Widening the Bottleneck at the Quantum-Classical Interface

Abstract

Quantum computing's promise to solve complex, uncomputable problems is likely to require systems with billions of qubits. Creating an efficient interface between such a vast number of qubits and classical systems represents an enormous challenge, critical to fulfilling that promise. In this thesis, I demonstrate that the readout and control subsystems comprising this interface can be engineered in new ways that overcome current non-scalable limitations.

Connecting billions of qubits to classical systems demands innovative solutions to address power dissipation, readout and control overhead, and space constraints. These issues stem from the bottleneck in the cryogenic fridge housing the quantum systems, where ultra-low temperature requirements impose space and cooling power limitations. Reducing the per-qubit load is a requirement of scale-up, and while many novel approaches have been proposed, questions remain on their effect on the qubits themselves.

If future qubits have significantly higher fidelities than those available today, then far fewer are needed to achieve the same level of effectiveness. My first experiment explores this possibility by examining a measurement-only approach to topological states. Topological architectures have long promised qubits far more robust to noise compared to those of today, however engineering challenges have prevented their realisation. A measurement-only approach alleviates some of these challenges, and to validate this idea, a trivial topological state (the Aharonov-Bohm effect) is measured using a dispersive gate sensing (DGS) technique. We find that this state can be read out using DGS, with the measurement speed far surpassing that of conventional transport measurement methods. Applying DGS to other architectures also reduces their per-qubit load, as it eliminates the need for a discrete charge sensor, thereby decreasing the number of connections required for a quantum device.

Billions of qubits will require billions of...