Micron CEO Sanjay Mehrotra announced a fundamental shift in the economics of the semiconductor industry. In his view, memory is no longer a cyclical commodity but has become a strategic resource of the artificial intelligence era.
Speaking at the site of a new semiconductor plant in Boise, Idaho, Mehrotra outlined a key thesis: the traditional model, where high demand triggered the construction of new capacity and overproduction crashed prices, has lost its relevance. AI has created a fundamentally different, more sustainable level of consumption.
Memory as Infrastructure
According to Micron's CEO, customers no longer choose suppliers solely by price, as with ordinary commodities. Chips are now tailored to specific processors and systems, and the performance of an entire server critically depends on this pairing.
"That's why I call memory the strategic infrastructure of the AI era. Today, it's impossible to imagine artificial intelligence without memory. AI-based systems require more and more memory," Mehrotra emphasized.
The Numbers Speak for Themselves
Micron's financial reports fully confirm these statements. Revenue for the third fiscal quarter soared to $41.46 billion, compared to $9.30 billion in the same period a year earlier. Gross margin reached an impressive 84.6%, up from 37.7% a year ago, and the company expects to hit around 86% in the fourth quarter.
The main challenge now is production capacity. Mehrotra admitted that factories cannot keep up with demand: data centers alone require about 50% more chips than Micron is capable of producing. The company is already implementing an ambitious plan with $250 billion in investments in manufacturing and research within the United States.
Among future demand drivers, Micron's CEO cited manufacturers of autonomous vehicles, robotics, and smart home appliances with AI elements.
The Flip Side of the Coin
However, not everyone shares this optimism. In June, a class-action lawsuit was filed against Micron, Samsung, and SK Hynix: plaintiffs accuse the manufacturers of collusion and artificially inflating DRAM memory prices. Chinese companies, including CXMT, are also ramping up production, which could bring back intense price competition—the very thing Mehrotra says the industry has moved away from thanks to AI.
My analysis: Mehrotra's statements are not just corporate optimism but a marker of a paradigm shift. AI has indeed created a structural shortage, but history teaches us that high margins always attract new players. The question is not whether cyclicality will return, but when exactly and how severe it will be. Investors should closely monitor capacity expansion in China and regulatory actions—these factors could quickly change the current picture.