China Unveils World’s First Brain-Like Mini Supercomputer, the BIE-1

In a breakthrough that could reshape the global computing landscape, Chinese scientists have unveiled the world’s first brain-like intelligent computer — the BIE-1, a refrigerator-sized device that rivals room-sized supercomputers while consuming just a fraction of their power.

The BI Explorer computing system (BIE-1) was developed by the Guangdong Institute of Intelligence Science and Technology (GDIIST) and introduced during a high-tech forum in the Guangdong-Macau In-depth Cooperation Zone on Friday.A Supercomputer in a Mini Fridge

Unlike conventional supercomputers that occupy entire rooms and consume megawatts of power, the BIE-1 is compact enough to fit in a home or office. The institute says the system can plug directly into a standard household socket, consuming only one-tenth the power of traditional supercomputers, translating to a consumption level of roughly 2,000 to 3,000 watts.

“A traditional computing centre is like a building — requiring massive investment and enormous energy,” said Nie Lei, co-director of the GDIIST Intelligent Computing Systems Laboratory. “The BIE-1 is the size of a mini refrigerator and can be deployed anywhere — even in mobile environments — while maintaining supercomputer-level performance.”

This compact device was jointly developed with Zhuhai Hengqin Neogenint Technology and Suiren (Zhuhai) Medical Technology, both start-ups incubated by GDIIST.

Brain-Inspired Architecture

At the heart of the BIE-1 is a neural network architecture inspired by the human brain, capable of intuitive reasoning and adaptive learning. The system’s intuitive neural network (INN) mimics how neurons connect, allowing it to learn from limited data, interpret information autonomously, and process multiple sensory inputs — text, speech, and images — simultaneously.

This brain-like design allows the BIE-1 to achieve training speeds of 100,000 tokens per second and inference speeds of 500,000 tokens per second, performance metrics typically associated with GPU clusters in advanced data centers.

High Power in a Compact Package

Despite its size, the BIE-1 is packed with hardware muscle. It integrates 1,152 CPU cores, 4.8 terabytes of DDR5 memory, and 204 terabytes of local storage — figures comparable to high-performance computing (HPC) nodes used in AI research facilities.

Even under heavy workloads, the system’s CPU temperature remains below 70°C, and its low-noise cooling makes it suitable for quiet environments like classrooms or hospitals.

90% Less Energy, 100% More Potential

Supercomputers typically consume enormous amounts of power not only for computation but also for cooling. By contrast, the BIE-1 achieves up to 90% lower energy consumption while maintaining equivalent performance. This could make high-end computing far more sustainable and widely accessible.

According to GDIIST, the device is ideal for AI training, edge computing, and personal intelligent systems, opening the door for widespread use of brain-inspired computing beyond research labs and government facilities.

Applications Across Sectors

The BIE-1’s flexibility allows it to serve diverse industries:

• Healthcare: Real-time home health monitoring and diagnostics.

• Education: Personalized tutoring systems adapting to each student’s pace.

• Enterprise: Intelligent assistants capable of automating office workflows.

• Research: Compact data analysis platforms for field scientists and engineers.

GDIIST researchers say the system could even be used to power autonomous drones, mobile labs, or smart city infrastructures, all while cutting down drastically on power needs.

China’s Push Toward AI Hardware Independence

The BIE-1 represents more than just a technical milestone — it also underscores China’s growing effort to achieve self-reliance in high-performance and AI computing amid global chip tensions.

In recent years, China has invested heavily in neuromorphic and analog AI chips, including systems reportedly 1,000 times faster than Nvidia GPUs for certain tasks. The BIE-1 builds on this foundation, combining software innovation with indigenous hardware to bypass conventional limitations.

Toward the Future of “Everywhere AI”

By miniaturizing supercomputing into a household-friendly form, GDIIST envisions a future where intelligent computing becomes as common as home Wi-Fi routers.

“This is not just a smaller supercomputer — it’s a step toward embedding intelligence everywhere,” said Nie Lei. “From hospitals to classrooms to living rooms, high-end AI will soon be within everyone’s reach.”

If proven commercially viable, the BIE-1 could transform how computing power is distributed, making advanced AI available to small businesses, researchers, and even individuals — a move that could fundamentally democratize the next generation of artificial intelligence.