SEOUL, June 25 (AJP) - Micron Technology Inc. delivered a stunning 74 percent revenue jump for the quarter ended May and projected the AI-driven memory boom to extend well into 2027, setting the stage for another stretch of red-hot earnings for its larger South Korean peers in the memory market powering the AI age from datacenters to smart devices.
The U.S. chipmaker posted fiscal third-quarter revenue of $41.46 billion, a 74 percent jump from $23.86 billion in the previous quarter.
Looking ahead, Micron forecast fiscal fourth-quarter revenue of about $50.0 billion, with Chief Executive Officer Sanjay Mehrotra expecting tight supply-demand conditions for both DRAM and NAND to persist beyond calendar 2027.
Beyond the upbeat outlook, Micron unveiled what may prove a more consequential shift for the memory industry.
The company said it has signed 16 long-term Strategic Customer Agreements (SCAs), securing approximately $22 billion in customer financial commitments.
"We expect these SCAs to significantly enhance the durability and predictability of Micron's strong financial performance," Mehrotra said during the earnings call.
The move toward take-or-pay contracts gives Micron unprecedented visibility over future demand through 2030, marking a structural shift from the industry's traditionally cyclical, spot-driven business model.
The agreements signal similar arrangements for SK hynix and Samsung Electronics to pursue similar long-term customer commitments as they finance increasingly expensive capacity expansions for high-bandwidth memory (HBM) and advanced DRAM.
Unlike previous memory cycles, where producers relied heavily on volatile spot pricing and short-term contracts, the AI era is encouraging chipmakers and hyperscale customers to secure supply years in advance.
Micron said its HBM capacity remains fully booked through 2026, with customer discussions already extending into 2027, reflecting sustained demand from AI datacenters and next-generation computing infrastructure.
The development reinforces expectations that the current memory upcycle may prove more durable than previous booms, supported less by consumer electronics demand than by long-term AI infrastructure investment from hyperscalers, cloud providers and enterprise customers.
Copyright ⓒ Aju Press All rights reserved.
