Graph-Based Quantum Computing Simulation

Abstract

The simulation of quantum systems on classical computers faces significant challenges due to the exponential scaling of computational resources required to accurately represent quantum states and their evolution. This work proposes a novel approach to quantum computing simulation based on graph representation and graph neural networks (GNNs). We represent quantum systems as graphs, where nodes correspond to quantum bits (qubits) and edges represent the interactions between them. The dynamics of the quantum system are then learned using GNNs, which can efficiently capture complex correlations and dependencies within the system. This method offers a potential pathway to scaling quantum simulations by exploiting the inherent parallelism and learning capabilities of graph-based models. The core claim of this research is that simulating quantum systems on classical computers is computationally intractable for large systems. The core mechanism involves representing quantum systems as graphs and employing GNNs to learn system dynamics. This approach provides a new way to tackle the problem of quantum simulation.