TL;DR · 30-second read
The Short Version
The United States government closed a gap that had let China get hold of advanced computer chips. Shares of companies that make chips for artificial intelligence fell, and a broad index of 500 large American companies dropped 2%.
Micron, America’s biggest maker of memory chips, has about $14.6 billion in sales tied to China. How much of that survives now depends on government permission, not on whether Chinese customers want to buy.
Separately, Nvidia, the best-known maker of artificial intelligence chips, said its leader of 33 years resigned on November 11.
AP reported that further drops in Nvidia and other AI stocks pulled the S&P 500 down 2%. Boise Dev reported that Micron shares tumbled after the Trump administration closed a loophole that had allowed China access to advanced chips, and 24/7 Wall St. put Micron’s China-linked revenue at $14.6 billion. Yahoo Finance reported that AMD and Broadcom sank on the tighter rules and that AI cloud providers CoreWeave and Nebius also fell; Quiver Quantitative reported that Monolithic Power Systems declined on chip-sector weakness and the new export curbs.
The selloff follows CoreWeave’s November 10 third-quarter report, which raised full-year revenue guidance, and a November 12 Nvidia filing disclosing that CEO Jensen Huang had resigned, with CFO Colette Kress named interim chief executive.
Executive Summary
Washington has closed another route by which Chinese buyers could reach advanced US chips, and markets repriced the AI hardware chain in response. Micron is the clearest exposure: it has $14.6 billion in China-linked revenue, and for whatever share of that the tightened rule reaches, the deciding factor is now a US export license rather than customer demand.
The timing sharpens the point. Days earlier, CoreWeave reported a $132.5 billion revenue backlog, more than 4.6 gigawatts of contracted power and raised guidance — evidence that AI infrastructure demand outside China remained strong. The selloff reflects a change in who is allowed to buy, not a reported fall in how much buyers want.
For data center operators, cloud buyers and investors, the practical takeaway is that policy risk now sits alongside power, supply and financing as a variable in AI infrastructure planning. Nvidia’s leadership change adds a second uncertainty ahead of its November 17 earnings report.
For China Sales, the Export License Is the Order That Counts
Export controls act as a gate in front of the order book. Once a chip falls into a controlled category, selling it to a Chinese buyer requires a license from the US government, and without that license the sale cannot ship regardless of what the customer will pay. That is the mechanism behind this story: for whatever slice of Micron’s $14.6 billion in China-linked revenue the tightened rule touches, the variable that matters is Washington’s licensing posture, not customer appetite.
AI memory already works this way. Since December 2024, US rules have restricted exports to China of high-bandwidth memory (HBM) — stacks of DRAM chips mounted beside an AI processor to feed it data fast enough. Beijing runs its own gate: in 2023 China’s cyberspace regulator told operators of critical infrastructure to stop buying Micron products after a security review. Micron’s China business has been shaped by policy on both sides for years; closing another access route narrows it further.
The $14.6 billion is a ceiling on exposure, not an estimate of loss. China-linked revenue can include products no export rule covers, such as conventional memory for phones and PCs. The fair reading is that a material revenue stream now depends on license decisions Micron does not control — which is a different risk from a slowdown in orders, and one that no amount of AI demand can offset.
The Demand Data Pointed the Other Way
Days before the selloff, CoreWeave — a specialist AI cloud provider, sometimes called a neocloud — reported third-quarter numbers that show no softening in AI infrastructure demand. Its November 10 earnings release put its revenue backlog (contracted revenue not yet delivered) at about $132.5 billion as of September 30. It added more than 200 megawatts of active power to exceed 1.7 gigawatts, grew contracted power by more than 900 megawatts to over 4.6 gigawatts, raised full-year revenue guidance to $13.0–$13.3 billion and lifted the low end of its capital spending forecast to $37–$39 billion because it expects to deliver infrastructure to customers sooner than planned.
The same release noted a multi-year agreement with Solidigm securing priority access to enterprise SSD capacity. When a large buyer signs contracts for priority access to storage, it signals that memory and storage supply is contested outside China. That matters for Micron: whether restricted China volume can be redirected to non-Chinese buyers is the difference between lost revenue and reshuffled revenue — a question the company has not yet answered.
Set side by side, the evidence supports a narrow conclusion: what changed is the set of permitted buyers, not the depth of the order books reported by the companies building AI capacity.
Why the Selloff Reached Power Chips and AI Clouds
The decline spread well beyond memory: AMD (AI accelerators), Broadcom (custom AI chips and networking silicon), Monolithic Power Systems (power-management chips for servers), and the AI clouds CoreWeave and Nebius. Their common thread is that valuations rest on the total volume of AI buildout worldwide. When policy removes or restricts a buyer region, investors re-examine the whole chain and price in the possibility of further rules.
For the cloud providers, the link is less direct. CoreWeave’s disclosures do not tie its business to China, and its customers named in the third-quarter release are US and European AI labs and enterprises. It has, however, broadened its international footprint to more than 1.3 gigawatts of contracted power, including India, Indonesia and Spain. Absent any company statement that the rule touches those deployments, the moves in CoreWeave and Nebius read as sector-wide repricing rather than a change in their own contracted demand.
Nvidia Faces Policy Pressure and a Leadership Change at Once
Nvidia disclosed in an 8-K filed November 12 that its board accepted Jensen Huang’s resignation as president, CEO and director on November 11, effective immediately. CFO Colette Kress is serving as interim CEO while keeping the finance role, and longtime director Harvey C. Jones was named the company’s first-ever board chair. Nvidia did not give a reason for the resignation.
In Nvidia’s announcement, Jones said the board expects operations, customer commitments and the product roadmap to continue without interruption. For infrastructure buyers, roadmap continuity is the operative issue: CoreWeave reported bringing multi-rack Vera Rubin NVL72 systems online in the quarter. Nvidia reports third-quarter fiscal 2027 results on November 17, the first opportunity for its interim leadership to address the export rule’s effect on its own business.
Background
Since October 2022, the US has used export controls to limit China’s access to advanced AI chips and the tools to make them, tightening the rules several times, including restrictions on high-bandwidth memory in December 2024. Micron, headquartered in Boise, Idaho, is the largest US maker of memory chips, producing DRAM, NAND flash and HBM for AI systems; in 2023 China’s cyberspace regulator also told critical-infrastructure operators to stop buying Micron products.
CoreWeave, based in Livingston, New Jersey, rents GPU computing capacity for AI workloads and reports customers including OpenAI, Cohere and Perplexity. Nvidia, the dominant supplier of AI accelerators, had been led by Jensen Huang for 33 years before his November 11 resignation. Source: More drops for Nvidia and other AI stocks drag the US market lower, pulling the S&P 500 down 2% — AP News on the AI-led market decline. Related coverage: Boise Dev, 24/7 Wall St., Yahoo Finance, Quiver Quantitative. Primary sources: CoreWeave Form 10-Q for the quarter ended September 30, 2026 (filed 2026-11-12); CoreWeave Form 8-K (filed 2026-11-10); CoreWeave Reports Strong Third Quarter 2026 Results (Exhibit 99.1); Nvidia Form 8-K (filed 2026-11-12); NVIDIA Announces CEO Transition (Exhibit 99.1).Sources

