Description
In recent years, spin qubits have achieved remarkable progress, surpassing 99% fidelity in both single- and two-qubit operations for the first time. Despite these advancements, several challenges remain critical to the successful development of this technology. One such challenge for spins in silicon metal-oxide-semiconductor devices involves engineering a global drive capable of addressing potentially millions of spin qubits, while meeting stringent requirements for field uniformity and minimizing induced electric noise. A proposed solution in the literature involves utilizing an off-chip dielectric resonator made of potassium tantalate (KTaO₃) to deliver the global drive to spin qubit devices. This approach effectively reduces the device's footprint and mitigates heating concerns.
Characterizing the noise introduced by the resonator, as well as the mechanisms through which it couples to the qubits, is essential for mapping the strengths and limitations of this architecture. In this work, we estimate the impact of the resonator on qubits performances, focusing on driving speed and coherence times. We disentangle noise sources that are proportional to the amplitude of the magnetic field generated, from noise mechanisms that are only activated when the qubits are resonant with the dielectric resonator. Furthermore, by sweeping the orientation of the external DC magnetic field and tracking the Rabi frequency of the qubits, we reconstruct the orientation of the field generated by the resonator. Our findings are supported by finite-element method (FEM) electromagnetic simulations, which indicate that currents forming in the gate nanostructure account for the observed difference in qubits performances, and Hamiltonian simulations, detailing how different components of the electric field perceived by the qubits directly affects coherence metrics. Understanding and reconstructing the noise introduced by this driving architecture represents a crucial step toward the scalable implementation of spin-qubit quantum processors.
| I am the presenting author | Yes |
|---|