We present an exact synthesis approach for computing Exclusive-or Sum-of-Products (ESOP) forms with a minimum number of product terms using Boolean satisfiability. Our approach finds one or more ESOP forms for a given Boolean function. The approach can deal with incompletely-specified Boolean functions defined over many Boolean variables and is particularly fast if the Boolean function can be expressed with only a few product terms. We describe the formalization of the ESOP synthesis problem with a fixed number of terms as a decision problem and present search procedures for determining ESOP forms of minimum size. We further discuss how the search procedures can be relaxed to find ESOP forms of small sizes in reasonable time. We experimentally evaluate the performance of the SAT-based synthesis procedures on completely-and incompletely-specified Boolean functions.

Exclusive-or sum-of-products (ESOP) expressions are used as intermediate representations in quantum circuit synthesis flows, and their complexity impacts the number of gates of the resulting circuits. Many state-of-the-art techniques focus on minimizing the number of product terms in a ESOP expression, either exactly or in a heuristic fashion. In this paper, we investigate into ESOP optimization considering two recent quantum compilation flows with opposite requirements. The first flow generates Boolean functions with a small number of Boolean variables, which enables the usage of methods from exact synthesis; the second flow generates Boolean functions with many Boolean variables, such that heuristics are more effective. We focus on the reduction of the number of T gates, which are expensive in fault-tolerant quantum computing and integrate ESOP optimization methods into both flows. We show an average reductions of 36.32% in T-count for the first flow, while in the second flow an average reduction of 28.23% is achieved.

No abstract

Quantum compilation is the task of translating a quantum algorithm implemented in a high-level quantum programming language into a technology-dependent instructions flow for a physical quantum computer. To tackle the large gap between the quantum program and the low-level instructions, quantum compilation is split into a multi-stage flow consisting of several layers of abstraction. Several different individual tasks have been proposed for the layers in the flow, many of them are NP-hard. In this article, we will describe the flow and we will propose algorithms based on Boolean satisfiability, which is a good match to tackle such computationally complex problems. This article is part of the theme issue ‘Harmonizing energy-autonomous computing and intelligence’.

No abstract