Nvidia's Artificial Intelligence (AI) Chips Still Need Memory. Here's Why the Micron Sell-Off Has Gone Too Far.

Growth investors have bid up Micron over the past year based on its role supplying high-bandwidth memory (HBM) to AI hyperscalers. Google’s TurboQuant, a model-compression algorithm that lowers AI models’ memory requirements, could materially reduce future HBM demand and partially offset the bullish case for Micron.

Analysis

Growth investors have bid up Micron over the past year based on its role supplying high-bandwidth memory (HBM) to AI hyperscalers. Google’s TurboQuant, a model-compression algorithm that lowers AI models’ memory requirements, could materially reduce future HBM demand and partially offset the bullish case for Micron.