An Ultra-Thin, Foldable AI Chip That Could Make Wearables Fully Independent

Wearable devices are becoming smarter every year. However, most of them still rely heavily on nearby smartphones to perform basic computing tasks. This limitation may not last much longer, as researchers from Tsinghua University and Peking University have developed a flexible artificial intelligence chip thinner than a human hair that can be folded thousands of times.

FLEXI Chip Features

According to a report by the technology news website Digital Trends, cited by Al Arabiya Business, the chip—known as FLEXI—could allow future wearable devices to run AI independently, without the need to constantly send data to a phone or the cloud.

What makes FLEXI unique is not only its computing capability but also its physical design. According to the research paper, the chip is built as a thin plastic film using low-temperature polycrystalline silicon circuits manufactured on a flexible substrate.

Because the entire system sits on this bendable surface, the chip can be folded, stretched, rolled, or even twisted without damaging its embedded AI circuits.

This makes it perfectly suited for wearable devices such as smart patches or health monitors that need to closely adapt to the shape of the human body.

To test its durability, FLEXI underwent extreme stress tests. It survived more than 40,000 bending and folding cycles to a radius of just one millimeter without any decline in performance.

Excellent Performance in Practical Tests

FLEXI also delivered impressive results in real-world testing. When used for health monitoring, it detected irregular heartbeats with 99.2% accuracy and tracked daily activities such as walking and cycling with 97.4% accuracy.

The chip proved highly energy-efficient as well, consuming less than 1% of the power used by conventional chips. According to projections, the flexible chip could cost less than one dollar per unit when mass-produced.

The researchers stated that their next step is to integrate more sensors into the chip and increase its complexity, paving the way for flexible, AI-powered wearables to become part of everyday life.