Markets News
StocksSeptember 15, 20262 min read

Memory Stocks Slide as AI Leaders Call for a Slower Race

Micron, SanDisk and SK Hynix fell sharply as investors reassessed the infrastructure spending underpinning the AI boom.

Micron Technology shares fell about 6% to $916.03 in early trading Monday, while SanDisk dropped roughly 6% to $1,531. SK Hynix’s U.S.-listed shares sank 7% to $175.98, extending a global retreat in the memory complex as investors questioned whether the AI buildout can keep accelerating at its current pace.

The immediate trigger was unusual: executives at the companies driving frontier-model development began arguing that the industry should ease off the throttle. Anthropic Chief Executive Dario Amodei called for a coordinated slowdown in a weekend essay, while OpenAI CEO Sam Altman backed the idea. Elon Musk also joined the push, turning what might have been a policy debate into a direct challenge to the assumptions embedded in AI-stock valuations.

Memory suppliers sit close to the center of those assumptions. Micron sells high-bandwidth memory used alongside advanced AI accelerators, while SK Hynix is one of the market’s largest HBM suppliers. SanDisk is more exposed to NAND flash and solid-state storage, but its investment case has increasingly included AI data centers, where inference workloads require large stores of model weights, cached data and user interactions.

That makes the stocks unusually sensitive to any hint that model training, deployment or data-center construction might be delayed. Micron has said the expansion of AI data-center capacity is driving demand for high-performance and high-capacity memory, while SanDisk recently told investors that AI inference is reshaping the memory hierarchy and could push enterprise data-center flash demand toward 1.2 zettabytes by 2030.

Monday’s reaction was a valuation reset, not evidence that customers had canceled orders. The Kospi fell 3.3% as Samsung Electronics and SK Hynix slid, and U.S. chip shares followed the overnight move.

The market is pricing a narrower margin for error. Even a slower frontier-model timetable could preserve long-term demand, but it would weaken the urgency behind the next wave of capital spending. For memory makers, where earnings can swing sharply with supply, pricing and utilization, that distinction matters immediately.

MUSNDKSKHYAnthropicOpenAI

This article was produced with the help of AI technology.
Source: Yahoo Finance

Comments (0)

Log in to join the discussion.Log in

No comments yet - be the first to weigh in.