Due to the significant progress made in the implementation of quantum hardware, efficient methods and tools to design corresponding algorithms become increasingly important. Many of these tools rely on functional representations of certain building blocks or even entire quantum algorithms which, however, inherently exhibit an exponential complexity. Although several alternative representations have been proposed to cope with this complexity, the construction of those representations remains a bottleneck. In this work, we propose solutions for efficiently constructing representations of quantum functionality based on the idea of conducting as many operations as possible on as small as possible intermediate representations -- using Decision Diagrams as a representative functional description. Experimental evaluations show that applying these solutions allows to construct the desired representations several factors faster than with state-of-the-art methods. Moreover, if repeating structures (which frequently occur in quantum algorithms) are explicitly exploited, exponential improvements are possible -- allowing to construct the functionality of certain algorithms within seconds, whereas the state of the art fails to construct it in an entire day.
翻译:由于在实施量子硬件方面取得重大进展,设计相应算法的有效方法和工具变得日益重要。许多这些工具依赖某些构件的功能代表,甚至完全的量子算法,然而,这些构件本身就具有指数复杂性。虽然已提出若干备选表述来应对这一复杂性,但这些表述的构建仍是一个瓶颈。在这项工作中,我们根据尽可能小的中间表达法 -- -- 使用决定图作为具有代表性的职能描述法 -- -- 来高效构建量子功能的表述法,实验性评估表明,采用这些解决方案可以建立预期的表述法,其速度要快于最先进的方法。此外,如果明确利用重复结构(经常在量子算法中出现),那么指数改进是可能的 -- -- 允许在数秒内构建某些算法的功能,而工艺的状况则无法在一整天内构建。