arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2306.05024quant-phcs.ET
收藏

用于促进 HHL 及其他算法的 NISQ 实现的近似查找表与任意函数旋转

Approximative lookup-tables and arbitrary function rotations for facilitating NISQ-implementations of the HHL and beyond

Petros Stougiannidis, Jonas Stein, David Bucher, Sebastian Zielinski, Claudia Linnhoff-Popien, Sebastian Feld

AI总结:

为促进 HHL 算法的 NISQ 实现,提出一种新颖的电路近似技术,通过空间高效的多项式函数旋转和查找表转换降低算术子例程的电路深度,在早期应用维度下实现了更小电路且保持良好近似。

摘要:

许多具有可证明加速的量子计算有前景的应用都以 HHL 算法为核心。由于硬件限制及其在已知实现中对量子比特和门的大量需求,其在近期量子计算机上的执行是不可行的。为了促进此类 NISQ 实现,我们提出了一种新颖的电路近似技术,该技术增强了 HHL 中的算术子例程,这些子例程在小规模环境中是特别消耗资源的组件。为此,我们描述了多项式函数的空间高效旋转的算法实现,这些旋转不需要在量子电路内部进行显式的算术计算。我们展示了如何通过提供一种简单而强大的近似技术来降低此类电路的深度。此外,我们提供了一种算法,将任意函数旋转的查找表转换为允许应用该近似技术的结构。这使得能够为许多多项式和非多项式函数实现近似旋转电路。在现实的早期应用维度下获得的实验结果表明,与现有技术相比有显著改进,在实现良好近似的同时生成了较小的电路。

英文摘要:

Many promising applications of quantum computing with a provable speedup center around the HHL algorithm. Due to restrictions on the hardware and its significant demand on qubits and gates in known implementations, its execution is prohibitive on near-term quantum computers. Aiming to facilitate such NISQ-implementations, we propose a novel circuit approximation technique that enhances the arithmetic subroutines in the HHL, which resemble a particularly resource-demanding component in small-scale settings. For this, we provide a description of the algorithmic implementation of space-efficient rotations of polynomial functions that do not demand explicit arithmetic calculations inside the quantum circuit. We show how these types of circuits can be reduced in depth by providing a simple and powerful approximation technique. Moreover, we provide an algorithm that converts lookup-tables for arbitrary function rotations into a structure that allows an application of the approximation technique. This allows implementing approximate rotation circuits for many polynomial and non-polynomial functions. Experimental results obtained for realistic early-application dimensions show significant improvements compared to the state-of-the-art, yielding small circuits while achieving good approximations.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑