HKU Scientists Unveil 10mK Brain-Inspired Chip for Quantum Control and Deep Space Missions
Updated
Updated · ScienceDaily · Jun 27
HKU Scientists Unveil 10mK Brain-Inspired Chip for Quantum Control and Deep Space Missions
1 articles · Updated · ScienceDaily · Jun 27
Summary
10 millikelvin is the operating point of HKU’s new neuromorphic chip, where a single silicon carbide transistor mimicked neuron-like electrical spiking near absolute zero.
The design targets a core quantum-computing bottleneck: today’s control electronics sit away from qubits because they generate too much heat, adding wiring and complicating scale-up.
Silicon carbide MOSFETs produced a strong S-shaped negative differential resistance below 2 kelvin, letting the team program stable cryogenic behavior from the material’s atomic properties rather than self-heating.
HKU said the approach could be thousands of times more energy-efficient than conventional electronics and can be manufactured in existing industrial foundries on 300-mm wafers.
The researchers also cascaded the artificial neurons into larger networks, pointing to uses in quantum error correction, real-time control and electronics for the Moon or deep-space environments.
With AI's energy crisis looming, could this brain-inspired chip be the blueprint for sustainable intelligence?
This chip brings quantum controls into the cold, but what unforeseen problems will this unprecedented integration create?
What new industries will be unlocked by electronics that thrive in the extreme cold of deep space?
HKU Unveils 10 Millikelvin Neuromorphic Chip: A Breakthrough for Cryogenic Quantum Computing and Space Missions
Overview
Researchers at Hong Kong University have developed a brain-inspired neuromorphic chip that can operate at extremely low temperatures, down to 10 millikelvin. This chip, recently published in Nature Communications, is the first programmable neuromorphic hardware to work at such cold conditions. It uses silicon carbide MOSFETs, which show a special negative differential resistance effect at cryogenic temperatures. This effect lets a single transistor mimic the energy-efficient spiking of biological neurons. These artificial neurons can also be connected to form larger networks, opening new possibilities for advanced data processing in extreme environments.