Are Nvidia's Customer Relationships Concentrated? Quanta and Samsung Compared
July 30, 2026
Altsets
Research by Altsets Research
Two quantified Nvidia customer relationships show 0.53% of Nvidia revenue tied to Quanta and 0.04% tied to Samsung, illustrating why famous customer names do not prove concentration.
Data used:Altsets Supply Chain Intelligence: 90k+ entities, 400k+ relationships, 20+ years of history.
Key findings
- The displayed Nvidia to Quanta relationship is associated with 0.53% of Nvidia revenue and 1.76% of Quanta's cost base.
- The displayed Nvidia to Samsung relationship is associated with 0.04% of Nvidia revenue and 0.10% of Samsung's cost base, while Amazon, Microsoft, and Super Micro Computer are structural-only in the supplied graph.
The two quantified downstream customer relationships visible in this Altsets Nvidia graph do not by themselves show heavy customer concentration.
The mapped Nvidia to Quanta Computer relationship is associated with 0.53% of Nvidia revenue and 1.76% of Quanta's cost base. The mapped Nvidia to Samsung Electronics relationship is associated with 0.04% of Nvidia revenue and 0.10% of Samsung's cost base. The same graph also shows downstream connections from Nvidia to Amazon, Microsoft, and Super Micro Computer, but no comparable percentage metrics are visible for those edges in the supplied view.
That distinction matters. A list of recognizable customers can look concentrated even when the quantified relationships shown are small relative to the supplier's revenue base.
What the graph actually answers
The graph supports two different observations. First, Nvidia has multiple visible downstream customer relationships. Second, two of those relationships carry displayed economic percentages:
| Nvidia customer relationship | Nvidia revenue percentage | Customer cost percentage |
|---|---|---|
| Quanta Computer | 0.53% | 1.76% |
| Samsung Electronics | 0.04% | 0.10% |
The percentages are directional. The Nvidia revenue percentage asks how much of Nvidia's revenue is associated with the customer relationship. The customer cost percentage asks how much of the customer's cost base is associated with Nvidia. They should not be combined into one dependency score.
A famous customer is not automatically a concentrated customer
This is one of the most useful distinctions supply-chain data can add to company research. Seeing Amazon, Microsoft, Super Micro Computer, Samsung Electronics, and Quanta Computer in the same downstream network creates a strong visual impression. But company-name recognition is not a measure of revenue concentration.
For the two relationships with visible percentages, the supplier-side revenue shares are 0.53% and 0.04%. Those values are small enough that neither displayed edge, by itself, supports the conclusion that Nvidia is highly dependent on that customer. The correct conclusion is narrower: the customer relationship exists, and the visible economic share is limited relative to Nvidia's revenue base in this snapshot.
Quanta looks more material than Samsung in the displayed percentages
Quanta's relationship is associated with 0.53% of Nvidia revenue, compared with 0.04% for Samsung Electronics. That makes the displayed Quanta relationship about thirteen times larger on the supplier-revenue percentage measure. The customer-side percentages also differ.
Nvidia is associated with 1.76% of Quanta's cost base and 0.10% of Samsung's. Again, the Quanta relationship appears more economically visible from both sides within this comparison. That does not mean Quanta is Nvidia's most important customer overall. It only compares the two quantified downstream edges shown here.
Why structural relationships still matter
The unquantified Amazon, Microsoft, and Super Micro Computer edges should not be ignored. A structural relationship can still be useful for:
- customer discovery;
- event monitoring;
- earnings read-through research;
- product-cycle research;
- identifying possible shared demand drivers;
- deciding which relationships deserve additional metric retrieval.
What the structural edge does not support is a numerical concentration claim. Missing percentage data should remain unknown. It should not be converted to zero, and it should not be assumed to be large simply because the customer is well known.
What this means for Nvidia earnings research
Suppose one of the visible downstream customers reports a major change in capital spending or demand. The graph gives an investor a set of customer relationships to investigate. The percentage data then helps determine whether the event appears large enough, relative to Nvidia's revenue base, to deserve priority.
For Quanta, the visible supplier-revenue percentage is 0.53%. For Samsung, it is 0.04%. An investor should therefore avoid treating a headline at either company as automatically material to Nvidia. The event would still need to be connected to the specific business represented by the relationship.
Customer concentration requires a broader denominator
A proper customer-concentration analysis should not stop at two edges. It should ask:
- How many customer relationships are mapped?
- Which have quantified supplier-revenue percentages?
- How much of total supplier revenue is represented by the quantified set?
- Are several customer entities economically linked to the same end market?
- Are unquantified relationships missing because of data coverage or because the exposure is genuinely small?
- Has the customer mix changed over time?
The current graph is enough to reject one simplistic interpretation: recognizable customer names do not automatically prove customer concentration. It is not enough to calculate Nvidia's complete customer-concentration profile.
Why this is different from sector analysis
Sector analysis would classify Nvidia as a semiconductor company and its customers by their own industries. Relationship analysis asks a more direct economic question: which named counterparties sit downstream, and how large do those relationships appear relative to each side? That can reveal small but real customer relationships that sector labels cannot show. It can also prevent overreaction to a customer headline when the mapped revenue exposure is limited.
A bounded conclusion
Within the supplied graph, Quanta Computer is the more economically visible of the two quantified Nvidia customer relationships, at 0.53% of Nvidia revenue and 1.76% of Quanta costs. Samsung Electronics is associated with 0.04% of Nvidia revenue and 0.10% of Samsung costs. Amazon, Microsoft, and Super Micro Computer are structurally visible but unquantified in this view.
The right conclusion is therefore not "Nvidia is customer-concentrated" or "Nvidia is fully diversified." It is that customer concentration has to be measured, not inferred from a list of famous company names. For metric definitions and limitations, read the Altsets supply-chain data methodology. Browse the broader Company Dependency Research collection for other bounded relationship analyses.