Benchmarking Quantum Circuit Emulators on HPC Cluster

Quantum circuit emulators are essential tools for testing and studying quantum algorithms, particularly when access to quantum hardware is limited. In this thesis, we focus on tensor network-based emulators, which simulate quantum circuits using tensor network methods—an approach that efficiently represents the wavefunction, especially when entanglement across qubits is limited. We compare the performance of two tensor network-based emulators: QMatcha Tea, which runs on both CPU and GPU, and cuQuantum, which is optimized for GPU acceleration, when executed on an HPC cluster. Both simulators utilize tensor networks but they differ in hardware optimizations and specific implementation techniques. We test both simulators on the same quantum circuit, the Quantum Fourier Transform (QFT), across varying numbers of qubits. Our study benchmarks these tools in terms of computational efficiency, scalability, and accuracy, providing valuable insights into their strengths and limitations for classically simulating quantum algorithms.