The Micron AI memory chip story reached a new peak last week, with the Boise, Idaho-based memory maker briefly surpassing both Meta and Tesla by market capitalisation on Thursday before settling back. By Friday’s close, Micron held a market cap close to $1.27 trillion, while Meta sat at $1.39 trillion and Tesla at $1.42 trillion. The gap is now very much closeable. That took a while to type without laughing, given that Micron spent the better part of the previous decade trading below $100 a share.
The stock has risen over 236% in the past month alone, closing Friday at $1,132. That is not a typo.
What the Earnings Actually Said
The immediate catalyst was a set of third-quarter results that the word ‘blockbuster’ barely covers. Revenue quadrupled year-over-year to $41.45 billion, while profit surged from $1.88 billion to $28.2 billion over the same period. Micron also guided for fourth-quarter revenue of between $49 billion and $51 billion, which is the kind of forward guidance that makes analysts reach for their upgrade buttons.
Zoom out a little and the trajectory is just as steep. According to Micron’s fiscal Q4 2025 earnings call prepared remarks, full-year fiscal 2025 revenue grew nearly 50% to a record $37.4 billion, with gross margins expanding by 17 percentage points to reach 41%. Those are not the numbers of a commodity cyclical quietly muddling through.
Dig into the product mix and the AI story becomes even clearer. Citing Micron’s Q4 2025 earnings call, Yahoo Finance reported that High-Bandwidth Memory revenue reached nearly $2 billion in fiscal Q4 alone, implying an annualised run rate of roughly $8 billion. DRAM revenue hit $9 billion for the quarter, up 69% year-over-year. NAND, the less glamorous sibling, came in at $2.3 billion, down 5% year-over-year, a reminder that not every segment is riding the same wave. For Q1 2026, Micron guided to $12.5 billion in revenue, a gross margin of 51.5%, and earnings per share of $3.75.
Why the Micron AI Memory Chip Case Is About More Than One Good Quarter
The perennial bear case on memory chip makers has always been the cycle: build capacity, watch demand evaporate, sell product at a loss, repeat. Micron has been here before, and so has anyone who has owned the stock. The question Wall Street is now wrestling with is whether AI fundamentally breaks that pattern, or merely delays the next bust.
Micron’s answer is structural: long-term supply agreements. The company disclosed 16 strategic customer agreements spanning data centre, consumer, and automotive segments, framing them as a transformation of its underlying business model. The most eye-catching of these was announced on 22 June 2026, when Micron and Anthropic disclosed a multi-pillar strategic agreement covering memory and storage AI architecture design, supply and demand commitments for HBM, DRAM, and SSDs, enterprise deployment of Claude across Micron’s own operations, and a strategic equity investment by Micron in Anthropic’s Series H funding round, according to Micron’s investor relations announcement.
That last point is worth pausing on. Micron is not just selling memory to Anthropic; it is taking an equity stake in one of the most closely watched AI labs in the world. That is a different kind of relationship from a standard volume supply contract.
Nvidia is also among the key buyers, as are the hyperscalers: Microsoft, Amazon AWS, Google, Meta, and Oracle. With so much of the AI compute stack dependent on HBM and DRAM, the shortage condition nicknamed ‘RAMageddon’ is predicted to persist into 2027, pushing up prices on everything from Apple products to Xbox consoles in the process.
William Blair analyst Sebastien Naji captured the new consensus in a research note: ‘Given the strong likelihood of continued ASP growth in the coming quarters and improving revenue visibility thanks to a rapidly expanding set of long-term agreements (SCAs) with key customers, we see potential for more durable earnings growth and reiterate our Outperform rating.’
The ‘durable’ qualifier is the whole argument in one word. Memory has historically been the opposite of durable. If the SCA structure actually delivers through a demand softening, Micron will have earned the Nvidia comparison. If it does not, the stock’s history below $100 will serve as a fairly pointed reminder of how these cycles typically end.
The next real test arrives with Q1 2026 results. That $12.5 billion revenue guidance and 51.5% gross margin target will either confirm the new reality or start the conversation about whether the cycle just got postponed rather than abolished.
