Micron CEO: AI Fundamentally Changes the Cyclical Logic of the Memory Industry
Nashnova编辑部
Micron CEO Sanjay Mehrotra says AI has fundamentally rewritten memory's supply-demand equation — data-center customers want roughly 150% of what Micron can commit to ship, and the boom-bust cycle may be breaking. Yet the stock trades at only ~7× forward P/E, signaling the market is not fully convinced.
Why "no memory, no AI"?
Mehrotra's core argument: AI systems demand memory that is larger, faster, and lower-power — all three at once. That turns memory from a commodity part into strategic infrastructure.
This means → memory is no longer a plug-and-play component you swap between vendors. It is becoming a system-level co-design element, tightly coupled with the processor.
The supply gap is already concrete — data-center customers want roughly 1.5× what Micron can actually deliver.
Is data-center the only demand driver?
Mehrotra listed growth vectors beyond the data center: autonomous vehicles, robotics, AI-enabled smartphones, and the emerging wave of agentic AI — AI systems that autonomously execute multi-step tasks.
In plain terms = it is not just cloud giants fighting for memory. Everything from cars to phones is adding more of it.
A striking comparison: the original iPhone carried 128 MB of DRAM. Today's flagship phones pack 8–12 GB — a 60×-plus increase, and that is just one device category.
Can long-term contracts really flatten the memory cycle?
The memory industry has historically lived and died by the spot market, with prices swinging hard on supply-demand shifts. Mehrotra says the pattern is changing: Micron has signed five-year strategic agreements with 16 customers, with more added since.
This means → customers lock in supply, Micron locks in demand. Both sides are actively dampening cyclical risk.
This reflects a deeper shift: customers increasingly pull memory into the co-design process with processors and systems early on. Vendor selection is no longer just about price.
$25 billion on U.S. soil — building what?
The interview took place at Micron's fab construction site in Boise, Idaho. The site will ultimately house two fabs, each roughly the size of ten football fields.
The first Boise fab is expected to begin wafer production by mid-2027, as part of Micron's $25 billion-plus U.S. manufacturing and R&D plan.
In plain terms = Micron is betting real capital on the thesis that the cycle will not snap back the way it used to.
Is the market buying it?
Micron currently trades at only about 7× forward earnings — a textbook cyclical-stock discount.
This means → the market's core assumption remains that memory's cycle will eventually turn, and today's boom is not sustainable.
Mehrotra's rebuttal rests on rising long-term-agreement coverage and diversifying demand sources. But the thesis ultimately needs actual supply-demand data to prove out — the gap between the CEO's conviction and the market's pricing has not yet closed.
Content is for reference only, not financial advice.