Executive Summary : | Artificial intelligence (AI) has revolutionized our world, but classical neural networks running on von-Neumann computing architecture have limitations compared to the human brain. These architectures have a data bus that is loaded by data movement and have a slower speed than the processor. These architectures are power-hungry and not sustainable for complex AI applications. To evolve and help various industries and emerging technologies, two compute paradigms (neuromorphic computing and quantum computing) are proposed. Neuromorphic computing, also known as brain-inspired computing, mimics the human nervous and cerebral systems, but its hardware implementation limits scalability, generalization, and reconfigurability. This research project proposes an unorthodox approach to neuromorphic computing, which takes inspiration from quantum correlation (entanglement and superposition). The proposed neuromorphic quantum computing platform is energy-efficient, scalable, and eliminates strict requirements such as isolation from the environment, milli-Kelvin temperature, and decoherence. The core strength of quantum computing is the quantum entanglement of qubit states, which is intrinsically related to the state of other qubits. The proposed neuromorphic quantum computing platform uses spiking/non-spiking silicon artificial neurons in a network to perform sequences of arbitrary unitary operations, such as Hadamard gates, CNOT-gates, and rotations in Hilbert space. These gates can be quantum computing relevant for information processing and solving complex real-world problems. |