We revisit the Kitaev model for fault tolerant quantum computing on a square lattice with underlying quantum double D(G) symmetry, where G is a finite group.
Category (publication): SDK
Every Classical Sampling Circuit is a Quantum Sampling Circuit
We introduce “Q-marginals,” which are quantum states encoding some probability distribution in a manner suitable for use in Quantum Monte Carlo Integration (QMCI) and show that these can be prepared directly from a classical circuit sampling for the probability distribution of interest.
Application-Motivated, Holistic Benchmarking of a Full Quantum Computing Stack
We propose “application-motivated” circuit classes for benchmarking: deep, shallow and square quantum circuits. Using systems made available by IBM Q, we examine their performance, showing that noise-aware compilation strategies may be beneficial and that device connectivity and noise levels play a crucial role in the performance of the system.
Architecture-Aware Synthesis of Phase Polynomials for NISQ Devices
We propose a new algorithm to synthesise quantum circuits for phase polynomials, which takes into account the qubit connectivity. This work focuses on the architectures of current NISQ devices. The resulting algorithm generates circuits with a smaller CNOT depth than those currently used in Staq and Tket, while improving the runtime with respect to the former.
Using Reinforcement Learning to Perform Qubit Routing in Quantum Compilers
We demonstrate that the qubit-routing problem has a natural interpretation as a reinforcement learning problem. The results show state-of-the-art performance when qubit routing is treated as an abstracted problem and suggest that reinforcement learning may lead to further gains being made when addressing backend optimisation more generally.
A Generic Compilation Strategy for the Unitary Couple Cluster Ansatz
We describe a compilation strategy for Variational Quantum Eigensolver (VQE) algorithms which use the Unitary Coupled Cluster (UCC) ansatz. This strategy reduces cx depth by 75.4% on average and by up to 89.9% compared to naive synthesis for a variety of molecules, qubit encodings and basis sets.
Dual-Parameterized Quantum Circuit GAN Model in High Energy Physics
In a collaboration with CERN, we propose a dual-PQC GAN for generative modelling applications in High-Energy Physics. This development enables the generation of samples from an ensemble of typical images – something not possible with conventional qGAN architectures.