Micron Technology has crossed above Meta Platforms in total market capitalization, a milestone driven by explosive AI infrastructure spending that has supercharged demand for HBM and high-density DRAM. MU's most recent fiscal year showed revenue of $37.4B — up nearly 49% year-over-year — with gross margins recovering to 39.8% and diluted EPS of $7.59, a dramatic improvement from the trough of the memory cycle.

The comparison is striking because Meta is a far larger revenue generator ($201B, +22% YoY) with a 30% net margin and $23.49 in diluted EPS — roughly three times Micron's earnings per share. The market-cap flip reflects how aggressively investors are pricing AI memory upside into MU relative to what is effectively a mature, highly profitable social/advertising platform.

The bull case for MU rests on the HBM supply constraint narrative: Micron is one of only three companies (alongside Samsung and SK Hynix) capable of supplying high-bandwidth memory at scale, and AI model training and inference are structurally memory-hungry workloads. If AI capex from hyperscalers continues to accelerate, MU's pricing power and margin expansion could justify the re-rating.

The bear tension is real: memory is a cyclical commodity business, and MU's net margin of 22.8% — while recovered — is still well below Meta's 30%, on less than a fifth of Meta's revenue. Any demand softening, oversupply from Samsung, or capex pullback from hyperscalers could compress MU's multiples quickly. META, meanwhile, trades on durable advertising cash flows with a growing AI-monetization story of its own.

What to watch: MU's next earnings print for HBM pricing and volume commentary, any shift in hyperscaler capex guidance, and whether META's own AI infrastructure buildout starts to re-attract valuation multiple expansion.