How to Read Customer Inventory Before Calling It Bad News for Suppliers
September 2, 2026
Altsets
Research by Altsets Research
Break a customer's inventory increase into raw materials, work in process, and finished goods before deciding whether the change signals a production ramp or weakening demand for exposed suppliers.
Data used:Altsets Supply Chain Intelligence: 90k+ entities, 400k+ relationships, 20+ years of history.
Key findings
- Nvidia inventory rose from 21.403B USD to 31.575B USD between January and July 2026, but raw materials and work in process increased while finished goods declined.
- The inventory mix creates a different supplier read-through than a finished-goods buildup, especially for SK Hynix and Micron given their quantified Nvidia revenue exposure.
To decide whether rising customer inventory is bad news for suppliers, separate raw materials, work in process, and finished goods, then compare those changes with management's explanation of production and demand. A higher total can reflect a production ramp rather than a demand glut. Nvidia's fiscal Q2 2027 balance sheet is a useful example.
Total inventories rose from 21.403B USD at January 25, 2026 to 31.575B USD at July 26, 2026. That is a large increase. But the components moved differently:
| Nvidia inventory | Jan. 25, 2026 | Jul. 26, 2026 |
|---|---|---|
| Raw materials | 3.807B USD | 11.341B USD |
| Work in process | 8.822B USD | 13.377B USD |
| Finished goods | 8.774B USD | 6.857B USD |
Raw materials and work in process rose sharply while finished goods fell. That looks very different from a simple pile-up of unsold finished products.
Total inventory is not enough
A top-line inventory increase can come from weak demand, prebuilding ahead of demand, product transition, longer lead times, supply assurance, manufacturing ramp, geographic restrictions, or strategic component purchases. The composition helps narrow the interpretation. If finished goods are rising rapidly while shipments slow, the signal can be concerning.
If raw materials and work in process are rising while finished goods decline, the company may be building into future production. The filing still needs operating context.
Why suppliers should care
Altsets maps SK Hynix and Micron as suppliers to Nvidia. The displayed supplier-revenue exposure is Nvidia at 27.88% of SK Hynix revenue and Nvidia at 17.62% of Micron revenue. That means Nvidia inventory dynamics are economically relevant enough to deserve attention for both suppliers. The question is not "inventory went up, therefore supplier demand goes down." The question is: What kind of inventory went up, and what does that imply about future procurement?
Nvidia's current mix points toward a production question
From January to July 2026: raw materials increased by about 7.534B USD, work in process increased by about 4.555B USD, and finished goods decreased by about 1.917B USD. That pattern is consistent with a company carrying more inputs and partially processed products while holding less finished inventory. It does not prove future supplier orders will rise.
But it gives a very different starting hypothesis from a finished-goods glut. An investor should then read production-ramp, demand, and product-transition commentary.
Export restrictions complicate the inventory picture
Nvidia also disclosed H20 and H200 inventory and purchase-obligation charges tied to export restrictions. This means not all inventory growth should be attributed to healthy demand. Some inventory can become stranded by regulation.
Other inventory can support a strong product ramp. The same balance sheet can contain both dynamics. That is why component-level inventory disclosure is useful.
Supplier exposure does not reveal which inventory bucket they feed
A mapped supplier relationship does not tell you whether a supplier's products sit in raw materials, work in process, finished goods, or purchase commitments at the customer. For memory suppliers, the operational role can vary by product and production stage. The network gives economic relevance.
The accounting table gives inventory composition. Product research is needed to connect them.
Look at purchase obligations too
Nvidia reported 2.138B USD of excess inventory purchase obligations at July 26, 2026, down from 2.739B USD at January 25. Purchase obligations matter because suppliers can remain supported by committed orders even when final demand changes. But excess commitments can also become a source of future charges.
An investor researching suppliers should therefore monitor inventory composition, purchase commitments, cancellation rights, provisions, product transitions, and customer demand. No single line item answers the supply-demand question.
A better supplier read-through framework
Suppose a major customer reports a 50% inventory increase. Before treating that as bad news for suppliers, ask:
- Did raw materials rise?
- Did work in process rise?
- Did finished goods rise?
- Did revenue also accelerate?
- Is a new product ramp underway?
- Are there excess-inventory provisions?
- Are purchase commitments increasing or decreasing?
- Which suppliers are economically exposed to the customer?
Supply-chain data answers the last question. The filing answers the first seven. Combining them creates a much more useful supplier signal.
A repeatable inventory-read-through workflow
- Identify the customer with a large inventory change.
- Break inventory into raw materials, work in process, and finished goods.
- Compare the direction of each component.
- Review inventory provisions and purchase obligations.
- Read product-ramp and demand commentary.
- Map suppliers with meaningful customer exposure.
- Validate product relevance.
- Distinguish production buildup from unsold finished-goods accumulation.
- Track the same inventory composition over subsequent quarters.
The management-guidance cross-check shows how public commentary across connected companies can confirm or contradict a demand narrative. This article adds a balance-sheet signal: whether the customer's inventory composition supports the same story. For relationship definitions and limitations, read the Altsets supply-chain data methodology.