Hardware-efficient quantum computation and error correction in bosonic systems

Abstract

Quantum computing has the potential to eclipse its classical competitors, but only if the number of high-quality qubits can be scaled up. Large-scale quantum systems are impeded by the formidable hardware resources needed to combat growing amounts of errors from hardware imperfections. Previous efforts have mainly focused on either optimizing quantum hardware or finding new quantum algorithms. This thesis explores synergies between the system-specific hardware physics and algorithm design that together yield more than the sum of their parts in the quest for scalable quantum computation in bosonic systems. We present an algorithm for generating nonclassical states of light, using full-quantum X( 2 ) nonlinearities, that transcends previous limits on conversion efficiency. We show that such nonlinearities-which enable highly efficient three-wave mixing between quantized signal, idler, and pump fields-can be employed in two systematic frameworks for quantum computing. The first, which utilizes X(2) interactions' fundamental symmetries and recognizes that photon-loss is their dominant source of errors, provides a set of hardware-efficient quantum error-correction codes and their associated encoded universal gates. One of our codes achieves a constant rate of protected photons, a necessity for robust large-scale quantum computation. The second framework provides hardware-efficient universal quantum control facilitating plug-and-play application of machine learning algorithms. It takes constraints on the hardware resources and control-error models as inputs, and returns robust control pulse shapes for high-fidelity quantum gate execution. The transformative performance gains obtained from this hardware-efficient approach offer potential for scalable quantum computation using available quantum devices.