← Back to KHAO

DeepSeek · Nvidia · Huawei · China ·

Huawei-led team argues it post-taught DeepSeek's 1.6-trillion-parameter model, 1,000 Ascend 910C chips tapped in teaching

2 min read

Compiled by KHAO Editorial — aggregated from 1 source + 6 references discovered via search. See llms.txt for citation guidance.

◎ Multiple-sources

Image accompanies the article at Tom's Hardware. No description was extracted from the source.

A research group that includes Huawei Technologies says it completed full-parameter post-training of DeepSeek's V4-Pro, a 1.6-trillion-parameter model.

Key facts

Summary

The revelation is evidence that Chinese accelerators can now handle a training-class workload on domestic silicon, the part of the AI pipeline Chinese firms have had the most trouble moving off Nvidia hardware under U.S. export controls. The Ascend 910C is Huawei's current flagship AI accelerator, a dual-die part that returned roughly 60% of an Nvidia H100's inference performance in earlier DeepSeek testing. Post-training is the “tuning” stage that follows the much larger pre-training phase. Post-training then shapes behavior through instruction-following, safety alignment, and task-specific data.

Read full article at Tom's Hardware →

#DeepSeek #Nvidia #Huawei #China