TechCrunch Mobility: China’s Self-Driving Industry Has a Backup Plan for Nvidia

China prepares chip alternatives
At Web Summit, co-founder Paddy Cosgrave highlighted Chinese founders and companies, saying future superstars would come from China rather than California. Nvidia’s importance to autonomous vehicle technology also featured in conversations at the conference.
Leo Wang, CFO of Chinese autonomous vehicle company Pony.ai, said computing demand already exceeds supply. The company asks engineers to use training resources efficiently, including midtier options rather than high-end computing when possible.
Pony.ai is also using more chips made in China. Wang said domestic chips offer another option for AI training and give the company flexibility.
A backup plan already underway
Wang said Chinese companies have been preparing for years for the possibility of stricter export controls on Nvidia chips. He said alternatives exist and have become more mature, although they may not yet be as cost-effective.
Nvidia supplies AI computing technology to the autonomous vehicle industry in different ways. Some companies use its automotive processors to run driving software; others use its GPUs to train models. Mercedes-Benz, Zoox and Wayve have full-stack partnerships with Nvidia.
Other mobility developments
May Mobility plans to use Volkswagen’s all-electric ID. Buzz for a global ride-hailing platform and is targeting more than 10,000 autonomous vehicles by 2031. Volkswagen said it is offering an automated-driving-ready base vehicle to companies developing their own systems; this does not mean its MOIA group is leaving Mobileye.
Key points
- Pony.ai is managing a computing shortage with more efficient training and domestic chips.
- Leo Wang said Chinese firms have prepared alternatives to Nvidia chips for years.
- May Mobility targets more than 10,000 autonomous vehicles through 2031.
- Lucid reported $350 million in third-quarter revenue and a net loss of $830.9 million.

