AIToday

Chip stocks plunge 7%–9% on AI demand slowdown fears

Yahoo Finance AI2d ago4 min read
Chip stocks plunge 7%–9% on AI demand slowdown fears

Key takeaway

Semiconductor stocks fell sharply—with the broader index dropping over 7%—after reports surfaced that large cloud providers such as Meta may have over-invested in AI infrastructure and could be reducing future spending. A Citi analyst also questioned whether the returns on AI capex justify continued high spending rates. The selloff reflects a potential shift away from the two-year assumption that GPU and memory demand would remain limitless.

Summaries like this, in your inbox every morning.

Sign up free →

3 Key Points

  • What happened

    The Philadelphia Semiconductor Index dropped over 7%, with Amtech falling 9.2% and IPG Photonics falling 8.6%. The selloff was triggered by concerns that large cloud providers may be slowing AI infrastructure spending, a Citi analyst's warning about returns on AI capex, and reports that Meta plans to lease out spare AI computing capacity.

  • Why it matters

    For two years, the semiconductor sector assumed GPU and memory demand would remain insatiable. If Meta—which guided to as much as $145 billion(約23兆円) of capex for the year—has spare capacity to sell, it signals hyperscalers (large cloud providers) may have over-built, which could shrink future orders for GPUs, memory, and storage chips. This appears to shake confidence in the sector's growth foundation.

  • What to watch

    The market treats this as significant but not yet fundamental—Amtech's stock moves greater than 5% occur frequently (72 times over the last year). A similar shock hit 9 days ago when SK Hynix's slower memory expansion spooked the sector; the Philadelphia Semiconductor Index opened down roughly 7% that day as well.

FAQ

What specifically triggered the semiconductor selloff?
A Citi analyst questioned whether large cloud platforms would continue high rates of spending on AI infrastructure if they could not show investors the cost was generating returns. Additionally, reports that Meta planned to sell access to its AI computing power sparked fears of future overcapacity in the industry.
Why does Meta's spare capacity matter to chip makers?
If Meta—guided to as much as $145 billion(約23兆円) of capex for the year—has enough spare capacity to lease out, it signals hyperscalers may have over-built. This suggests future orders for GPUs, high-bandwidth memory, and NAND could shrink.

Discussion

No discussion yet for this article

Stay ahead with AI news

Get curated AI news from 200+ sources delivered daily to your inbox. Free to use.

Get Started Free

Free · takes 30 seconds · unsubscribe anytime

1 minute a day. The AI essentials.

200+ sources · Email / LINE / Slack

Get it free →