Why Big Tech’s AI Spending Lifted SK Hynix

The chain investors are pricing in

The rally can be summed up as a four-step chain:

1. Big Tech keeps spending on AI infrastructure. Morningstar/Dow Jones linked SK Hynix’s record quarter to robust AI investments by big global technology companies .
2. AI systems need accelerators, not just ordinary servers. TNW noted that in AI servers, the memory sitting next to the graphics processor has become increasingly important .
3. HBM rides beside the accelerator. SK Hynix has been reported as a major supplier of HBM products for Nvidia AI hardware .
4. Strong demand tightens memory pricing. TNW pointed to a $725 billion hyperscaler capex ramp and a 20% HBM price rise as part of the backdrop for SK Hynix’s stock strength .

In other words, the market is not treating SK Hynix as a generic semiconductor name. It is treating the company as a leveraged play on the memory bottleneck inside AI infrastructure.

Why HBM, not just GPUs, became the swing factor

The public AI hardware story often starts with GPUs. For SK Hynix investors, the crucial part is the memory attached to those processors. HBM is built for the kind of data movement AI systems require, and TNW captured the shift by noting that the most expensive part of an AI server used to be assumed to be the graphics processor — but increasingly, attention has moved to the memory beside it .

That matters because memory-company earnings are highly sensitive to both volume and pricing. SK Hynix attributed its strong third-quarter performance to brisk shipments of higher-end HBM3E products and higher prices for DRAM chips, including chips used in data servers for AI training . So AI data-center spending is not just a theme for SK Hynix. It can translate into shipments, pricing power and earnings expectations.

Why SK Hynix got singled out

SK Hynix is not merely benefiting from a broad memory-cycle recovery. Its specific appeal comes from its exposure to the Nvidia-centered AI accelerator supply chain. Morningstar/Dow Jones called SK Hynix the main supplier of high-bandwidth-memory products to Nvidia, and AFP said the company dominates the HBM market and is a key supplier for Nvidia .

The U.S. revenue data also supports that link. TrendForce, citing Korean reports and a company filing, said SK Hynix’s U.S.-market revenue surged 2.6 times in 2024, fueled by rising AI memory-chip demand from major U.S. technology companies. The same report said U.S.-market sales accounted for more than half of SK Hynix’s annual revenue . That helps explain why headlines about U.S. AI infrastructure spending can quickly affect sentiment toward SK Hynix.

The rally was backed by numbers

This was not only a story about future promise. Reported results gave investors something concrete to point to.

AFP reported that SK Hynix’s second-quarter 2025 operating profit climbed almost 70% from a year earlier to 9.21 trillion won, while revenue reached 22.23 trillion won — both all-time highs . Later, Morningstar/Dow Jones reported record quarterly earnings, with HBM3E shipments and higher DRAM prices cited as important drivers . The Straits Times also reported that SK Hynix logged a record operating profit of 11.4 trillion won in the September quarter and said customers had already secured its entire memory-chip lineup for 2026 .

The share-price move reflected that earnings momentum. ChosunBiz reported that SK Hynix shares topped 600,000 won for the first time in November 2025, after starting the year in the 170,000-won range and rising 262% in 306 days. The rally was steep enough for the stock to be designated an investment warning issue .

What the current story assumes

At that point, the stock was pricing in more than the latest quarter. The bullish case rests on three main assumptions.

First, Big Tech AI infrastructure spending needs to stay strong enough to keep demand flowing into AI accelerators and memory . Second, HBM supply needs to remain tight enough to support pricing and profitability, a point highlighted by TNW’s question of when supply will catch up with demand . Third, Nvidia’s AI chip cycle needs to keep translating into orders for SK Hynix, given the company’s reported role in Nvidia’s HBM supply chain .

Those assumptions may be plausible based on recent reporting, but they are still assumptions — not guarantees.

Risks investors should watch

The same chain that lifted the stock can also work in reverse.

• HBM supply could catch up. TNW framed the harder question as when supply catches demand, which would matter for pricing expectations .
• Big Tech spending could slow. If record earnings are being driven by robust AI investment from major technology companies, a slower investment pace could weaken the demand story .
• Nvidia exposure cuts both ways. Being a key HBM supplier to Nvidia is a strength, but it also makes investor sentiment sensitive to Nvidia’s AI hardware cycle .
• Valuation and volatility are no longer minor issues. A 262% rise in 306 days and an investment-warning designation show how much optimism had already been reflected in the share price .

The bottom line: U.S. Big Tech’s AI spending helped lift SK Hynix because AI data-center investment does not stop at GPUs. It creates demand for the high-bandwidth memory that sits alongside AI accelerators. As long as Big Tech capex, Nvidia’s accelerator cycle and tight HBM supply reinforce one another, investors have a reason to connect AI spending headlines to SK Hynix. If any part of that chain weakens, the same connection could become a source of volatility .