Nvidia Expands Support for Chinese AI Models as SEC Warns of U.S. Restriction Risks
Updated
Updated · CNBC · Aug 27
Nvidia Expands Support for Chinese AI Models as SEC Warns of U.S. Restriction Risks
3 articles · Updated · CNBC · Aug 27
Summary
Nvidia on Wednesday highlighted new optimizations for Chinese open models including DeepSeek V4 Flash and Alibaba’s Qwen 3.8, extending its “local AI initiative” as those systems gain global traction.
The push is aimed at keeping Nvidia’s hardware the default AI stack before Chinese chipmakers such as Huawei and Alibaba lock in developers with their own model optimizations.
An SEC filing tied to second-quarter earnings warned that White House limits on supporting apps built on Chinese-origin open models like DeepSeek, Qwen or Kimi could materially hurt Nvidia’s business.
Washington’s scrutiny has intensified as Chinese open models improve and spread in the U.S. and abroad, while Nvidia and more than 20 companies including Microsoft and Meta have urged policymakers to avoid early curbs.
The fight over open models is becoming a broader U.S.-China tech battleground, with analysts warning cheaper Chinese systems could become defaults in developing markets and shape future political alignment.