DeepSeek-V3 Technical Report
DeepSeek-AI, 2024, DeepSeek-V3 Technical Report (arXiv preprint arXiv:2412.19437)
Subscriber deck — free during early access
1 cards in “6 Conclusion, Limitations, and Future Directions” · whole deck
6 Conclusion, Limitations, and Future Directions
- What limitations of DeepSeek-V3 do its authors name?Deployment: the recommended serving unit is large (prefilling alone uses 32 GPUs), a burden for small teams, and generation speed, though over twice DeepSeek-V2’s, could still improve. They expect better hardware to ease both.
- DeepSeek-AI, 2024, DeepSeek-V3 Technical Report (arXiv preprint arXiv:2412.19437), §6
Source
- DeepSeek-AI. 2024. DeepSeek-V3 Technical Report. In arXiv preprint arXiv:2412.19437. https://arxiv.org/abs/2412.19437