Spiking Neuromorphic Chip Encodes Quantum Entanglement Correlations

Analog neuromorphic chips, inspired by structural and dynamical properties of the biological brain, show a high energy efficiency in running artificial neural-network architectures for the profit of generative applications. Together with recent proposals for artificial neural networks to encode quantum states, this encourages employing such hardware systems as platforms for simulations of quantum system or quantum state tomography.
Here we report on the realization of a prototype using the spike-based BrainScaleS hardware developed in the context of European’s Human Brain Project (HBP). This chip realizes fast analog dynamics to boost computationally expensive tasks.
The probabilistic implementation of quantum states is achieved through Bayesian sampling by the spiking neurons. Furthermore, a specific Hebbian learning scheme exploits the hardware speed and allows for a variety of network topologies. Training the hardware-encoded network to represent maximally entangled quantum states of up to four qubits reaches high fidelities. Extracted Bell correlations for two-qubit states convey that non-classical features are captured by the analog hardware, demonstrating the feasibility of simulating quantum systems with spiking neuromorphic chips.