Monkey-Like AI: Chinese Researchers Develop Thinking Computer
Darwin Monkey: China’s Brain-Like Computer Mimics Macaque Intelligence, Ushering in New Era for AI and Neuroscience
august 2, 2025 – In a significant leap forward for artificial intelligence and neuroscience, Chinese engineers have unveiled a groundbreaking new generation of brain-like computer, codenamed “Darwin Monkey.” This revolutionary system, developed by the national key lab of brain intelligence at Zhejiang University, is designed too meticulously mimic the intricate workings of a macaque monkey’s brain, promising to unlock unprecedented capabilities in AI and offer profound insights into the complexities of biological cognition. The announcement, made on August 2, 2025, places China at the forefront of a burgeoning field that seeks to bridge the gap between artificial and biological intelligence.
the Architecture of Artificial Cognition: Darwin 3 Chips and Unprecedented Scale
At the heart of Darwin Monkey lies a elegant array of 960 Darwin 3 chips. These are the third-generation brain-like neural processing units, independently developed by the Zhejiang University lab. This proprietary technology represents a significant advancement in neuromorphic computing, a field dedicated to designing hardware that emulates the structure and function of the human brain.The sheer scale of the Darwin Monkey system is staggering.The chip array supports an astonishing more than 2 billion pulsed neurons and over 100 billion synapses. This massive capacity allows Darwin Monkey to process information in a manner far more akin to biological brains than conventional silicon-based computers. Synapses, the junctions between neurons, are crucial for learning and memory, and the vast number of simulated synapses in Darwin Monkey enables it to handle complex patterns and relationships within data.
This immense neural and synaptic density positions Darwin Monkey as the world’s first brain-like computer to seamlessly integrate advanced thinking abilities with critical sensory and cognitive functions. Unlike previous AI systems that might excel in one area, Darwin Monkey is engineered to simultaneously process and understand vision, hearing, language, and learning, mirroring the integrated nature of primate cognition. This holistic approach is a key differentiator, moving beyond specialized AI to a more generalized, adaptable form of artificial intelligence.
Powering the future: Energy Efficiency in Neuromorphic Computing
A critical aspect of Darwin Monkey’s design, and a testament to the advancements in neuromorphic engineering, is its power consumption. Under typical operating conditions, the system consumes approximately 2,000 watts of power. While this might seem substantial, it is remarkably efficient when compared to the computational power it delivers and the energy demands of traditional supercomputers performing similarly complex tasks. This focus on energy efficiency is paramount for the future of AI, especially as systems become larger and more sophisticated, aiming for lasting and scalable deployment. The growth of low-power, high-performance neuromorphic chips like Darwin 3 is a significant step towards creating AI that can operate effectively without exorbitant energy footprints.
Applications and Implications: From AI Models to Brain Science Research
The research team at Zhejiang University has not only built the hardware but has also successfully deployed several bright applications on Darwin Monkey,demonstrating its practical utility and immense potential.
Running Advanced AI Models
One of the most compelling applications is its ability to run the deepseek brain-like large model. This integration allows Darwin Monkey to perform a wide range of sophisticated tasks,including:
Logical Reasoning: The system can analyze complex scenarios,identify logical fallacies,and draw coherent conclusions,a capability crucial for advanced problem-solving and decision-making.
Content Generation: Darwin Monkey can generate human-like text, creative content, and even code, showcasing its prowess in natural language processing and creative AI.
Mathematical Problem-Solving: The computer can tackle intricate mathematical problems, demonstrating a deep understanding of numerical relationships and computational logic.
The ability to run these advanced AI models on a brain-like architecture signifies a paradigm shift.It suggests that neuromorphic hardware can provide a more efficient and potentially more intuitive platform for the next generation of AI algorithms, moving beyond the limitations of current deep learning frameworks.
Simulating Biological Brains: A New Frontier for Neuroscience
Beyond its AI capabilities, Darwin Monkey offers revolutionary possibilities for brain science research. Leveraging its powerful neuronal and synaptic resources, the system can preliminarily simulate animal brains of varying neuron sizes. This includes:
C. elegans: A simple nematode worm with a well-mapped nervous system of 302 neurons, serving as a foundational model for understanding basic neural circuits.
Zebrafish: A vertebrate model organism with a more complex brain, offering insights into sensory processing and motor control.
Mice: Mammalian models widely used in neuroscience research, providing a closer approximation to human brain structures and functions.
* Macaques: Primates whose brains share significant similarities with human brains, making them invaluable for