Brain-inspired computing applies the operating mechanisms of biological neural networks to computer system design — creating low-power, highly parallel, and efficient systems that begin to resemble the brain itself.
In a breakthrough that pushes the boundaries of artificial intelligence and neuroscience, researchers at Zhejiang University unveiled a brain-inspired computer powered by a dedicated neuromorphic chip with more than two billion neurons. The achievement positions the college at the forefront of the field.
A new way of computing
The human brain is an extraordinarily efficient system, integrating vision, hearing, language, learning, reasoning, and planning at a fraction of the energy a conventional processor would need for the same work.
Key specifications
- 960 Darwin-III neuromorphic chips across 15 blade servers
- 2.25 billion spiking neurons; over 200 billion synapses
- Typical power draw of roughly 2,000 watts
We are not only building faster computers — we are exploring a new way of computing, and rethinking how machines can learn.
Source: ZJU English Portal. This sample renders every element the article template supports, so a new article only needs content — never bespoke styling.