SK Hynix Has More Nvidia Exposure Than Micron on Every Displayed Relationship Measure
September 14, 2026
Altsets
Research by Altsets Research
Nvidia represents 27.88% of SK Hynix revenue versus 17.62% of Micron revenue. The SK Hynix relationship is also larger by relationship size and by Nvidia cost share in the supplied data.
Data used:Altsets Supply Chain Intelligence: 90k+ entities, 400k+ relationships, 20+ years of history.
Key findings
- Nvidia represents 27.88% of SK Hynix revenue versus 17.62% of Micron revenue, making the displayed SK Hynix supplier-side exposure about 1.58 times larger.
- The same ordering appears in relationship size, 21B USD versus 9.8B USD, and Nvidia cost share, 27.33% versus 14.00%, without implying that either stock should move proportionally to those percentages.
SK Hynix has the larger Nvidia relationship on every displayed Altsets measure. Nvidia represents 27.88% of SK Hynix revenue versus 17.62% of Micron revenue. The displayed relationship sizes are 21B USD for SK Hynix-Nvidia and 9.8B USD for Micron-Nvidia, while the Nvidia cost percentages are 27.33% for SK Hynix and 14.00% for Micron.
Nvidia is about 1.58 times more important to SK Hynix by supplier revenue share
Supplier revenue percentage is the cleanest metric for asking which memory supplier is more dependent on Nvidia as a customer. Dividing 27.88 by 17.62 gives about 1.58. In the supplied estimates, Nvidia therefore occupies roughly one and a half times as much of SK Hynix's revenue base as it does Micron's.
That does not mean SK Hynix stock should move 1.58 times as much as Micron stock after Nvidia news. Revenue concentration is not stock beta. It does mean that Nvidia demand deserves more weight inside the SK Hynix customer-concentration thesis than inside the Micron thesis when the comparison is limited to these displayed relationship estimates.
The customer-side metrics point in the same direction
The relationship is also larger from Nvidia's side in the supplied data. SK Hynix is associated with 27.33% of Nvidia's cost base compared with 14.00% for Micron. The relationship-size estimates show the same ordering at 21B USD versus 9.8B USD.
Those percentages use a different denominator from supplier revenue share, so they should not be combined into one dependency ratio. The useful conclusion is simply that all three displayed measures independently rank SK Hynix-Nvidia above Micron-Nvidia: supplier revenue share, Nvidia cost share, and relationship size.
Public company evidence supports treating both as real AI-memory relationships
Nvidia's fiscal 2026 filing names both SK Hynix and Micron among the companies from which it purchases memory. SK Hynix and Nvidia also announced a multi-year technology partnership in 2026 around next-generation memory for AI infrastructure. Micron has separately announced HBM4 production designed for Nvidia's Vera Rubin platform.
The public evidence establishes product and strategic relevance. The Altsets data adds the comparative economic layer. Instead of merely concluding that both memory suppliers are tied to Nvidia, the relationship metrics answer the sharper question of which displayed relationship is larger.
The comparison matters most when one Nvidia event affects both suppliers
A major Nvidia demand change can place Micron and SK Hynix in the same event analysis because both have direct customer exposure. The supplier revenue percentages then provide a reason to treat the relationships differently rather than assigning equal exposure because both companies sell memory.
A stronger Nvidia event could matter to SK Hynix through a larger displayed customer concentration while Micron can have different product mix, pricing, capacity, other customers, and valuation. The data establishes the difference in economic exposure, not the final stock outcome.
The larger relationship makes SK Hynix the cleaner Nvidia-demand readthrough of the two
If the research question is narrowly about Nvidia demand, the supplied data gives SK Hynix the stronger direct relationship signal. A larger share of SK Hynix revenue is associated with Nvidia, the relationship size is larger, and the displayed Nvidia cost percentage is larger. That makes SK Hynix the more concentrated Nvidia-demand exposure of the two companies in this comparison.
This does not mean SK Hynix is automatically the better stock when Nvidia is strong. A cleaner readthrough can also mean greater downside if Nvidia demand weakens, changes suppliers, or shifts product mix. The relationship is useful because it sharpens the conditional thesis: an investor expressing a view specifically on Nvidia memory demand should treat SK Hynix and Micron as different exposures rather than interchangeable memory names.
Micron can still outperform for reasons outside the shared customer
The comparison also shows the limit of customer exposure as a stock-selection rule. Micron can have different pricing, capacity, product timing, manufacturing execution, valuation, and exposure to customers other than Nvidia. A smaller Nvidia percentage can reduce one concentration while leaving Micron better or worse positioned on other dimensions.
That is why the conclusion belongs at the relationship level first. SK Hynix is more exposed to Nvidia in the supplied data. The stock conclusion comes later, after the investor decides whether greater Nvidia sensitivity is desirable at the current point in the cycle and whether the market has already priced that relationship more aggressively into one company than the other.
The conclusion is that the shared customer does not create equal exposure
Both Micron and SK Hynix are meaningful Nvidia suppliers, but the supplied Altsets data does not show equal dependence. SK Hynix has the larger Nvidia relationship by supplier revenue share, relationship size, and Nvidia cost share, with Nvidia representing 27.88% of SK Hynix revenue versus 17.62% of Micron revenue. That is the starting point for comparing how one Nvidia catalyst can matter differently to two memory suppliers.
The same-shock different-outcomes guide explains why shared customer exposure does not guarantee the same stock response. The shared-customer relative-value guide explains how a common customer can define a comparison set without becoming a hedge ratio.
For relationship definitions and evidence limits, read the Altsets methodology.