Neuromorphic quantum computing

Area of Science:

• Quantum Computing
• Artificial Intelligence
• Neuromorphic Computing

Background:

• Correlations are fundamental to quantum computation.
• Correlations may also be key for artificial intelligence, neuromorphic computing, and biological computing.
• Exploring 'correlated computing' requires new frameworks.

Purpose of the Study:

• To demonstrate that neuromorphic computing can perform quantum operations.
• To explore the potential of correlated computing by linking neural networks and quantum systems.
• To develop a novel probabilistic computing approach beyond traditional Markov chains.

Main Methods:

• Representing spiking neuron states (active/silent) as Ising spin states (two states).
• Constructing a quantum density matrix from Ising spin expectation values and correlations.
• Learning quantum gates by adjusting neural network parameters while preserving quantum correlations.

Main Results:

• Demonstrated that neuromorphic systems can execute quantum operations.
• Showcased the ability to learn quantum gates in a two-qubit system through neural network parameter changes.
• Established that learned parameter changes adhere to constraints ensuring quantum correlations.

Conclusions:

• Neuromorphic computing offers a viable platform for performing quantum operations.
• The proposed probabilistic computing model transcends Markov chains, utilizing classical probability distribution constraints akin to quantum entanglement.
• This work provides a foundational step towards systematic exploration of correlated computing.