Recently, Fibocom completed a strategic investment in Yeats AI. To date, Yeats AI has received backing from leading industrial investors and venture capital institutions including Ant Group, CASSTAR, Shenzhen Capital Group, TCL Capital, Heli Capital, Unity Ventures, Tongge Ventures, Awinic Electronics, CDB Technology Ventures and Chaohui Capital.
Yeats AI specializes in on-device AI computing, with a focus on NPU architecture and hardware-software co-optimization. The investment will further strengthen Fibocom’s collaboration with upstream on-device computing partners and enhance its ecosystem for AI-powered devices.
As On-Device AI Moves into Products, Energy-Efficient Computing Becomes a Key Enabler
As multimodal AI, AI agents and other capabilities continue to extend to end devices, next-generation products such as AI smartphones, AI PCs, smart glasses and robots are accelerating their evolution toward greater intelligence. On-device AI is moving beyond model deployment toward real-world product applications.
According to Frost & Sullivan, China’s on-device AI market is expected to reach RMB 307.7 billion by 2029, with a five-year CAGR of 39.9%. Behind this rapid growth, balancing power consumption, performance and cost remains a key challenge for large-scale adoption.
This is particularly important for emerging devices such as smart glasses and wearables, where size, thermal management and battery life are critical considerations. Chip energy efficiency directly affects computing performance, battery life and overall user experience.
As an architecture optimized for AI computing, NPU technology can improve the processing efficiency and energy efficiency of on-device workloads. As more AI models are deployed on devices, energy-efficient computing and the ability to adapt to different models and product form factors are becoming essential foundations for the large-scale adoption of on-device AI.
Advancing On-Device AI Computing Through Ecosystem Collaboration
The founding team of Yeats AI previously worked on Apple’s Neural Engine team. The company focuses on highly energy-efficient inference for next-generation AI-native hardware and is developing a unified general-purpose computing platform to provide cross-device hardware and software solutions for diverse Agentic AI and Physical AI applications, enabling coordinated development across architecture, hardware and software.
A Yeats AI representative said, “Bringing on-device AI into real-world products requires closer integration between underlying computing capabilities and device requirements. Fibocom’s experience in smart modules, its customer base across intelligent devices, and its industry application expertise provide a strong foundation for accelerating the adoption of on-device computing technologies in real-world products.”
Chen Shijiang, Senior Vice President and Secretary of the Board at Fibocom, said, “This investment represents a key step in Fibocom’s full-stack on-device AI strategy. Yeats AI’s expertise in NPU architecture and on-device computing complements Fibocom’s product portfolio for intelligent devices, further strengthening the collaborative pathway from foundational computing capabilities to intelligent devices and real-world applications.”
As AI-powered devices continue to evolve, Fibocom will leverage wireless connectivity and artificial intelligence as its core technology foundations, continue collaborating with high-quality partners across the value chain, and accelerate the deployment of intelligent capabilities across devices and application scenarios—driving industries from “Everything Connected” toward “Everything Intelligent.”
