IT Home On October 12, according to the Securities Times, AI infrastructure company Wwenxinqiong has completed a C-round financing of 2.3 billion yuan. The total financing amount since its establishment reaches 4.3 billion yuan, with both the single-round and cumulative amounts setting new records in China’s AI native infrastructure sector.
Public information shows that Wuwenxinqiong was established in 2023 and is a technology transfer enterprise of Tsinghua University and Shanghai Jiao Tong University.
IT Home noticed that, according to a source cited by Bloomberg this month, Wuwenxinqiong has submitted an IPO application to the Hong Kong Exchanges and Clearing Limited in secret form over the past few months. It plans to raise hundreds of millions of US dollars, with the goal of listing in the first half of 2027.
Further reading
Shanghai Wuwenxinqiong Intelligent Technology Co., Ltd. is the full name of the company. It was established in May 2023 and is incubated by the NICS-EFC Laboratory team from the Department of Electronic Engineering at Tsinghua University. It is also a technology transfer enterprise for the achievements of science and technology between Tsinghua University and Shanghai Jiao Tong University.
In terms of the team, co-founder and CEO Xia Lixue holds a Ph.D. from Tsinghua University. He studied under Professor Wang Yu, the head of the Electronic Engineering Department at Tsinghua University. He was included in the AI2000 list of the world’s most influential AI scholars (Top100 in the chip category). Founder Wang Yu has long been engaged in research on intelligent chips and high-efficiency circuit systems. In 2016, he invested in Shenjian Technology through the conversion of knowledge achievements. This company was acquired by Silexis in 2018 and later integrated into the AMD system. The other co-founder, Dai Guohao, serves as the chief scientist and is an associate professor with a long-term teaching appointment at Shanghai Jiao Tong University.
Unlike chip design companies such as BiRen Technology and Suiyuan Technology, as well as large model companies like Zhipu and MiniMax, Wwenxinqiong takes a “middle layer” approach: it does not develop GPU/AI accelerator hardware itself, nor does it directly build underlying large models. Instead, it focuses on engineering-level challenges such as heterogeneous computing power scheduling, software and hardware coordination optimization, and large-scale cluster utilization rates, and proposes an M×N technical paradigm that connects “M types of models” with “N types of chips”. Specifically, the “One Center” computing power distribution center has been deployed with 37,000P of computing power, covering 16 mainstream chips. Its thousand-card-scale heterogeneous chip mixed training platform has scalability for tens of thousands of cards. It has achieved mixed training of large models using four domestic chips from Huawei Ascend, Tianlian Smart Chip, MuXi, and Moore Thread, respectively with two imported chips from AMD and NVIDIA, with a maximum computing power utilization rate of 97.6%.
In the cross-domain scenario, Wunwenqiong and China Telecom completed the first large-scale cross-domain mixed training technology verification in China with a distance of over 4,000 kilometers. This involved carrying out cross-domain training tasks from Hami in Xinjiang to Shenzhen in Guangdong, achieving efficient and stable training of a 30 billion parameter model.
“Backyard”Token factory corresponds to the Agentic MaaS large model service platform. The company’s pioneering cross-cluster heterogeneous PD separation architecture (PDD) reduces the first Token delay (TTFT) by 51.5%, while the cost per single Token can be reduced by 37.5%. Over the past year, its Token factory has led to a 10-fold reduction in inference costs.
In terms of customers and the ecosystem, Wenwenxinqiong has reached cooperation with leading large model companies such as Zhipu and Yuezhi Andi. Its industry solutions cover areas like entertainment games, healthcare, legal services, and energy power. Customers include China Mobile, China Unicom, Southern University of Science and Technology, and Migu Culture Technology, as well as government and research institutions.
In its previous financing process, Wuwenxinqiong completed a Series A financing of nearly 500 million yuan in September 2024. Over more than one year after its establishment, the cumulative financing reached nearly 1 billion yuan. In November 2025, it completed a Series A+ financing of nearly 500 million yuan. In May 2026, it announced receiving over 700 million yuan in financing, with Hangzhou High-tech Capital Group and Huiyuan Capital as co-investors.
The investors include cross-border Chinese institutions, industrial investors, and financial investment firms. Chinese-owned institutions include Hangzhou High-tech Capital, Social Security Zhongguancun Fund, Zhuhai Technology Industry Group, Xiamen Guoxing Investment, and Xuhui Capital; industrial investors include Tencent, Baidu, Lenovo, and Xiaomi; financial investment firms include Sequoia China, Qiming Venture Capital, Hongtai Fund, Dachen Finance Intelligence, and Shunwei Capital.