Chinese researchers have achieved a significant milestone in artificial intelligence (AI) technology with the introduction of a fully flexible AI chip, as detailed in a paper published on January 29 in Nature. This breakthrough, developed by teams from Tsinghua University, Peking University, and other institutions, aims to enhance applications such as wearable health monitoring devices and flexible robotics.
The increasing convergence of AI with the Internet of Things (IoT) has escalated the demand for lightweight, efficient, and flexible computing hardware. Traditional silicon-based chips, while common, are rigid and ill-equipped to adapt to the human body or complex surfaces. Meanwhile, existing flexible processors face limitations including low operating frequencies and high power consumption, rendering them unsuitable for data-intensive tasks such as neural network inference.
The research team’s innovation, termed the FLEXI series of digital compute-in-memory AI chips, uses domestic manufacturing processes to overcome the historical barriers to high-performance edge AI computing found in flexible electronics. Utilizing low-temperature polycrystalline silicon thin-film transistors, the FLEXI chip is ultra-thin, bendable, and characterized by ultra-low power consumption and high energy efficiency.
According to the team, the chip’s compute-in-memory architecture integrates memory units with computing units within a single structure. This integration effectively eliminates the time and energy overhead associated with data movement, thereby enhancing computational efficiency. The smallest variant, known as the FLEXI-1, measures just 31.12 square millimeters and houses 10,628 transistors, operating at an exceptionally low power consumption of just 55.94 microwatts.
Yan Bonan, an assistant professor at the Peking University Institute for Artificial Intelligence and one of the study’s authors, noted that the chip can endure over 40,000 cycles of 180-degree bending without any degradation in performance. Furthermore, it has performed stably in long-term tests lasting up to six months. This chip supports neural network compression and features one-click deployment, thereby augmenting its on-chip intelligence. Despite its modest capacity of 1 kilobit, a single flexible chip has achieved an impressive arrhythmia detection accuracy of 99.2%, positioning it as a key component for next-generation wearable medical devices, flexible brain–computer interfaces, and intelligent robotic systems.
The development of the FLEXI chip could pave the way for advanced applications in healthcare technology and robotics, facilitating the integration of AI into everyday life through more adaptable and efficient systems. As the demand for flexible and efficient computing solutions continues to grow, this innovation may represent a crucial advancement in the evolution of AI and its applications in various sectors.
See also
Tesseract Launches Site Manager and PRISM Vision Badge for Job Site Clarity
Affordable Android Smartwatches That Offer Great Value and Features
Russia”s AIDOL Robot Stumbles During Debut in Moscow
AI Technology Revolutionizes Meat Processing at Cargill Slaughterhouse
Seagate Unveils Exos 4U100: 3.2PB AI-Ready Storage with Advanced HAMR Tech
