metal-library
is a powerful framework made to expedite the design and simulation phase of a superconducting quantum devices. Its primary objective is to reduce the time spent by design engineers by providing an initial estimation of a components characteristics. We do this by comparing a component's geometry to experimentally verified quantities of qubits and cavities.
As a design engineer, your current workflow probably looks like:
PI tells you they want a qubit w/ certain target observables. For example: "Give me a TransmonCross (
This is super long and boring.
What if it could be faster?
Luckily that's the point of this library! In our case, you just tell the library, "Hey I want a TransmonCross coupled to a cavity w/ this
How can I trust the output of this library?
All simulation done in this library have been verified by experiment.
What I mean is:
- Get experimental characteristics from fabricated devices
- Get theoretical characteristics by running simulation of the same devices.
- If there's "good enough" agreement, we run the same simulation setup on various geometries close to the fabricated devices.
What do you mean by "good enough" agreement
NOTE TO SELF: ask Eli what to put here or if to rewrite?
- Install Qiskit-Metal. See here
- Clone repository into a directory of your choice
cd <REPO PATH>
git clone https://github.com/LFL-Lab/metal-library
- Install dependencies in your Qiskit Metal conda environment.
conda activate <QISKIT METAL ENV>
cd metal-library
pip install .
To update the repository
conda activate <QISKIT METAL ENV>
cd <REPO PATH>
git pull
pip install --upgrade .
from metal_library import