# DeepSeek reportedly plans to pre-train V5 on Huawei Ascend chips

> Original title: DeepSeek plans to pre-train its next-gen V5 model on Huawei Ascend chips, per STAR Market Daily sources.

- Source: X.PIN
- Published: 2026-11-04T15:15:00.000Z
- Bad Influence: https://badinfluence.news/items/enim06sl5xehccth7a3b8fli8
- Original: https://x.com/thexpin/status/2117999057398354315

## Summary

DeepSeek reportedly plans to pre-train its next-generation V5 model on Huawei Ascend chips, according to STAR Market Daily sources. Until now, Ascend has handled only DeepSeek inference and V4-Pro post-training. The post links this to DeepSeek open-sourcing its Ascend compute and communications stack, including DeepGEMM, FlashMLA and DeepEP, on September 30.

## Post

DeepSeek plans to pre-train its next-gen V5 model on Huawei Ascend chips, per STAR Market Daily sources.

Until now, Ascend only handled DeepSeek inference and V4-Pro post-training. Suddenly, DeepSeek open-sourcing its full Ascend compute & comms stack (DeepGEMM, FlashMLA, DeepEP) on Sept 30 makes total sense.

![](https://badinfluence.news/api/img-proxy?u=https%3A%2F%2Fpbs.twimg.com%2Fmedia%2FHWSXC87awAA_Trv.jpg&mode=thumb&exp=1794009600&sig=b46c84b39359f947)