Where is DeepSeek's next AI model?
9 days ago
- #DeepSeek
- #Huawei Chips
- #AI Development
- DeepSeek has delayed the launch of its R2 model due to challenges in training with Huawei Ascend AI chips.
- The company has not introduced major updates to its AI models in recent months, aside from minor revisions.
- Market expectations for a new foundation model shortly after R1's release have not been met, with no announced schedule for R2.
- Training advanced AI models is costly and complex, requiring significant computing resources, data, and algorithms.
- OpenAI took two and a half years to release GPT-5 after GPT-4, highlighting the time investment needed for such advancements.
- DeepSeek faced persistent technical issues while training the R2 model with Huawei's chips, as reported by The Financial Times.