Hai-Tian ZHANG

AI is undeniably a revolutionary force that could potentially transform our society. However, traditional AI hardware has high energy consumption and low computational efficiency, which remains significantly inferior to the human brain's low energy usage and high efficiency. Therefore, discovering new mechanisms to address the limitations of computing power has become one of the most pressing issues, and constructing brain-like devices is one of the primary routes to overcome this challenge.

Hai-Tian Zhang introduced a novel approach to bridging the gap between artificial intelligence devices and the human brain. By precisely manipulating the motion of hydrogen dopants in strongly correlated oxide thin films, he has developed artificial neuron and synapse hardware that feature ultra-low energy consumption and ultra-fast responses.

Traditional oxide neural devices rely on the movement of large cations (e.g., Ag+) and anions (e.g., O2-), consuming energy at the pJ level. In contrast, Hai-Tian used hydrogen ions (protons), which are only 0.01% the size of O2-. Taking advantage of their small size and low movement energy barrier, the devices he developed require only fJ-level energy for nanosecond neural computations, making their efficiency comparable to that of the human brain.

Moreover, through various novel neuromorphic computing devices, he has introduced breakthrough functionalities in AI, particularly demonstrating performance that far exceeds traditional neural networks in handling complex environmental changes, such as a 250% improvement in decision-making ability and a 300% enhancement in dynamic information processing capability.