The booming field of quantum computing offers unprecedented exponential computational speedup. However, there are still a lot of challenges to realizing large-scale quantum computers. One of them is related to the ultra-low-power dissipation requirement for cryogenic electronics controlling quantum bits (qubits) and correcting their errors. Commercial nano-transistors dissipate too much power at low temperatures to be used for cryogenic electronics in a future quantum computer consisting of millions of qubits. Here we propose an alternative, non von Neuman processor architecture based on brain-neuron-mimicking artificial neural networks functioning down to 1 Kelvin (-272 Celsius). Fabricated within standard Si transistor technology, our cryo-neuromorphic processor offers fast and ultra-low-power computation capabilities to control large numbers of qubits. Moreover, it can be co-integrated with spin qubits on the same silicon chip, thus offering great scalability prospects.

2022

2025