Sanjay Mehrotra, CEO of Micron, has declared a fundamental shift in the economics of the memory market. In his view, the era of artificial intelligence has transformed memory from a cyclical commodity into a strategic resource that determines the performance of all AI infrastructure.
Speaking at the site of a new semiconductor plant in Boise, Idaho, Mehrotra outlined a radically changed industry landscape. Micron, it should be noted, is executing a massive capital investment plan of $250 billion in U.S. manufacturing and research, which in itself speaks to long-term confidence in demand.
Why Mehrotra calls memory infrastructure
Traditionally, the memory market operated on a cyclical basis: a surge in demand triggered the construction of new capacity, followed by oversupply and price collapses. Mehrotra argues that AI has shattered this model. Now, customers can no longer choose a supplier solely by price tag, as with a typical commodity.
Memory chips are increasingly designed for specific processors and systems. The overall performance of a server now directly depends on how efficiently memory is linked to compute cores. "That's why I call memory the strategic infrastructure of the AI era," he stated. "Today, it's impossible to imagine AI without memory. AI-based systems require ever more memory."
What the numbers show
Micron's financial reports fully support these claims. Revenue for the third fiscal quarter soared to $41.46 billion, compared to $9.30 billion in the same period last year. Gross margin reached 84.6%, up from 37.7% a year earlier, and the company forecasts around 86% for the fourth quarter.
The bottleneck now is production. Mehrotra acknowledged that plants do not cover customer demand: data centers alone require approximately 50% more chips than Micron produces. And this is just the beginning. Among future buyers, he cited manufacturers of autonomous vehicles, robots, and AI-powered home appliances.
However, not everyone shares this optimism. In June, a lawsuit was filed against Micron, Samsung, and SK Hynix, with plaintiffs accusing the manufacturers of collusion and price inflation on DRAM memory. Chinese companies like CXMT are also ramping up output, which could bring back the fierce price competition—the very thing that, according to Mehrotra, the industry has moved away from thanks to AI.
My take: Mehrotra's arguments about a structural change in demand look convincing, but history teaches us caution. AI does create new, more sustainable demand, yet growing competitor capacity and legal risks could quickly bring cyclicality back to this market. Investors should closely watch the balance between ambitious plans and actual production capabilities.