Beijing, February 13 – DeepSeek’s AI models have given Chinese chipmakers like Huawei a competitive advantage over U.S. giants such as Nvidia. While Chinese companies have faced challenges in producing high-end chips for AI training, DeepSeek’s focus on inference optimizes computational efficiency, helping to bridge the performance gap between Chinese and U.S. chips.
Chinese AI chipmakers, including Huawei, Hygon, and EnFlame, have embraced DeepSeek’s open-source models and low fees to boost AI development amid U.S. export restrictions. Despite this progress, Nvidia remains dominant due to its CUDA platform’s extensive software capabilities, which remain difficult for Chinese firms to replicate.
