• Micron Technology Inc. warns that the global memory-chip shortage will persist beyond 2026, driven by insatiable AI demand and structural supply gaps.
  • The company is prioritizing enterprise AI customers over consumer markets, exiting certain product lines and accelerating capital expenditure.
  • The prolonged tightness is expected to keep pricing elevated for DRAM and NAND, affecting data-center builds and device costs.

Micron Technology Inc. signaled that the global memory-chip shortage will persist beyond 2026, as AI-driven demand continues to outpace capacity expansion. The warning comes amid a broader industry shift toward specialized memory for AI workloads, with Micron pivoting away from consumer segments to focus on high-value enterprise customers.

“We see demand for memory in AI data centers remaining robust for years to come, and supply will struggle to keep pace,” CEO Sanjay Mehrotra said during a recent investor call. He emphasized that even with aggressive investment in new fabrication plants, the supply-demand imbalance is structural, not cyclical.

Micron has reduced its exposure to consumer memory, discontinuing certain brands to reallocate production capacity. The company plans to boost capital expenditure significantly in 2026, with a particular focus on high-bandwidth memory (HBM) used in AI accelerators. A person familiar with the matter said Micron is in talks to acquire a site in Taiwan to accelerate DRAM output.

“The memory industry is undergoing a paradigm shift,” said Dan Hutcheson, an analyst at TechInsights. “It’s not just about capacity; it’s about the type of memory being produced. AI requires specialized, high-performance chips that take longer to manufacture.”

Market data reflects the tightening. Spot prices for DDR5 DRAM have risen 15% year-to-date, while NAND flash contract prices are up 8% in the first quarter. Analysts expect pricing to remain elevated through 2027, as new fabs take 18–24 months to reach full production.

Correction: An earlier version of this article misstated the timeline for a new fab in Taiwan; the deal is still under negotiation. This article has been updated to reflect that.

The shortage has broader implications. Hyperscalers like Microsoft (MSFT) and Amazon (AMZN) are facing longer lead times for server deployments, while PC and smartphone makers are bracing for higher component costs. Micron’s shift away from consumer memory may exacerbate shortages in those segments, though the company argues the enterprise focus is necessary to support the AI ecosystem.

Contacted for comment, a spokesperson for Micron declined to elaborate beyond the CEO’s public remarks. The company’s next earnings report, due in April, will likely provide further clarity on capex plans and customer commitments.