Micron’s dynamic random‐access memory (DRAM) and high‐bandwidth memory (HBM) chips have become indispensable components in the AI compute ecosystem. Unlike traditional semiconductors that focus on processing, these memory devices feed massive data streams to accelerators, cutting latency and power draw. In December, Micron announced that its entire HBM output for calendar 2026 was fully contracted weeks before year‐end 2025, reflecting an order book of more than $14 billion at premium pricing. With only two peers worldwide producing similar products, the company is leveraging structural supply constraints to command price increases of 30–45% quarter‐over‐quarter for its cutting‐edge modules.

To relieve supply bottlenecks and capture long‐term revenue growth, Micron has committed over $200 billion to expand U.S. production capacity. The plan includes a major extension of its Virginia fabrication complex and two new gigafab campuses in Idaho and New York slated to come online in phases from 2027 through 2030. Additionally, the company signed a letter of intent to acquire a Taiwanese wafer fab for $1.8 billion, securing advanced DRAM process nodes. These initiatives aim to double the company’s bit output by fiscal 2029, targeting an annual run‐rate above 400 exabytes of memory storage.

Fiscal Q1 2026 results underscored Micron’s momentum, with revenue soaring 57% year‐over‐year to $13.6 billion, driven by 69% growth in DRAM sales and a 40% uptick in HBM revenue. Non‐GAAP gross margin expanded by 1,730 basis points to 45.5%, while quarterly free cash flow exceeded $8.4 billion, a record. Management’s Q2 outlook calls for revenue growth of 132% year‐over‐year and gross margins near 48%. Despite this explosive growth, the shares trade at just 5.5 times forward sales and 11 times forward earnings—ratios well below long‐term peers— positioning Micron as a compelling buy for investors banking on the AI memory shortage persisting through 2028.