What Happens When a Major Customer Starts Competing With Its Supplier?
September 14, 2026
Altsets
Research by Altsets Research
A customer can remain commercially important while developing internal technology that narrows future dependence, creating a relationship investors should analyze by product scope rather than label as simply bullish or bearish.
Data used:Altsets Supply Chain Intelligence: 90k+ entities, 400k+ relationships, 20+ years of history.
Key findings
- The supplied network maps Nvidia to Microsoft as a customer relationship, while current public evidence shows both a deep Nvidia-Microsoft AI infrastructure partnership and Microsoft's deployment of its first-party Maia 200 accelerator.
- Internal customer technology does not automatically end the supplier relationship; investors should identify which workloads are being internalized, which still require the supplier, and whether bargaining power changes before revenue does.
A major customer does not always stay only a customer. In technology markets, the same company can buy a supplier's products, build around those products, and simultaneously invest in internal technology that could reduce its future dependence on the supplier.
That creates a relationship investors should not force into a simple bullish or bearish label. When a customer becomes a competitor, the important question is which parts of the relationship are being internalized and which parts are becoming even more valuable.
Microsoft and Nvidia show why customer relationships can become strategically ambiguous
The supplied Altsets network maps Nvidia to Microsoft as a customer relationship. Public evidence also shows a deep current partnership. Nvidia said in June 2026 that the companies were expanding work across Azure, Windows, Microsoft Foundry, and next-generation AI infrastructure.
At the same time, Microsoft has developed its own Maia 200 AI accelerator. Microsoft describes Maia 200 as a first-party inference accelerator designed to improve the economics of AI workloads inside Azure.
Those facts can coexist. Microsoft can remain an important Nvidia customer while also trying to own more of the silicon stack itself.
Internal development does not automatically mean customer loss
It is tempting to see a customer's in-house product and immediately conclude that the outside supplier will lose the business. That can be too simplistic.
Large customers often operate heterogeneous infrastructure. Microsoft has publicly said it continues to modernize its fleet with its own silicon alongside technology from Nvidia and AMD. Its own accelerator can absorb certain workloads while demand for outside GPUs continues to grow elsewhere.
The investment question is therefore not whether Microsoft has a competing chip. It is whether Maia changes the portion of workloads for which Nvidia remains the preferred supplier.
The relationship can become more valuable and more competitive at the same time
Nvidia and Microsoft are still collaborating deeply across AI infrastructure. Microsoft can use Nvidia technology for one part of its fleet while developing Maia to improve economics for another.
That creates a more interesting dependency than a normal customer relationship. The supplier may benefit from the customer's rapid growth while facing a long-term incentive for the customer to internalize high-cost or strategically important components.
An investor should monitor both sides of that tension.
Product scope becomes more important than the customer name
A relationship map can tell the investor that Microsoft matters to Nvidia. It cannot by itself tell the investor which workloads are at risk from Maia 200.
That requires product-level evidence. Maia 200 is currently positioned around inference. Nvidia serves a much broader range of training, inference, networking, software, and accelerated-computing workloads.
The narrower the overlap, the weaker the competitive threat to the overall relationship. The broader the overlap becomes, the more the customer relationship may need to be re-underwritten.
Customer bargaining power can change before revenue visibly changes
A customer building an internal alternative can gain leverage even before it materially reduces purchases.
The customer may have more negotiating power on pricing, product roadmaps, supply commitments, or deployment timing because the supplier is no longer the only credible path.
That means the strategic meaning of the relationship can change before a revenue decline appears in financial statements.
Supply-chain data helps identify which supplier-customer pair deserves that deeper monitoring.
The same pattern exists outside semiconductors
Automakers can develop more components internally while continuing to buy from outside suppliers. Cloud companies can build custom hardware while still expanding purchases from merchant vendors. Large retailers can develop private-label products while remaining major customers of branded suppliers.
The common question is whether the customer is replacing the supplier, complementing the supplier, or using internal capability mainly to improve bargaining power and resilience.
Those are very different outcomes for the investment thesis.
The conclusion is to monitor the direction of dependence
A major customer's internal technology program should not automatically trigger a sell decision. It should trigger a more precise question about which part of the commercial relationship is becoming less dependent on the supplier.
The most important signal is not that the customer became a competitor. It is whether the customer's need for the supplier is shrinking, shifting, or becoming concentrated in a narrower part of the stack.
The customer concentration guide explains why a large customer can create both visibility and risk. The thesis-change-log guide shows how new evidence can be compared with the assumptions behind an existing position.
For relationship definitions and evidence limits, read the Altsets methodology.