Nvidia backs DeepSeek and Qwen optimizations despite Washington concerns

Nvidia is expanding support for leading Chinese artificial intelligence models as it seeks to keep its hardware at the center of the global AI ecosystem despite growing political scrutiny in Washington. The chipmaker highlighted optimizations for open models including DeepSeek V4 Flash and Alibaba’s Qwen 3.8 as part of a broader initiative designed to make popular AI systems run efficiently on Nvidia hardware. The strategy comes as Chinese models gain adoption among developers globally, including in the US, because of their improving capabilities and relatively low costs. Nvidia warned in a regulatory filing that potential US restrictions on supporting applications built on Chinese open-source foundation models could materially affect its business. The company specifically cited DeepSeek, Qwen and Kimi, saying restrictions on providing products and services that support applications built on those models could limit its ability to compete. Nvidia argues that supporting models regardless of where they are developed gives users an incentive to remain within the American technology ecosystem. An Nvidia employee told CNBC that developers using both American and Chinese models are likel...