Advanced simulations and calculations on quantum computers require high fidelity implementations of quantum circuits. The universal gateset approach builds complex unitaries from manygates drawn from a small set of calibrated high-fidelity primitive gates, which results in a lower combined fidelity. Compiling a complex unitary for processors with higher-dimensional logical elements, such as qutrits, exacerbates the accumulated error per unitary because a longer gate sequence is needed. Optimal control methods promise time and resource efficient compact gate sequences and, therefore, higher fidelity. These methods generate pulses that can, in principle, directly implement any complex unitary on a quantum device. In this work, we demonstrate that any arbitrary qutrit gate can be realized with high fidelity. We generated and tested pulses for a large set of randomly selected arbitrary unitaries on two separate qutrit compatible processors, LLNL Quantum Device and Integration Testbed (QuDIT) standard QPU and Rigetti Aspen-11, achieving an average fidelity around 99 %. We show that the optimal control gates do not require recalibration for at least three days and the same calibration parameters can be used for all implemented gates. Our work shows that the calibration overheads for optimal control gates can be made small enough to enable efficient quantum circuits based on this technique.
Material defects fundamentally limit the coherence times of superconducting qubits, and manufacturing completely defect-free devices is not yet possible. Therefore, understanding theinteractions between defects and a qubit in a real quantum processor design is essential. We build a model that incorporates the standard tunneling model, the electric field distributions in the qubit, and open quantum system dynamics and draw from the current understanding of two-level system (TLS) theory. Specifically, we start with one million TLSs distributed on the surface of a qubit and pick the 200 highest coupling systems. We then perform a full Lindbladian simulation that explicitly includes the coherent coupling between the qubit and the TLS bath to model the time dependent density matrix of resonant TLS defects and the qubit. We find that the 200 most strongly coupled TLSs can accurately describe the qubit energy relaxation time. This work confirms that resonant TLSs located in areas where the electric field is strong can significantly affect the qubit relaxation time, even if they are located far from the Josephson junction. Similarly, a strongly-coupled resonant TLS located in the Josephson junction does not guarantee a reduced qubit relaxation time if a more strongly coupled TLS is far from the Josephson junction. In addition to the coupling strengths between TLSs and the qubit, the model predicts that the geometry of the device and the TLS relaxation time play a significant role in qubit dynamics. Our work can provide guidance for future quantum processor designs with improved qubit coherence times.