Record-smashing earnings and an 86% gross margin forecast suggest consumers will foot the bill for AI's insatiable appetite for memory.
Micron Technology just posted the kind of quarter that makes Wall Street analysts look like they’re playing darts blindfolded. The chipmaker reported fiscal Q3 2026 revenue of $41.5 billion against expectations of $35.7 billion, alongside adjusted earnings per share of $25.11 versus the roughly $20.5 consensus. Shares surged approximately 16% in premarket trading.
The AI memory squeeze is just getting started #
Micron’s Q4 revenue guidance landed in the $49 billion to $51 billion range. Analysts had been expecting roughly $43.2 billion.
Projected gross margins for the upcoming quarter sit at around 86%.
CEO Sanjay Mehrotra pointed to the root cause during the earnings announcement on June 24: supply constraints in the high-bandwidth memory (HBM) market are expected to persist beyond 2027. When demand consistently outstrips supply in any market, the entity selling the scarce good gets to name its price. Micron is currently that entity, alongside competitors like SK Hynix, which also saw its shares climb on the back of Micron’s results.
Anthropic deal and the broader AI buildout #
Two days before the earnings report, on June 22, Micron announced a strategic partnership with Anthropic. The deal includes a memory and storage supply agreement alongside an investment in Anthropic’s Series H funding round.
Tokenized shares bridge chip profits to crypto markets #
Micron launched tokenized shares on the Solana blockchain roughly two days before the earnings report. The tokenized MU shares allow for round-the-clock trading, effectively untethering Micron’s stock from the constraints of traditional market hours.
The timing was arguably strategic. Anyone holding tokenized MU shares could react to the after-hours earnings beat in real time, rather than waiting for the next morning’s opening bell.
What this means for investors #
The risk sits on the demand side. Memory chips are notoriously cyclical. The current upcycle is historically extreme, but cycles are called cycles for a reason.
Whether the tokenized shares convergence accelerates or stalls will depend largely on regulatory clarity around tokenized securities, a question that remains very much unresolved heading into the second half of 2026.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our