Skip to content
Analysts link 1.5 GW of smuggled Nvidia chips to Huang’s departure
Depth
1 outlet · newest source 194h ago
AI hardware

Analysts link 1.5 GW of smuggled Nvidia chips to Huang’s departure

SemiAnalysis estimates massive illicit GPU exports to China may have contributed to the CEO’s sudden exit from the chipmaker.

Key points

  • SemiAnalysis estimates over 1.5 gigawatts of restricted Nvidia hardware reached China.
  • The firm suggests this smuggling, alongside personal scandals, may have driven Jensen Huang’s departure.
  • The estimate excludes compliant chips like the H20 and H800, focusing on banned models.

SemiAnalysis, a semiconductor research firm, published a newsletter claiming that widespread illicit exports of advanced AI accelerators to China may have driven Jensen Huang’s departure from Nvidia. The analysts argue that this factor, combined with revelations about his personal relationships, could explain the sudden leadership change. according to the published report, on 2026-11-14, a Guardian article had asserted the CEO stepped down, though that event appeared fabricated and lacked independent corroboration at the time.SemiAnalysis correlates 1.5 GW of illicit GPU exports with Jensen Huang’s departure, citing potential compliance issues.

Evidence: reportedMass smuggling of up to 1.5 million Nvidia AI GPUs could be the underlying cause of Jensen Huang's departure, analysts claim

aifleet-tomshardware

What changed

According to Tom's Hardware, SemiAnalysis believes Huang’s departure may have been driven not only by personal scandals but also by illicit exports. The firm points to staggering volumes of banned hardware moving into China. They estimate that more than 1.5 gigawatts of restricted Nvidia hardware have reached the country despite U.S. export controls. This figure represents the total power capacity of the smuggled chips, not just their count.

Evidence: reportedMass smuggling of up to 1.5 million Nvidia AI GPUs could be the underlying cause of Jensen Huang's departure, analysts claim

aifleet-tomshardware

Evidence: reportedMass smuggling of up to 1.5 million Nvidia AI GPUs could be the underlying cause of Jensen Huang's departure, analysts claim

aifleet-tomshardware

How it works

SemiAnalysis’ Accelerator Model tracks AI accelerator production and destinations to derive this estimate. The calculation likely multiplies package-level rated power by the number of GPUs, though the report does not explicitly detail the conversion method. If they were B200 SXM units rated at 1,000 watts, the number would be around 1.5 million units.

Evidence: reportedMass smuggling of up to 1.5 million Nvidia AI GPUs could be the underlying cause of Jensen Huang's departure, analysts claim

aifleet-tomshardware

What the evidence means

The estimate excludes processors Nvidia developed specifically to comply with U.S. rules, such as the H20 or H800. SemiAnalysis does not disclose the exact number of accelerators shipped illegally or their split between product generations. The firm claims the U.S. Department of Commerce has reached a similar conclusion about the scale of unauthorized shipments, but provides no independent verification of that agency's findings. The data remains an analyst estimate, not a confirmed government statistic.

Evidence: reportedMass smuggling of up to 1.5 million Nvidia AI GPUs could be the underlying cause of Jensen Huang's departure, analysts claim

aifleet-tomshardware

Cost and market position

For product teams, this volume represents an enormous amount of computing infrastructure. This suggests that significant Nvidia compute power is operating outside U.S. regulatory oversight, potentially altering market dynamics for AI training resources in restricted regions.

Why it matters

  • Product teams must understand that export controls may be less effective than assumed, with large-scale smuggling operations persisting.
  • The comparison to xAI’s Colossus clusters highlights the sheer scale of illicit compute, which could impact global GPU availability and pricing.
  • Analysts linking corporate leadership changes to illicit exports signals heightened scrutiny on executive accountability.
  • The exclusion of compliant chips from the estimate suggests a clear market distinction between legal and illegal hardware flows.

Topics: Nvidia, export controls, AI infrastructure, SemiAnalysis, Jensen Huang

Updated

For readers

Business

  • The scale of smuggling rivals major AI data centers, indicating robust black-market supply chains.

Investors

  • Leadership changes linked to illicit exports may signal increased scrutiny for chipmakers.

Policy

  • The estimate suggests current U.S. export controls are being circumvented on a massive scale.

Ask the article

Local keyword search of this article; no generated answers.

Did this article give you the evidence you needed?

No vote saved on this device.
Your choice stays on this device. No public tally.

Questions readers ask

Does this include Nvidia’s China-compliant chips?

No, the estimate excludes chips like the H20 and H800 designed to meet U.S. rules.

Is the U.S. government confirming these numbers?

SemiAnalysis claims the Department of Commerce reached a similar conclusion, but provides no proof.

Connected stories

This storyNvidia urges cloud providers to reduce Chinese client exposureNvidia urges cloud providers to NYT says Trump briefed on Nvidia CEO’s affair year ago, contradicting vague timelineNYT says Trump briefed on NvidiaNew York Times reports Nvidia CEO Jensen Huang had affair with VP Kimberly PowellNew York Times reports Nvidia CENvidia probe expands as reports link CEO relationship to $7.5B in Chinese chip salesNvidia probe expands as reports

Reading receipt

  • SemiAnalysis estimates over 1.5 gigawatts of restricted Nvidia hardware reached China.
  • The firm suggests this smuggling, alongside personal scandals, may have driven Jensen Huang’s departure.
  • The estimate excludes compliant chips like the H20 and H800, focusing on banned models.

Still open: How exactly did SemiAnalysis convert chip counts to gigawatts without disclosing the methodology?

Still open: What specific evidence supports the claim that the U.S. Department of Commerce agrees with these numbers?

Time spent: 0:00 · Depth: Full

Sources

  1. Tom's Hardware

Related stories

Background reading