Micron Surges as AI Servers Guzzle Memory — and the Company Can’t Make it Fast Enough

CNBC, Reuters, and Bloomberg contributed to this report.

Micron just delivered the kind of earnings report that makes investors forget last week’s tech wobble.

Shares of Micron Technology jumped about 15% after the memory-chip maker crushed expectations for its fiscal first quarter and tossed out guidance that basically screamed one thing: AI demand is outrunning supply, and it’s not even close.

On the earnings call, Micron executives said data centers — especially the ones powering artificial intelligence — are driving a surge in demand for the company’s memory and storage products.

And they weren’t subtle about it. Business chief Sumit Sadana put it bluntly:

Micron is “more than sold out,” with a “significant amount of unmet demand,” and the company expects demand to stay well above supply for the foreseeable future.

A big piece of that story is high-bandwidth memory (HBM) — the premium, high-performance memory that AI accelerators and servers depend on. Micron now expects the total addressable market for HBM to hit $100 billion by 2028, growing at a 40% compound annual rate.

Micron’s quarter came in way ahead of Wall Street’s projections:

• Adjusted EPS: $4.78 vs. $3.95 expected;
• Revenue: $13.64B vs. $12.84B expected.

Then came the real attention-grabber: guidance.

• Next-quarter revenue: about $18.70B vs. $14.20B expected;
• Next-quarter adjusted EPS: about $8.42 vs. $4.78 expected.

That’s not a “slightly better” outlook — that’s Micron telling the market the pipeline is packed and pricing power is back.

Analysts didn’t waste time.

• JPMorgan raised its price target, pointing to a favorable pricing setup.
• Bank of America upgraded the stock to buy.
• Morgan Stanley went even bigger, calling Micron’s upside one of the largest they’ve seen in US semiconductor history (outside of Nvidia), and arguing that if AI keeps growing, memory could be the next big beneficiary — not just the processors.

The AI trade has been getting punched around lately, with investors questioning whether all the infrastructure spending will pay off quickly enough. Micron’s report offers a pretty direct counterpoint: the AI buildout is already showing up in real demand, real shortages, and real pricing leverage.

In other words: chips are great, but memory is where the bottleneck — and the money — may be.