Micron CEO Sanjay Mehrotra said Thursday that memory chips have become "strategic infrastructure" for AI, warning that AI models cannot scale without them. The comments came as the memory chipmaker announced a $10 billion investment in a new research facility in Boise, Idaho, and its stock rose 4% on the day.
"Memory is no longer a component in a system; memory is the strategic infrastructure for AI," Mehrotra said in an interview with CNBC's Jim Cramer. "Without memory, you cannot make AI smarter or faster, without memory you can't scale up AI. AI is advancing and a lot of context is getting generated. All of that content has to be stored, has to be processed through memory."
Micron's $250 billion U.S. buildout
The research lab is part of a broader push to expand Micron's U.S. footprint. In July, the company outlined plans to invest more than $250 billion in U.S. manufacturing, research, and development. The new facility, called Micron Research Labs, will focus on memory technologies, advanced compute architectures, packaging, and future semiconductor manufacturing.
The initiative drew public support from Nvidia CEO Jensen Huang and Apple CEO Tim Cook. Huang called memory reinvention "one of the great challenges of the AI era" and said Micron is helping "drive the breakthroughs that will define the next era of AI and computing." Commerce Secretary Howard Lutnick described the project as "the first dedicated Memory Research Lab" in the U.S., saying it would ensure memory technology does not become a constraint on innovation.
The stakes extend beyond Micron's own balance sheet. As Generative AI and LLM systems grow more complex, they generate enormous amounts of context that must be stored and processed - which means demand for memory chips is tied directly to AI adoption. For executives planning AI infrastructure budgets, supply constraints in memory could become a strategic bottleneck.
Analyst and trader views
BMO Capital Markets initiated coverage on Micron with an Outperform rating and a $1,300 price target, implying a 33% upside from the stock's last close. BMO sees Micron benefiting from a "prolonged memory supercycle," driven by constrained supply and "exceptionally strong" demand across all product lines.
Retail sentiment on Stocktwits remained bearish, though some traders pointed to Micron's valuation. "Micron's earnings are conservatively expected to jump by 111% in fiscal 2027 to $154.89 per share," one trader said, arguing the stock could reach $2,323 if it trades at 15 times earnings.
MU stock gained 4% on Thursday and was up 0.3% in overnight trading.
Why this matters for executives and strategy
For leaders planning AI investments, Micron's positioning is a reminder that the AI supply chain runs on more than GPUs. Memory is a finite, increasingly scarce input - and its cost and availability will shape the economics of AI deployments. Companies building AI strategies should factor memory supply dynamics into their infrastructure planning, not just compute capacity. The broader AI for Executives & Strategy picture here is simple: the companies that understand the full hardware stack - including memory - will make better decisions about where to invest and when to move.
Your membership also unlocks:
