Search
DeepSeek to Release V4 Large Model as Early as Next Week: Prioritizing Testing on Domestic Chips like Huawei's, Excluding NVIDIA and AMD
1 day ago /
Read about 0 minute
Author：小编
DeepSeek plans to release the new V4 version of its multimodal large language model next week. This version natively supports image, video, and text generation, marking the first major update since the launch of the R1 reasoning model in January this year. It is expected to meet domestic demand for low-cost open-source models. During the testing phase, DeepSeek granted testing access to domestic hardware suppliers like Huawei but did not provide pre-release version access to NVIDIA and AMD. This move breaks industry norms, allowing Chinese chip manufacturers to optimize software for their processors in advance and gain a market edge, with the underlying strategy warranting deep consideration.
-
C114 Communication Network -
Communication Home
7 X 24 Track global technological trends
