Streamlining superconducting qubit design with an open source workflow

The ability to engineer the Hamiltonian of superconducting qubits makes them a promising platform for quantum computing and sensing systems. However, translating an effective circuit Hamiltonian into a physical device layout requires advanced modeling tools. Electromagnetic finite element method (FEM) simulations are essential for superconducting qubit design to extract device characteristics such as classical circuit parameters, energy participation ratios (EPR) for circuit quantization, and crosstalk. Current practice in the quantum computing field relies heavily on commercial software, such as COMSOL and Ansys HFSS, for electromagnetic device simulation. As an open-source alternative, we present a workflow for the design and modeling of superconducting qubit devices using Gmsh for mesh generation, AWS Palace for FEM modeling, and ParaView for mode visualization. We also compare mesh generation between Gmsh and Cubit, a closed-source meshing toolkit. Finally, we benchmark the performance and accuracy of AWS Palace with respect to simulation parameters, mesh size (with and without adaptive mesh refinement), and scaling across high-performance computing (HPC) resources.