), in its fiscal 2025 Form 10-K, discloses that “for fiscal year 2025, an indirect customer which primarily purchases our products through system integrators and distributors, including through Customer B, is estimated to represent 10% or more of total revenue, attributable to the Compute & Networking segment.” That disclosure shows the two-layer structure: the company tracks concentration at the level of the end buyer it can identify, even when the immediate customer is a channel partner.

What concentration disclosure conveys

Concentration cuts in both directions, and the filing reports the fact rather than judging it. A large customer is a confirmation of demand and a single point of failure at the same time: the loss, deferral, or in-housing of one major buyer's orders would materially affect revenue, which is why the company is required to name the exposure. The disclosure also interacts with segment reporting, because concentration is often attributed to a specific segment, and with supply-side concentration, since the same companies that buy the most chips may also be developing their own. A reader should track the concentration disclosure year over year: a customer crossing or falling below the 10% threshold, or a shift from direct to indirect attribution, changes the risk profile.

Where the 10% threshold comes from

The familiar 10%-of-revenue trigger for naming a major customer comes from the accounting requirement to disclose major customers in the financial-statement footnotes, while the narrative dependence disclosure under Item 101 and the risk-factor treatment under Item 105 are separate and judgment-based. A company can therefore disclose customer dependence in its risk factors even when no single customer crosses the 10% accounting threshold, if the concentration is otherwise material. The two disclosures complement each other: the footnote gives the quantitative line, and the risk factor explains the business consequence of losing or depending on that buyer. A reader who only checks the footnote can miss a concentration that management considers material but that is spread across a small group rather than a single named customer.

The indirect-sales structure also complicates how concentration trends over time. Because a chip designer attributes revenue to the end buyer it can identify even when sales route through distributors and system integrators, the named exposure can move between direct and indirect categories from year to year as channel arrangements change. A customer that bought directly in one year may appear as an indirect customer the next, or vice versa, without the underlying demand changing. That is why the year-over-year read matters more than any single period: the question is not only whether a customer crosses the threshold, but whether the company's overall dependence on a narrow set of end buyers is rising or falling, and whether the disclosure attributes that dependence to a particular segment whose growth the rest of the business is riding on.

Concentration on the supply side mirrors the customer side and is disclosed in parallel. A chip maker that depends on a small number of foundries, memory suppliers, or packaging providers carries supplier-concentration risk just as it carries customer-concentration risk, and both appear in the risk factors and the business description. For the AI-accelerator and high-bandwidth-memory market, the two can overlap: a company that buys the most chips may also supply critical components or be developing its own silicon, so the same counterparties appear on both sides of the disclosure. Reading the customer- and supplier-concentration disclosures together shows how much of a company's revenue and how much of its supply rest on a narrow set of relationships, which is the fuller measure of dependence than either disclosure alone.

The durable reading is that concentration disclosure quantifies dependence, not quality. The filing states which customers cross the reporting threshold and in which segment; it does not predict whether those customers will stay. For an analyst, the risk-factor and footnote language — read alongside the segment attribution — is the grounded source for how much of a chip company's revenue rests on how few buyers.

What the record shows: customer-concentration disclosure flags when a few customers drive a large share of revenue, with a 10%-of-revenue threshold for naming a major customer in the financial-statement footnotes and a separate judgment-based dependence disclosure under Item 101 and risk factors under Item 105 of Regulation S-K; NVIDIA's fiscal 2025 Form 10-K discloses an indirect customer estimated at 10% or more of total revenue attributable to its Compute & Networking segment; and the year-over-year trend is the meaningful read.