Nvidia, AMD and Intel stocks are crashing: here’s why this selloff is different

Nvidia, AMD and Intel stocks are sliding sharply on Monday, but investors are not reacting to another disappointing semiconductor earnings report.

The catalyst is more fundamental, as AI executives are questioning whether the race to build increasingly powerful models should continue at its current speed.

Nvidia fell more than 2% before the bell, while AMD dropped about 5% and Intel nearly 6%. Nasdaq 100 futures were down 1.72%.

If frontier AI development is paced, some data-centre computing demand could arrive more slowly.

Yet chip orders remain strong, leaving Wall Street to decide whether Monday’s sell-off is a reset or an overreaction.

Wall Street is questioning the speed of the AI cycle

Nvidia, AMD and Intel have benefited from expectations that AI workloads will require more computing power steadily.

Anthropic CEO Dario Amodei called on AI companies over the weekend to slow the pace at which model capabilities advance. OpenAI CEO Sam Altman and xAI’s Elon Musk broadly backed more restraint.

That introduces a new variable: what happens if the industry intentionally slows frontier progress?

ServiceNow, Adobe and Workday rose in premarket trading as investors reconsidered how quickly AI might disrupt software.

Brian Jacobsen, chief economist at Annex Wealth Management, told Reuters that “the strongest arguments for caution are those grounded in evidence, not fear,” while warning investors about incumbents protecting their positions.

Monday is not primarily questioning chip demand, but how quickly future demand arrives.

Slowing US AI may be harder than the sell-off assumes

The counterargument is that American labs cannot unilaterally stop the global AI race.

China continues investing in models, chips and sovereign computing infrastructure, while Washington focuses on preserving technological leadership.

Technology analyst Dan Ives said Amodei’s proposal was an important step towards industry self-regulation, but added that “the reality is China won’t slow down anytime soon.”

That constraint matters for Nvidia, AMD and Intel.

Pacing model releases does not mean hyperscalers cancel data centres or stop purchasing hardware.

Safety testing requires compute, while inference workloads, sovereign AI projects and deployment of existing models can keep absorbing chips if frontier gains arrive more slowly.

The market is pricing slower AI advancement, not confirmed weakness in semiconductor orders.

Chip demand was booming days before the sell-off

Four days earlier, Piper Sandler initiated bullish coverage across the AI semiconductor sector.

Analyst David O’Connor called Nvidia the “outright leader in AI compute with 80% market share” and set a $300 target.

He argued agentic AI had created a step-change in computing demand, with supply likely to remain constrained for another two to three years.

Piper also initiated AMD at Overweight with a $600 target, calling it an “Agentic AI Sweetspot” as server CPUs gain share and its Helios GPU platform ramps with OpenAI, Meta and Anthropic.

Intel received a Neutral rating and a $110 target, despite Piper saying agentic AI should support server-CPU demand.

Monday’s nearly 6% drop therefore lands on a company where Wall Street already sees greater execution and valuation risk.