Key takeaways
- Jensen Huang's resignation as Nvidia CEO sparked immediate concerns about potential export controls on Nvidia chips to China, a region where autonomous vehicle companies have deep dependencies on the company's AI computing technology.
- Pony.ai and other Chinese AV firms have spent years preparing contingency plans, including optimizing model deployment efficiency and investing in domestically produced chip alternatives, though these remain less cost-effective than Nvidia's products.
- The autonomous vehicle sector continues expanding globally with new funding rounds and deployments from companies like May Mobility and Newmo, even as geopolitical uncertainties around Nvidia's future and export policy persist.
Jensen Huang’s sudden resignation as Nvidia’s CEO—announced shortly after admitting to an affair with an alleged spy—sent ripples through technology industries this week, but nowhere more acutely than among China’s autonomous vehicle companies. The timing coincided with Web Summit, where Chinese founders dominated the conversation. Summit co-founder Paddy Cosgrave emphasized that “the superstars of the future are coming from China, not California” and praised China’s leadership in open source technology. Against this backdrop, Huang’s departure raised an immediate question for an entire industry: what happens if export controls on Nvidia chips to China tighten?
The answer, according to executives in China’s self-driving sector, is that contingency plans are already in motion.
Why Nvidia Matters to Autonomous Driving
Nvidia occupies a rarely challenged position as the backbone of modern autonomous vehicle development. The company supplies the AI computing technology that powers the industry—though not in a uniform way. Some autonomous vehicle companies rely on Nvidia’s automotive processors to run their self-driving software directly. Others use Nvidia GPUs primarily for training the machine learning models that power autonomous systems. The deepest partnerships take a full-stack approach, bringing Nvidia technology into every layer. Mercedes-Benz, Zoox, and Wayve exemplify this last category.
Huang has been among the most vocal American technology leaders in defending open source principles, making his leadership a symbol of a particular vision for how AI development could proceed globally. The prospect that export controls might intensify under his successor has focused minds in Beijing on what comes next.

China’s Nvidia Dependency
The Supply Crunch
Demand for Nvidia chips in China far exceeds available supply. Leo Wang, CFO of Chinese autonomous vehicle company Pony.ai, described this shortage as a significant operational pain point for the company. Pony.ai has roughly two years of experience navigating this constraint, and the problem has only deepened as the company scales its model training and deployment.
A Two-Pronged Response
Pony.ai’s strategy to address Nvidia’s scarcity operates on two fronts. First, the company has directed its engineers to optimize efficiency in how they deploy training models. Rather than always reaching for the most powerful—and scarce—high-end compute resources, the company now uses midtier alternatives when possible. This approach stretches available resources further without sacrificing capability.
Second, Pony.ai is actively shifting toward domestically produced Chinese chips. Wang noted that the Chinese chip ecosystem has received substantial attention and investment in recent years, creating viable alternatives to Nvidia for AI training workloads. These options provide “some flexibility,” in his words, even if they are not yet cost-competitive with Nvidia’s established offerings.
A Plan B Already Exists
When asked directly whether stricter export controls would threaten Pony.ai’s operations, Wang indicated that Chinese companies have been preparing for this scenario for years. “If you look at the whole situation, this has been developing for already a couple of years,” he said. “Everyone, I would say, is also practicing kind of like a Plan B for a few years.”
The alternatives already in existence differ materially from Nvidia’s offerings in one crucial way: they are less cost-effective. Yet Wang characterized them as a mature backup option—demonstrably more advanced than they were a few years ago—rather than a speculative future capability. The implication is that Chinese companies can operate without Nvidia if required, though at higher cost and possibly with reduced efficiency.
The Broader Autonomy Sector
New Deployments and Funding
Activity in the autonomous vehicle space continues regardless of geopolitical uncertainty. May Mobility announced plans to deploy more than 10,000 autonomous vehicles through 2031, operating on a global ride-hailing platform powered by the Volkswagen all-electric ID.Buzz. VW clarified that this partnership does not mark a departure from its use of Mobileye technology; rather, the ID.Buzz is being offered as an autonomous-driving-ready platform that allows other companies to integrate their own self-driving systems.
Japanese robotaxi startup Newmo raised ¥20.8 billion, roughly $135 million, in a funding round led by KDDI. Port, an urban logistics company, secured €14.5 million from the European Bank for Reconstruction and Development and Balnord.
Regulatory Scrutiny and Market Expansion
The National Highway Traffic Safety Administration is investigating a recall Waymo issued over a year ago regarding its robotaxis illegally passing school buses. According to people familiar with Waymo’s operations, the company has not recorded any school-bus-related incidents since March. The majority of the incidents that triggered the recall actually occurred after Waymo deployed the software update intended to fix the problem.
Tesla is expanding its robotaxi service to San Antonio, Texas, as it positions itself to compete more directly with Waymo. Production of the polarizing Cybertruck will continue through next summer before ending, CEO Elon Musk said this week.
Adjacent Moves
Lucid reported third-quarter revenue of $350 million, below Wall Street expectations, though its net loss narrowed to $830.9 million. The EV maker increased its delayed draw term loan from Saudi Arabia’s Public Investment Fund to $1.75 billion. Uber launched a parking reservation service in major U.S. cities about nine months after acquiring parking app SpotHero. The ride-hailing company also notified thousands of drivers that personal and driver’s license information was stolen during an October data breach at one of its technology providers.
ParkBee, a Dutch parking technology company, acquired Mobian’s city and parking operations.
Source: TechCrunch
Frequently Asked Questions
Why does Jensen Huang's resignation affect China's autonomous vehicle industry?
Huang was a vocal proponent of open source technology in the United States, and Nvidia is central to autonomous vehicle development worldwide, supplying AI computing technology to companies like Mercedes-Benz, Zoox, and Wayve. His departure raises concerns that export controls on Nvidia chips to China might tighten, disrupting the industry's primary source for AI computing.
What is Pony.ai's strategy to reduce Nvidia dependence?
Pony.ai is using two approaches: optimizing how efficiently its engineers deploy training models by using midtier compute resources instead of high-end ones, and increasing use of domestically produced Chinese chips as alternatives. These domestic options are less cost-effective than Nvidia's products but provide flexibility if export restrictions tighten.
Do Chinese companies have viable alternatives to Nvidia?
Yes, according to Pony.ai's CFO Leo Wang, Chinese companies have been preparing backup plans for years. Domestic chip alternatives exist and have matured significantly compared to a few years ago, though they remain less cost-effective than Nvidia's established offerings and represent a viable but imperfect option if exports are restricted.