Chinese startup DeepSeek has begun developing its own artificial intelligence chips. The company is creating processors for inference—the stage at which already trained models generate responses to user queries.
DeepSeek is already in talks with potential manufacturing partners and is hiring engineers to work on the project. The main goal is to reduce dependence on third-party suppliers, including NVIDIA and Huawei.
Many of the largest technology companies are currently developing their own AI chips, as specialized hardware allows for reduced costs and improved model performance. If DeepSeek manages to replicate its approach to creating affordable and efficient AI solutions at the hardware level, it could intensify competition in the market and exert additional pressure on NVIDIA's market position.
DeepSeek gained widespread recognition after releasing an open-source language model that, according to the company, achieved competitive performance at significantly lower computational costs compared to foreign counterparts.
Source: usnews.com
The Tech on Telegram Follow The Tech on Telegram
Keep up with the latest technology and startup news.