Micron’s record quarter spotlights memory as AI bottleneck

Micron reported fiscal third-quarter revenue of $41.5 billion and adjusted earnings of $25.11 per share. The company said customer commitments of $22 billion and remaining performance obligations tied to those deals amount to about $100 billion. Micron’s shares rose roughly 12% in after-hours trading and its market value moved above $1 trillion.

The $22 billion in commitments came from agreements with 16 strategic customers across data centers, consumer devices and automotive applications. Micron described contract terms that include take-or-pay provisions, cash deposits and pricing floors, which provide clearer demand visibility and some protection if market conditions weaken.

High-bandwidth memory, or HBM, featured in Micron’s reporting and analyst commentary. HBM sits next to advanced processors and delivers the high data throughput and low latency that large AI training and inference workloads require. Micron said the new contracts are aimed at securing capacity for those memory products.

Industry figures reacted to the earnings and contract disclosures. Daniel Newman of Futurum Group argued that current AI infrastructure buildouts are larger than commonly assumed and that memory should sustain higher pricing while supply remains constrained. D.A. Davidson analyst Gil Luria raised his price target on Micron to $2,000 from $1,500, citing improved visibility from long-term customer agreements. Art Hogan at B. Riley Wealth characterized the companys valuation milestone as reflecting investor reassessment of memory’s role in AI infrastructure.

Micron noted it will continue to increase capital spending to expand capacity. The company said the contract structures, including deposits and minimum-volume commitments, shift some demand risk to buyers and create a more predictable revenue stream for Micron compared with prior memory cycles.

Modern AI models move and store large volumes of data during training and inference, which increases demand for memory that can deliver both high bandwidth and low latency. As AI deployments grow across cloud data centers, edge devices and vehicles, memory capacity and supply chains have become a focal point for chipmakers and their customers.

Micron reported the revenue, profit and customer agreements alongside its quarterly results. Investors responded with a stock rally and several analysts updating forecasts and price targets based on the companys disclosed contract backlog and expected demand for high-bandwidth memory.