qiskit-metal
qiskit-metal copied to clipboard
qiskit-community
The 'Solvay Conference' on Advanced Topics In Simulating Superconducting Devices
What is the feature being requested?
Several discussions occured during the course of the tutorials on using Qiskit Metal which revolved around best practices in simulation and advanced topics. The idea was brought up that these are problems everyone has and there are pockets of information that exist throughout labs in the world. There doesn't seem to be a workshop/conference centered around this idea of model matching to experiments. It would be a good idea to maybe initiate some sort of laying out for the community these techniques that have been developed to push the field of device modelling forward.
What are use cases for this feature?
-Better understanding of superconducting qubits across more theoretical and advanced levels of application
Some possible ideas:
- Modelling the distribution of TLS's in devices and impacts on device performance
- How much BCS can we ignore in device design (when do we have to start treating devices as superconductors rather than perfect conductors)
- How can we capture ubiquitous phenomena like kinetic inductance which is often ignored
- Understanding how to model proximity effects between various superconductors or normal metals
- Best practices for numerical (full and block-) diagonalization of multiqubit Hamiltonians, including the choice of bases, truncation, tolerances and numerical methods.
- EPR Numerical diagonalization and Hamiltonian construction with ZPFs
- Comparison between EPR and impedance modeling.
- How to make sure the simulation is accurate
- did I catch all the spurious modes
- does the mesh make sense
- is the frequency sweep dense enough
- How to simulate in parts (not technically, but conceptually, how to design without significant crosstalks)
- bonds to connect separated ground planes and flux traps - simulating accurately is hard, when and how to ignore
- The voodoo of sharp corners in MW design
- Modeling the box / making sure it does not affect the chip
Tagging here @nbronn and @marcolincs, we may want to organize this midway or near end of open tutorials. I think for first one we can just self organize with Kallie. We can plan for something in 3-4 months from now.
