Hygon, a Shanghai-listed semiconductor firm, launches embedded processors targeting robotics and edge AI applications. The chips aim to decentralize AI processing, enabling deployment in factory floors and industrial machines instead of relying solely on cloud servers. This expansion follows Hygon’s prior focus on data center solutions.
Hygon develops AI chips for physical applications where machines interact directly with their environments. The company targets expansion beyond data centers into robotics, broadening its chip use cases. Hygon’s new chip designs aim to support real-world AI deployment in industries requiring direct machine-environment interaction.
UCloud deploys servers with Hygon’s Tianxi AI accelerators on its public-cloud platform, marking the first large-scale public-cloud rollout for the chip series. The move expands AI acceleration capabilities for cloud customers using UCloud’s infrastructure. Hygon’s Tianxi accelerators are designed to enhance performance for AI workloads.
Moore Threads and Hygon are seeing strong sales growth as they supply AI chips to Chinese customers amid rising demand. The companies are expanding production to replace Nvidia’s market share after U.S. export restrictions limited access to high-end AI accelerators. Domestic firms are accelerating development of alternatives to meet the growing needs of AI training and data centers.