China’s Darwin Monkey brain-like computer challenges AI boundaries with 2 billion artificial neurons

China’s ‘Darwin Monkey’: A New Milestone in Brain-Inspired AI Computing

Chinese engineers at Zhejiang University have introduced the world’s first brain-like computer featuring over 2 billion artificial neurons. Nicknamed the “Darwin Monkey,” its neuron count approaches that of a macaque, marking a significant step in advancing brain-inspired artificial intelligence (AI).

The Darwin Monkey builds on previous brain-inspired computing efforts by the Zhejiang University team. Comprised of 960 Darwin 3 neuromorphic chips, the system creates more than 100 billion synapses. This architecture brings us closer to replicating complex brain functions and achieving more advanced artificial intelligence.

Technical Achievements and Applications

This computing system has successfully handled tasks such as content generation, logical reasoning, and mathematical problem-solving. It utilizes DeepSeek’s brain-like large model, a Chinese AI company specializing in neuromorphic AI applications.

Beyond AI applications, the Darwin Monkey’s neuron and synapse resources can simulate the brains of various animals including macaques, mice, and zebrafish. Such simulations have potential to accelerate neuroscience research by providing detailed models of biological neural networks.

Neuromorphic Computing and Spiking Neural Networks

Neuromorphic computing mimics the brain’s neural network architecture using artificial neurons and synapses. This approach allows for efficient information processing that models cognitive functions like learning, memory, and decision-making. Compared to traditional AI, neuromorphic systems can offer faster, more adaptable problem-solving.

One key innovation is the use of spiking neural networks, which transmit data via discrete spikes rather than continuous signals. This method more closely reflects how biological neurons communicate, enabling more accurate brain simulations and potentially more energy-efficient AI.

Comparisons and Energy Efficiency

Intel’s Hala Point, announced in April 2023 with 1.15 billion neurons, was previously the largest neuromorphic system. It was deployed at Sandia National Laboratories to explore scalable AI solutions. The Darwin Monkey surpasses this with over 2 billion neurons while consuming around 2,000 watts of power, demonstrating significant progress in energy efficiency.

The Darwin 3 Chip: Core of the System

The Darwin Monkey relies on the Darwin 3 chip, developed in 2023 through collaboration between Zhejiang University and Zhejiang Lab. Each chip supports over 2.35 million spiking neurons and hundreds of millions of synapses. It also features specialized brain-inspired instruction sets and neuromorphic online learning mechanisms, enhancing its adaptability and performance.

Future Implications

Pan Gang, director of the State Key Laboratory of Brain-Computer Intelligence at Zhejiang University, highlighted the system’s large scale, high parallelism, and low power consumption as foundational for new computing paradigms.

This development follows the 2020 release of the “Darwin Mouse,” which featured 120 million artificial neurons. The Darwin Monkey represents progress in neural system integration, chip interconnection, and brain-inspired operating systems.

• World’s first brain-like computer with over 2 billion neurons
• Uses 960 Darwin 3 chips creating 100+ billion synapses
• Supports animal brain simulations for neuroscience
• Consumes approximately 2,000 watts under normal use
• Built with spiking neural networks for realistic brain modeling

For professionals interested in neuromorphic AI and brain-inspired computing, staying updated on such innovations is essential. Resources like Complete AI Training’s latest AI courses offer valuable insights into emerging AI technologies.