New York, Feb 17, 2026, 10:07 EST — Regular session underway.
- Micron slipped roughly 3% early as chip and tech names struggled in U.S. trade.
- Micron’s report outlined a years-long expansion plan targeting shortages in AI memory.
- Investors are eyeing whether big outlays will stretch the profit run, or just set up the next downturn.
Micron Technology, Inc. (MU.O) slid 2.6% to $400.88 early Tuesday, tracking losses across the chip sector. Investors reacted to renewed chatter about a huge capacity ramp linked to artificial-intelligence demand. The iShares Semiconductor ETF (SOXX) dropped roughly 1.9%. The Invesco QQQ Trust (QQQ) lost about 1%.
Micron’s fortunes are now tied closely to the AI data-center rush—not via processors, but thanks to memory. High-bandwidth memory, or HBM, is stacked DRAM positioned near AI chips to boost data speeds. Supply has stayed tight, which has lifted pricing throughout the memory segment.
Pressure from tighter conditions is now meeting sharper nerves over AI budgets. “The markets are taking each sector one-by-one and stress testing their business models,” Axel Botte, head of market strategy at Ostrum Asset Management, told Reuters. Botte’s comments came in a report that also highlighted worries about firms pouring too much cash into AI. (Reuters)
According to a Wall Street Journal piece via LiveMint, Micron is moving forward with a $200 billion slate of U.S. projects—among them, a $50 billion build in Boise, Idaho, and a massive $100 billion fab cluster outside Syracuse, New York. Production is set to run out toward 2027 and 2028. “I’ve never seen anything so disruptive as AI,” said Scott Gatzemeier, vice president at Micron. CFO Mark Murphy, speaking at an investor event, described the company’s business as being on “an extraordinary trajectory.” Brad Gastwirth of Circular Technology added, “We’re nowhere near the end of the shortage.” (mint)
HBM is drawing attention right now. It’s tougher to produce than standard DRAM, eating up more manufacturing capacity with each bit. That slows down supply adjustments and amps up the pressure whenever a production call gets made.
Micron has some short-term manufacturing progress outside the U.S. as well. In India, S. Krishnan, secretary at the electronics ministry, told an AI summit in New Delhi that Micron’s assembly, test, marking and packaging facility in Sanand, Gujarat, is gearing up to begin operations. “Micron would be starting production at their facility in India,” he said. (The New Indian Express)
Still, there’s an old pitfall here. Memory runs in cycles, and a wave of expansion can flip a shortage into oversupply fast—especially if demand for cloud or AI data centers falters, or competitors boost capacity sooner than anticipated. That pricing muscle? It could vanish just as fresh fabs come online.
Back in December, Micron’s earnings pointed to a sharp turn in the cycle. “Micron delivered record revenue and significant margin expansion,” said CEO Sanjay Mehrotra. Their forecast then? Non-GAAP gross margin around 68% for the fiscal second quarter. (Micron Technology)
The $200 billion headline number isn’t fresh. Micron had already mapped out about $200 billion in U.S. spending for manufacturing and R&D in Idaho, New York, and Virginia, according to a 2025 announcement detailing the chipmaker’s expanded domestic ambitions. (Micron Technology)
Traders have their eyes peeled for two things: hints that memory prices might be peaking, and signs that Micron is keeping a tight rein on its spending while demand remains strong. The company’s next earnings release lands around March 19, if Zacks’ earnings calendar is any guide. (Zacks)
