Chinese AI firm DeepSeek is moving to design its own chips for running trained models. The effort aims to cut reliance on suppliers including Nvidia and Huawei amid US export controls.
The company has begun hiring engineers and is in discussions with manufacturing partners, according to reports from July 7. The focus is on inference chips rather than training hardware.
DeepSeek rose to prominence with low-cost open-source AI models that compete with leading offerings. Its chip initiative remains early stage and would likely stay within China due to existing trade restrictions.
Industry observers note the development could intensify competition in the domestic AI sector if cost and efficiency gains are achieved.