Forget the doom headlines for a moment. Amid rising Treasury yields, a new Fed chair, and tariff litigation, one number flashed bright green this week: Micron Technology’s stock surged more than 16% in a single session, pushing its market capitalization above $1 trillion for the first time. The catalyst? A massive price target upgrade from UBS.
Micron makes DRAM and NAND flash memory — the kind of high-bandwidth memory (HBM) that’s essential for training and running large AI models. Every AI data center needs it. Every chip from Nvidia needs it. What’s different this cycle is that AI demand has created a structural, multi-year demand floor that analysts believe will prevent the kind of cyclical crash that historically followed memory chip bull runs.
The classic memory chip cycle: demand spikes, companies build capacity, supply overwhelms demand, prices crash. What’s changed? HBM — premium memory stacked directly on AI chips — can’t be substituted. It’s like needing premium fuel for a specific engine. As AI models grow larger, the memory required per training run grows geometrically. The largest AI labs are consuming billions in compute and memory per quarter — and that demand compounds.
Micron’s inclusion in the trillion-dollar club changes its investment profile: more institutional ownership, more index inclusion, more passive fund buying. It also means the easy-money phase may already be priced in after a 16% single-day jump. For investors seeking AI chip exposure without single-stock risk, semiconductor ETFs like SOXX or SMH provide broad exposure across Micron, Nvidia, AMD, and others.
After a 16% single-day surge, valuation has expanded significantly. Whether it’s a good buy depends on your belief in the multi-year AI chip demand story and comfort with semiconductor volatility. Consult a financial advisor before making individual stock decisions.
HBM is a type of memory chip stacked vertically and connected directly to AI processors, enabling extremely fast data transfer. It’s critical for training and running large AI models and commands a significant price premium over standard DRAM.
Nvidia makes the AI processors; Micron makes the memory those processors require. They’re complementary, not competitive. Nvidia has been the higher-profile AI winner, but Micron is increasingly recognized as an essential part of the same infrastructure buildout.