Flexible, Low-Power, Intelligent: China’s Latest Chip Innovation for Wearables
Chinese scientists have developed a flexible, low-power chip architecture that could transform wearable devices, enabling on-device AI processing and improved comfort.

By Indrani Priyadarshini

on February 4, 2026

A research team in China has unveiled a new class of flexible microchips designed to bring smarter, more efficient computing capabilities directly to the body — a development that could reshape the future of wearable electronics.

The work, recently published in Nature, addresses a key limitation of today’s wearable devices: reliance on rigid silicon chips. These traditional components, while powerful, can be uncomfortable when worn and often lack the adaptability needed for seamless integration with skin, fabrics, or other flexible surfaces.

Read More | No Lights, No Workers: Xiaomi’s Robot Factory Redefines Manufacturing

To overcome these constraints, researchers at Tsinghua University developed the FLEXI series of chips, built on low-temperature processes compatible with flexible plastic substrates. What sets these chips apart is their “compute-in-memory” architecture, which allows many computing tasks to occur inside memory cells rather than shuttling data back and forth between separate processing and storage units. This design significantly reduces energy consumption and accelerates processing — critical for small, battery-powered wearables.

Each FLEXI chip is lightweight, cost-effective (manufactured for under $1), and remarkably resilient. In testing, the chips endured more than 40,000 bending cycles without loss of performance and maintained stable operation for over six months. Despite their flexibility, they deliver a clock frequency of up to 12.5 megahertz with power draw measured in single-digit milliwatts — a promising balance of performance and efficiency.

Read More | India Opens First Government-Run AI Clinic to Speed Up Disease Detection

In practical demonstrations, the team used a FLEXI chip to power a neural network that tracked and analysed physiological signals, including heart rate, breathing, skin temperature, and moisture. The system was able to distinguish daily activities with 97.4% accuracy, all without needing to send data to a smartphone or remote server.

Tsinghua professor Ren Tianling described the FLEXI platform as a convergence of high performance, low power demand, and long-term durability. He suggests the technology could underpin a new generation of wearable devices that are not only more comfortable but also genuinely autonomous in processing data at the edge.

News Image
News Image