Technology news and discussion on TalkFuse.
China’s self-driving firms prepare for tighter Nvidia chip controls
Chinese self-driving companies are preparing for tighter US export controls on Nvidia chips, a key source of computing power for training and running autonomous driving systems. Pony.ai CFO Leo Wang said the industry has been developing alternatives for years, giving firms more flexibility if access to Nvidia hardware becomes more restricted.
Pony.ai is responding to a shortage of computing capacity by making its training models more efficient, sometimes using mid-tier resources, and turning to chips made in China. Wang said domestic options are more mature than they were a few years ago, though they are not yet as cost-effective. The article also notes Nvidia’s varied role in the sector, from supplying processors and training GPUs to partnering with some vehicle makers and self-driving firms.
In short
- Chinese self-driving firms have been preparing for tighter Nvidia chip restrictions.
- Pony.ai is improving efficiency and using domestically made chips.
- Chinese alternatives are more mature, but still less cost-effective.
Explained simply
Self-driving cars rely on large amounts of computing power to train their software and process information while driving. Nvidia, a US chipmaker, supplies processors and graphics chips used for these tasks.
The US can restrict exports of advanced technology to China, and tighter limits could make some Nvidia chips harder for Chinese companies to obtain. Pony.ai is one of China’s self-driving firms; it says companies have been developing other options, including chips made in China, although those alternatives are not yet as cost-effective.
Read the full article at the source →
Originally published by TechCrunch as “TechCrunch Mobility: The Nvidia backup plan for China’s self-driving industry”.