Making optimal control work for superconducting qubits

Optimal control is a powerful tool to find pulse sequences and shapes that implement quantum gates on given realistic hardware, such as superconducting qubits. For practical applicability, the intense filtering as well as the challenge of precise calibration need to be met. We illustrate this challenge along the example of the implementation of a controlled phase shift gate between transmon qubits coupled by a resonator. The pulse sequence reaches a high fidelity gate in short time but is highly sensitive to the characterization of the transmon frequencies. We show how a combination of robust optimal control, genetic algorithms, and plain debugging help meet these challenges.