Hypergraphic partitioning of quantum circuits for distributed quantum computing

TL;DR

This paper introduces a hypergraphic partitioning method for quantum circuits in distributed quantum computing, significantly reducing communication costs between partitions in NISQ devices.

Contribution

It presents a novel hypergraphic heuristic partitioning algorithm that decreases communication qubits, improving scalability of distributed quantum circuits.

Findings

Over 50% reduction in communication qubits for bipartite partitioning

Effective partitioning on benchmark circuits

Potential for improved multi-partition scenarios

Abstract

In the context of NISQ computers - Noise Intermediate Scale Quantum, it is a consensus that the distribution of circuits among processing agents is a viable approach to get greater scalability with small machines. This approach can increase the combined computational power at the cost of dedicating qubits to get the communication between partitions. Then comes the challenge of reducing this cost of communication-based on efficient circuit partitioning strategies. In this context, this work aims to reduce the consumption of communication qubits generated during the physical partitioning of quantum circuits in the distributed quantum computing scenario. We present a new method for partitioning quantum circuits in a hypergraphic representation, using a heuristic partitioning algorithm for this, reducing the number of communication qubits between the partitions. With this approach, we obtained partial results with a more than 50% reduction in the communication cost generated for the bipartite partitioning against a process done randomly on benchmark circuits. The expectation is that future experiments with multiple partitions will continue with equal or better results than previous works, supporting distributed quantum computing with machines in heterogeneous processing and communication scenarios.