Memory chips — the components that store the data AI models chew through — are back in the spotlight as investors debate how much room the current AI cycle has left to run.
According to a Yahoo Finance report, Citi is recommending that investors buy the dip in memory stocks, arguing that the AI cycle is gaining momentum rather than fading. The phrasing matters: "buying the dip" is analyst shorthand for treating a recent pullback in share prices as a discount rather than a warning sign. Citi's position, as reported, is that demand tied to AI capacity is still building.
That view is not the only one circulating. An opinion piece published on AOL.com argues that the best memory stock to own for the AI era is neither Micron nor Sandisk — the two names most American investors reach for first — but a Korean giant instead. It is an opinion column rather than a news report, and it points to something real about this corner of the market: memory manufacturing is concentrated among a small number of producers, several of them outside the United States.
Neither source, as provided, includes specific price moves, forecasts, or figures, so the size of any rally is not something these items establish.
The underlying logic is straightforward. Building AI capacity is not only about the processors that get the headlines. Every data center full of AI chips also needs vast amounts of high-speed memory to feed them, which turns memory makers into a second-order bet on the same buildout — and a more cyclical one, since memory has a long history of boom-and-bust pricing.
Why it matters: memory is one of the clearest places to watch whether the AI spending boom is genuinely broadening across the supply chain or concentrating in a handful of familiar names.