SK Hynix’s place in the memory business

SK Hynix is the pure-play memory rival that the “silent partner” pitch underplays. It is one of the three firms, alongside Micron Technology (MU) and Samsung, that control the global DRAM and HBM markets, and it has arguably been the leader in high-bandwidth memory since the technology’s earliest generations.

Where Micron is the only major US-based manufacturer and Samsung is a diversified conglomerate, SK Hynix is closer to a pure memory company. Its fortunes are tied, far more tightly than Samsung’s, to the price of DRAM and HBM, which makes it both the cleanest exposure to the memory cycle and the most volatile.

Why HBM leadership matters

SK Hynix was first to market with several HBM generations and has been a frequent supplier to NVIDIA’s accelerator lineup. Because NVIDIA still controls roughly 80% of the AI-accelerator market, a memory maker that supplies that ecosystem sits at the center of the AI buildout. That position is the core of SK Hynix’s appeal.

The pitch that resolves to Micron does not say much about SK Hynix, which is its own quiet irony. The memory story the pitch tells, a tight oligopoly supplying the AI boom, applies to all three firms, and SK Hynix is arguably the most exposed to it. Our 10X Project teardown makes that three-way point explicitly.

The Tesla HBM4 request

The most concrete recent data point for SK Hynix is the widely reported news that Tesla asked both SK Hynix and Samsung to develop HBM4 prototypes for its AI chips. The request matters for two reasons.

First, it shows HBM4, the successor to the current HBM3e standard, is moving from specification to sampling, with a marquee customer actively testing it. Second, it shows there is no exclusive supplier. Tesla went to two of the three memory makers, which means the HBM4 business is competitive rather than locked up. For SK Hynix, that is both a validation and a reminder that Samsung and Micron are chasing the same orders.

The pure-play trade-off

Owning SK Hynix is a bet on the memory cycle with very little else in the way. That is the appeal and the risk at the same time. When DRAM and HBM prices are strong, a pure play captures the full benefit; when the cycle turns, there is no display or phone business to cushion the fall.

Micron offers a US-based version of roughly the same exposure, vertically integrated from design through fabrication, and Samsung offers the diversified alternative. Our piece on Samsung semiconductor stock lays out the diversification side of the same choice.

The cyclical reality

SK Hynix carries the full weight of the memory cycle, and that cycle is severe. DRAM has spent decades swinging between oversupply and shortage, and the three competitors have a long history of ramping capacity at once, driving prices down, and then pulling back. SK Hynix has lived through every one of those swings.

The current hope is the HBM3e-to-HBM4 transition, which is supposed to bring better margins as each generation commands a higher price. The market-size figure the pitch leans on, roughly $33 billion for HBM by 2027, is a growth story, but it does not repeal the cycle. Our HBM3e explainer covers the handoff in detail.

The scoreboard behind the pitch

The “silent partner” pitch that anchors this whole cluster leans on a backward-looking number: Micron is described as up about 911% from a tease price of $92.50 to a recent price near $935.39, with a market cap around $1.06 trillion and roughly 15 times forward adjusted earnings. Those figures describe Micron, not SK Hynix, but they set the tone for the whole memory trade, and they are backward-looking by definition.

The pitch opened on the DeepSeek panic of late January 2025 and leaned on a promised conference call around April 23 that came and went with no supplier reveal. The takeaway for an SK Hynix investor is that the memory story is real while the urgency behind the marketing is not. The HBM4 request from Tesla is the concrete forward-looking data point; everything else in the pitch is a rearview scoreboard.

What it means for investors

For an investor, SK Hynix is the pure-play way to own the memory chokepoint of the AI buildout. The trend is real, the HBM leadership is real, and the Tesla prototype request is a concrete sign that HBM4 demand is forming. The risk is the cycle, plus the concentration of the business in a single, violently cyclical product category.

The useful frame is to treat SK Hynix as the cleanest, most volatile expression of the memory thesis, and to size the position for the boom-and-bust that comes with it. If you want the same thesis with a broader cushion, Samsung is the alternative; if you want a US-based pure play, Micron is the name. The pitch points at one of the three, but the story belongs to all of them.

Ready to see the research? Click here to access the Near Future Report.

NewsletterVetter is an independent publication. We receive compensation from some of the services we review through affiliate links. Nothing on this site is investment advice. Always do your own research.