Micron Technology CEO Sanjay Mehrotra has a message for anyone hoping the memory chip crunch would ease up soon: keep waiting.

During the company’s fiscal Q1 2026 earnings call, Mehrotra made clear that tight supply-demand dynamics for both DRAM and NAND flash memory will persist well into and beyond 2026. The culprit is a familiar one: artificial intelligence is eating the world’s memory supply, and there simply aren’t enough chips to go around.

The numbers behind the squeeze

Micron posted record quarterly revenue of $13.64B for fiscal Q1 2026, up from $8.71B in the year-ago period. That’s a 56% jump, and yet the company still can’t keep pace with demand.

Customers are currently receiving only 50% to 66% of the memory volumes they actually request. If a PC maker orders 100 units of memory, they’re getting somewhere between 50 and 66. That’s a structural bottleneck rippling through the entire electronics supply chain.

Micron’s AI-related memory products, particularly high-bandwidth memory used in data center GPUs, are fully booked for 2026. SK Hynix, one of the other major memory manufacturers, has reported being sold out of HBM through the same period.

Micron plans approximately 20% growth in bit shipments for 2026, but even that expansion isn’t expected to close the gap between what the market wants and what producers can deliver.

Why AI broke the memory market

The explosion of AI training and inference workloads has created unprecedented demand for high-bandwidth memory, a specialized type of DRAM that sits on top of GPU packages and feeds data to processors at extraordinary speeds. Every new AI data center build-out from the likes of Microsoft, Google, Amazon, and Meta requires massive quantities of the stuff.

That demand doesn’t just consume HBM supply. It pulls manufacturing capacity away from conventional DRAM and NAND production, tightening supply across the board.

PC and smartphone manufacturers are now seeking memory supply commitments extending beyond 2026, essentially locking in orders years in advance because they can’t trust the spot market to have inventory when they need it.

J.P. Morgan and Morningstar analysts have both forecast that supply tightness in the memory market will extend into 2027, aligning with Mehrotra’s outlook.

What this means for tech and crypto investors

Micron, SK Hynix, and Samsung, the three companies that collectively dominate global memory production, are positioned to benefit from elevated pricing power. When customers are getting half of what they order, the leverage sits squarely with the supplier.

Micron’s revenue trajectory tells that story clearly. Going from $8.71B to $13.64B in a single year without a proportional increase in volume means pricing is doing a lot of the heavy lifting.

Consumer electronics manufacturers dealing with constrained memory supply could see reduced production capabilities and potentially lower shipment volumes for PCs and smartphones.

Micron’s 20% planned increase in bit shipments for 2026 is being described in the context of an anticipated shortfall. J.P. Morgan and Morningstar project the shortage could extend into 2027, a supply-demand mismatch that doesn’t resolve in a quarter or two.