Micron reports record $54 billion quarter as AI memory boom accelerates

Micron Technology reported the largest quarterly revenue in the history of the memory-chip industry on Wednesday, capping a fiscal year in which demand for artificial-intelligence hardware transformed a famously boom-and-bust business into one of the most profitable corners of American manufacturing. The Boise, Idaho-based company said revenue for its fiscal fourth quarter, which ended September 3, reached $54.23 billion, in a filing submitted to the Securities and Exchange Commission.

The figure, disclosed in an 8-K report and an accompanying earnings release, is roughly five times what Micron generated in the same quarter a year earlier and about 31 percent above the prior quarter's already record-setting total. It caps a run of five consecutive quarterly records for the company, driven almost entirely by the surge in orders for the high-bandwidth memory chips that sit alongside processors from Nvidia and other AI accelerator makers in data centers.

The numbers

Micron's GAAP net income for the quarter was $37.70 billion, or $32.87 per diluted share, up from $3.20 billion, or $2.83 per share, in the same period of fiscal 2025. On a non-GAAP basis, net income came to $38.40 billion, or $33.42 per share. Operating cash flow for the quarter was $43.97 billion, compared with $5.73 billion a year earlier. The results, detailed in the company's investor relations announcement, beat the high end of Micron's own guidance, which had called for revenue of roughly $50 billion and non-GAAP diluted earnings near $30.73 a share.

For the full fiscal year, Micron's revenue totaled $133.19 billion, against $37.38 billion in fiscal 2025 — a more than threefold increase. Full-year GAAP net income reached $84.97 billion, compared with $8.54 billion the year before, and operating cash flow for the year was $89.68 billion. Micron's fiscal 2026 filings are among the string of 8-K reports the company has made to the SEC this year documenting successive record quarters, a pattern visible in its public filing history.

Looking ahead, Micron guided for first-quarter fiscal 2027 revenue of $61.5 billion, plus or minus $1.5 billion, with non-GAAP gross margin near 86.25 percent and non-GAAP diluted earnings of $38.15 per share, plus or minus a dollar. If achieved, that single quarter would exceed Micron's entire annual revenue as recently as fiscal 2024.

How the memory business got here

Micron makes DRAM and NAND flash memory, commodity components that for decades moved through a punishing cycle of oversupply, price collapses, plant closures and shortages. That cycle broke down over the past two years as cloud computing companies and AI labs began building data centers stuffed with graphics processors that each require enormous amounts of high-bandwidth memory, or HBM, to function. HBM chips, which stack memory dies vertically and wire them directly to a processor, carry far higher prices and margins than conventional memory.

Micron has said its entire HBM production for calendar 2026 was already sold out under multiyear, fixed-price agreements well before the year began, and that its newest generation, HBM4, is now shipping in volume to customers building AI servers. The company expects mass production of the following generation, HBM4E, to begin in fiscal 2027. To back the commitments, Micron disclosed that customers had signed 16 long-term Strategic Customer Agreements worth roughly $22 billion, including about $18 billion in upfront cash deposits, with 14 of those contracts guaranteeing a combined $100 billion in minimum future revenue.

That demand has pulled Micron's capital spending sharply higher. The company said fiscal 2027 capital expenditures would step up by more than $10 billion from fiscal 2026 levels to expand DRAM and HBM manufacturing capacity, including at its Idaho and New York sites, part of an industry-wide scramble by Micron, Samsung Electronics and SK Hynix to add capacity for a product category that scarcely existed as a distinct market five years ago.

Who is affected

The gap between AI-driven demand and available manufacturing capacity is now reaching well beyond data centers. Because DRAM and NAND production lines cannot be easily reallocated between HBM and conventional memory chips, the capacity manufacturers are redirecting toward AI customers is coming out of the supply that would otherwise go into personal computers, smartphones and other consumer electronics. Analysts at IDC have warned that the resulting shortage could push PC and smartphone prices higher through 2026, with forecasts of PC price increases of up to 8 percent and unit shipments declining as costs get passed to buyers. Dell, Lenovo, HP, Asus and Acer have each already announced coming price increases tied to memory costs, and industry analysts expect further increases as DRAM and solid-state-drive costs keep climbing. Budget smartphone brands, which operate on thin margins, are seen as most exposed, while Samsung and Apple have reportedly locked in 18 to 24 months of memory supply that insulates them for now. Cloud operators and AI infrastructure builders, by contrast, are the direct beneficiaries of Micron's expanded output, as are Micron's roughly 53,000 employees and its shareholders, whose stock has risen sharply over the course of fiscal 2026 alongside the company's results.

What executives and analysts are saying

Micron's chairman and chief executive, Sanjay Mehrotra, framed the results as validation of a yearslong bet that memory would become a strategic bottleneck for AI computing rather than a commodity afterthought.

"Micron delivered record fiscal 2026 results, and we expect an even stronger fiscal 2027," Mehrotra said in the company's earnings release, adding that "AI is becoming Super Intelligence, and memory enhances this intelligence and the competitiveness of our customers' platforms."

Micron has also been more explicit than most chipmakers about the severity of the supply crunch, with Mehrotra telling analysts on prior earnings calls this fiscal year that the company expects memory shortages to persist through the end of calendar 2027, and that it remains uncertain when supply will catch up with demand even after that. Wall Street analysts have responded by repeatedly raising price targets on Micron shares over the course of the year as each quarterly report has come in above the company's own guidance, a pattern that repeated again with Wednesday's results, which topped the midpoint of Micron's fourth-quarter revenue forecast by roughly $4 billion.

What happens next

Micron's own guidance suggests the current cycle has further to run: the company's forecast for the December quarter alone, at a midpoint of $61.5 billion, would mark a sixth consecutive quarterly record and represents a step up of more than 13 percent from the quarter just reported. Executives have tied that outlook to continued build-out of AI data centers by major cloud providers and to the ramp of HBM4E production next year, alongside the more than $10 billion increase in planned factory spending.

For consumers, the near-term trajectory points toward continued upward pressure on the price of computers, phones and other memory-dependent devices, with IDC and other research firms cautioning that the imbalance between AI and consumer demand is unlikely to ease before 2027. Regulators and competitors will also be watching how Micron, Samsung and SK Hynix deploy their expanded capacity, and whether the current wave of multibillion-dollar, multiyear supply contracts locks in pricing power for memory makers well beyond the current AI investment cycle.