Duality Quantum Computing with Subwave Projections

TL;DR

This paper introduces subwave-projection duality quantum computing (SWP-DQC), which enhances the flexibility and speed of quantum algorithms by enabling linear combinations of non-unitary operations through projections on subwaves.

Contribution

The paper proposes SWP-DQC, a novel extension of duality quantum computing that allows projections on subwaves, leading to faster algorithms and reduced qubit requirements.

Findings

SWP-DQC achieves an O(M) speedup over traditional DQC.

It enables linear combinations of non-unitary operations.

Optimized ground state preparation algorithm saves up to log2 N qubits.

Abstract

Duality quantum computing (DQC) offers the use of linear combination of unitaries (LCU), or generalized quantum gates, in designing quantum algorithms. DQC contains wave divider and wave combiner operations. The wave function of a quantum computer is split into several subwaves after the wave division operation. Then different unitary operations are performed on different subwaves in parallel. A quantum wave combiner combines the subwaves into a final wave function, so that a linear combination of the unitaries are performed on the final state. In this paper, we study of the properties of duality quantum computer with projections on subwaves. In subwave-projection DQC (SWP-DQC), we can realize the linear combinations of non-unitaries, and this not only gives further flexibility for designing quantum algorithms, but also offers additional speedup in the expected time complexity. Specifically, SWP-DQC offers an O(M) acceleration over DQC with only final-wave-projection in the mean time complexity, where M is the number of projections. As an application, we show that the ground state preparation algorithm recently proposed by Ge, Tura, and Cirac is actually an DQC algorithm, and we further optimized the algorithm using SWP-DQC, which can save up to $\log_2 N$ qubits compared DQC without subwave projection, where N is the dimension of the system's Hilbert Space.