Dephasing experiments and modelling demonstrating unified dc and pulsed flux control with low 1/f-noise

The execution of quantum algorithms depends critically on the fidelity of two-qubit gates. In superconducting QPUs, often such gates rely on fast frequency control of qubit and coupler frequencies through flux biasing. Conventional setups pose a scalability bottleneck, as they use a separate low-speed electronics and high-speed electronics, combined via a bias tee to balance the need for stable idle biasing and rapid gate modulation. Apart from the doubling of hardware, it introduces significant bias-tee–induced pulse distortions, which are only partially correctable due to the nonlinear, memory-dependent behavior of the capacitive elements. These distortions generate nonMarkovian errors that limit gate fidelity and scalability.

The Qblox Qubit Control Module (QCM) sets a new benchmark for scalable quantum control by providing a unified source for both stable DC biasing and fast flux pulsing, eliminating the bias tee, its associated distortions, and additional DC hardware. Performance is evaluated both at the flux-insensitive sweet spot and away from the sweet spot, where flux-noise sensitivity is elevated. A physical noise model is used to determine the operational limits of the QCM’s unified control approach. Finally, we highlight infield, state-of-the-art two-qubit gate fidelities in a tunable-coupler based chip.