In China, the release of the new DeepSeek AI model was postponed due to the problem with technical equipment. This is written by The Financial Times.
At the beginning of 2025, China’s authorities turned to DeepSeek developers calling for use systems based on Huawei processors instead of American technologies from Nvidia. But the developers had great difficulties when switching to Chinese analogues, and this delayed the release of a new version of the model.
The source notes that this situation shows that Chinese chips Huawei are not as good as American, and that Chinese IT companies are seriously dependent on technology from the USA. To help Deepseek developers to set up Chinese equipment and train AI, Huawei sent a team of engineers to their office. But even with their help, it was not possible to launch a new model.
In addition, FT informed sources reported that the output of the new version of Deepseek was also delayed due to the fact that the PRC authorities checked and prepared destructive documentation for the use of AI for a long time. The Chinese media suggest that the updated model will be presented in the near future, but the exact deadlines are not disclosed.
