Chinese AI startup DeepSeek is rapidly being adopted among domestic tech firms, particularly in the cloud services and semiconductor sectors.
Chinese tech firms, including Moore Threads, Hygon Information Technology, Huawei Technologies, Alibaba, Baidu, and Tencent, have been integrating DeepSeek’s LLM models into their cloud services and other operations.
Namely, Moore Threads and Hygon Information Technology have announced support for DeepSeek’s models using their AI chips.
Moore Threads has acknowledged that DeepSeek’s open-source V3 and R1 models have significantly advanced AI development and served as a source of inspiration for developers. To further the growth of the domestic AI ecosystem, Moore Threads plans to make its proprietary KUAE GPU intelligent computing cluster available to fully support the distributed deployment of DeepSeek’s V3 and R1 models. The given cluster is a comprehensive solution for AI data centers built on Moore Threads’ chip technology.
Hygon Information Technology has announced that its AI accelerators now support DeepSeek’s R1 and V3 models, aiming to enhance China’s AI capabilities. Some media reports suggest that Cambricon Technologies, a leading Chinese AI chip developer, is also joining the DeepSeek adoption wave, although there is not yet official confirmation of this fact.
Huawei Technologies, in collaboration with AI infrastructure startup SiliconFlow, is also integrating DeepSeek’s models into its Ascend cloud service. The partners introduced DeepSeek’s large language model V3 and reasoning model R1 to end users, with SiliconFlow offering access to these models at competitive rates. Meanwhile, major Chinese cloud service providers like Alibaba, Baidu, and Tencent have also made DeepSeek’s models available within their offerings, further solidifying the AI startup’s position in the market.
Analysts view these developments as a significant milestone for China’s AI industry, reducing reliance on advanced U.S. hardware and showcasing the capabilities of domestic technology.
Low-cost open-source AI models from DeepSeek took the tech world by storm just a week ago. Upon its release, the stock market saw significant declines in major tech shares, including Nvidia, Microsoft, and Meta. The novel solution intensified discussions about the global AI race and its implications for technology investments, raising questions about the sustainability of U.S. leadership in AI. The President of the United States, Donald Trump, said that the technology of DeepSeek should be a wake-up call for US industries to improve the existing AI capabilities.
DeepSeek’s efficient model training and open-source approach have garnered not only U.S. concerns but also international attention, positioning it as a notable player in the AI landscape. However, its association with the Chinese government and potential data security issues have raised caution among some international entities and experts who believe the tool might potentially spread misinformation or bias and share user data with the Chinese state authorities.
