Author | ZeR0 Editor | Moying
Xindongxi reported on October 29 that, according to the CSRC website, on October 28, Beijing-based embodied AI high-computing chip company Huixi Technology filed for listing counseling registration with the Beijing Securities Regulatory Bureau, initiating its A-share IPO process. The counseling institution is CITIC Securities.
Huixi Technology was established in April 2022 with a registered capital of CNY 42.53 million. Its legal representative is founder and CEO Xu Ningyi. The controlling shareholder is Hefei Yinxi Intelligent Technology Partnership Enterprise (Limited Partnership), holding 18.4280% equity with a special voting rights ratio of 68.7694%.
Xu Ningyi obtained his bachelor's, master's, and doctoral degrees from the Department of Electronic Engineering at Tsinghua University. He previously served as head of the Hardware Computing Group at Microsoft Research Asia, Chief Architect of Baidu's Intelligent Chip Division, and CEO of Zhenliang Intelligent, leading the design and large-scale commercialization of multiple AI chips. Among the other two co-founders, Zhang Jianyong graduated from the Department of Automotive Engineering at Tsinghua University with bachelor's and master's degrees, and from Shanghai Jiao Tong University with a doctoral degree; he formerly served as AVP of Autonomous Driving at NIO. He Guanghui is a professor at Shanghai Jiao Tong University, long engaged in research on high-efficiency digital system chips and high-performance chip interconnect technology, with academic achievements industrialized across multiple fields.
01.Lei Jun's Shunwei Capital Participated in Multiple Funding Rounds, Cumulative Equity Financing Exceeds CNY 2.5 Billion This Year
In its early days, Huixi Technology focused on intelligent computing platforms for vehicles. In February 2023, the company announced it had completed two funding rounds: Angel and Angel+. The Angel+ round was jointly led by Shunwei Capital under Lei Jun and Xiaomi Group, with participation from Guoqi Investment, Lianxing Capital, Kevins Fund, SenseTime Guoxiang Capital, MiraclePlus, Jinsha River Venture Capital, Lishi Capital, Qingyan Capital, Zhuoyuan Capital, and others; existing shareholders from the Angel round continued to increase their stakes. The Angel round was led by YuanSheng Capital and NIO Capital, with follow-on investment from prominent institutions such as ZhenFund, SEE Fund, and Cloud Nine Capital.
Just five months later, the company announced the completion of a strategic financing round of several hundred million yuan, led by Jingwei Hirain, with participation from Langmafeng Venture Capital, Taiming Capital, and Sanqi Interactive Entertainment; existing shareholder Lianxing Capital continued to increase its stake. Starting in 2025, Huixi Technology updated its positioning to build a computing platform for the era of embodied intelligence, and successively announced billion-level financings in 2026. One of these was the Pre-B round equity financing of over CNY 800 million announced in February this year, participated in by industrial guidance funds such as Liangxi Digital Mother Fund and Jingguorui Fund, along with multiple domestic embodied intelligence robotics companies, industrial CVCs, and seasoned industry investors; many existing shareholders continued to increase their stakes.
According to corporate credit inquiry platform Qichacha, investors in this round included AgiBot, GALBOT, and Dongfang Precision Science. Subsequently, in October this year, Huixi Technology announced that with the completion of Pre-B+ and B rounds, the company's cumulative equity financing scale within 2026 had exceeded CNY 2.5 billion.
This round was led by the Beijing Artificial Intelligence Industry Investment Fund, with joint participation from government funds and market-oriented investment institutions including the Beijing Robotics Industry Development Investment Fund, Yinshan Capital, Zhengxingu Capital, Jinfu Investment, Jiantou Investment, Huatai Baoli, Huimuyuanlun, Shanghe Momentum Fund, and the National Artificial Intelligence Industry Investment Fund; existing shareholders such as Shunwei Capital, YuanSheng Capital, Kevins Fund, SenseTime Guoxiang Capital, Bohua Industrial Investment, Guoqi Investment, Langmafeng Venture Capital, Sanqi Interactive Entertainment, Shengyi Capital, Xinglian Capital, and Zhuoyuan Asia Capital continued to increase their stakes.
02.Embodied AI Chips Mass Produced and Deployed, Collaborating with Leading Embodied AI Enterprises
Currently, Huixi Technology is dedicated to R&D of high-performance chips for embodied intelligence and building an industrial ecosystem, providing full-stack computing products and solutions covering chips, modules, software toolchains, and foundation models for robotics enterprises. Centered on the technical concept of "data closed-loop defines the chip," the company launched the high-performance AI SoC chip Guangzhi R1 based on a general-purpose programmable architecture, along with its accompanying computing platform, and released the Huixi Embodied computing product series during the 2026 World Robot Conference.
The Huixi Embodied computing product series includes Huixi Embodied R1 MAX and Huixi Embodied R1 PRO. Both products adopt a unified technical architecture and packaging design, targeting different performance and application needs: R1 MAX features 500TOPS INT8 AI computing power and a 24-core Cortex-A78AE CPU; R1 PRO features 375TOPS INT8 AI computing power and a 16-core Cortex-A78AE CPU.
Huixi Technology builds its technical system around the RPU computing kernel, RISE computing platform, and RhinoVLA model.
- RPU provides high-efficiency, ultra-low latency underlying computing support for multimodal perception, task understanding, reasoning decision-making, and action generation; RISE is built upon self-developed Huixi Embodied series chips, offering 500TOPS computing power, efficiently supporting frontier models such as VLA and world models, and has already been integrated into mass-produced products of multiple robotics enterprises; RhinoVLA provides standardized pre-trained foundation models and a complete development pipeline, helping robotics enterprises reduce investment in model R&D, adaptation, and deployment.
The company has collaborated with leading embodied intelligence enterprises such as AgiBot, Daxiao Robotics, X Square Robot, and Leju Robotics, continuously delivering engineering capabilities for Guangzhi R1 chips and RISE computing platforms. Product applications cover scenarios including industrial manufacturing, smart retail, commercial catering, and unmanned logistics, with shipment volumes of Huixi Embodied series computing products growing steadily.
In terms of ecosystem construction, Huixi Technology has partnered with upstream and downstream collaborators to launch the "Huixi Embodied Intelligent Computing Ecosystem Alliance," connecting industrial resources such as robot bodies, algorithm models, sensors, and basic software/hardware, promoting deep integration of technological innovation and application scenarios.
03.Conclusion: Embodied Intelligence Requires System-Level Computing Infrastructure
The embodied intelligence industry is currently developing rapidly, placing higher systematic requirements on underlying computing platforms. It not only demands higher single-chip AI computing power but must also simultaneously handle model inference, multi-sensor processing, real-time control, and data collection, while ensuring faster migration, deployment, and upgrading of algorithms.
By building complete computing platform capabilities from chips, core modules, domain controllers to software toolchains, Huixi Technology is driving embodied intelligence computing hardware to further become the underlying platform connecting models, data, and actions. The company plans to use humanoid robots as a "small entry point," with independently controllable computing platforms at the core, collaborating with upstream and downstream supply chain partners to build an open and collaborative embodied intelligence ecosystem, thereby driving the development of the embodied intelligence industry.