Upcoming DeepSeek AI model failed to train using Huawei’s chips

A startup developing a new AI model called DeepSeek faced technical difficulties during the training process. The company encountered persistent issues while attempting to use Huawei's Ascend chips, which are designed for AI workloads. Despite their efforts, the startup was unable to successfully train the DeepSeek model using the Huawei hardware. The article does not provide any further details about the nature of the technical problems or the reasons behind the failed training attempts. It also does not mention the specific capabilities or intended applications of the DeepSeek AI model. The news highlights the challenges that AI developers can face when working with emerging hardware and technologies. The inability to train the model using Huawei's chips suggests that there may be compatibility or performance issues that need to be addressed. The article does not indicate whether the startup was able to successfully train the model using alternative hardware or software solutions.