Algorithm-specific Performance Analysis of Transmon Qubit Devices*

Many demonstrations of quantum algorithms exhibit a tradeoff between the accuracy of the algorithm and the fidelity of the circuit that implements that algorithm. In particular, for algorithms that rely on Trotter decomposition to approximate a target unitary, increasing the number of steps reduces the algorithm error. However, in current devices, uncontrolled interactions with the environment and suboptimal control limit qubit coherence and gate fidelity, which ultimately restrict circuit depth. We analyze algorithm performance on transmon qubit devices using simulated, model-based, and experimentally measured process maps for gates, and compare a number of characterization metrics. In the case of a Trotterized algorithm, we determine the optimal operating point and predict the expected performance in good agreement with experiment.