China Unveils Darwin Monkey: World's Largest Brain-Like Computer

In a landmark achievement for neuromorphic computing, engineers at Zhejiang University in China have developed Darwin Monkey, a brain-inspired supercomputer that simulates over 2 billion artificial neurons—surpassing Intel’s Hala Point system (1.15 billion neurons) announced earlier in 2024.

The Darwin Monkey integrates 960 Darwin 3 neuromorphic chips, collectively simulating more than 100 billion synapses, approaching the complexity of a macaque brain. The chips were co-developed with Zhejiang Lab, supported by the provincial government, Alibaba Group, and the university itself. Darwin Monkey’s neuromorphic architecture enables it to mimic biological brain activity using spiking neural networks, combining memory and processing functions for vastly improved energy efficiency.

Key Highlights

• Darwin Monkey simulates over 2 billion neurons via 960 neuromorphic chips
• Surpasses Intel’s Hala Point (1.15B neurons) in scale
• Performs logical reasoning, content generation, and brain simulations

Remarkably, it operates at 2,000 watts, a meaningful step toward the brain’s biological benchmark of ~20 watts. The system has demonstrated tasks such as logical reasoning, math, and content generation, powered by a large AI model from Chinese firm DeepSeek. It is also capable of simulating various animal brains, including mice and zebrafish, presenting a powerful tool for neuroscience and AI research.

Also Read: Foxconn and NVIDIA to Construct AI Supercomputer in Taiwan

This development signifies China’s leap ahead in the neuromorphic computing race, reflecting the nation’s heavy investment in post-Moore’s Law chip technologies. With the neuromorphic computing market projected to hit $6.48 billion by 2024, breakthroughs like Darwin Monkey’s scale and efficiency could shape the next era of AI processing, robotics, and intelligent systems.