As overseas high-end AI chip export controls continue to be tightened, DeepSeek, a leading domestic model company, is accelerating its migration to a domestic computing power base. Recently, we learned from people familiar with the industry chain that DeepSeek V5, the next big model of DeepSeek, plans to use the Huawei Ascend 950 chip as the core training computing power, and the project pace points to next year. At this stage, DeepSeek's core R&D and docking team with Shengteng is based in Hangzhou, and the two sides have already completed a large number of low-level operator adaptations and customized development during the V4 phase. According to the cooperation path disclosed by people familiar with the matter, previously in the previous DeepSeek version, Shengteng Computing only undertook inference tasks and had not yet entered the training session; V5 will be its first major model to natively rely on the Shengteng 950 for training.

Disclaimer:Webull uses external vendor Google Translation Service for news translations where we endeavour to ensure these are correct, however, we recommend that you please double-check this information accordingly. Webull is not responsible for translation errors or issues.