Design and efficiency in graph-state computation

TL;DR

This paper introduces Etch, an open-source circuit-etching tool for quantum circuit transpilation to graph states, evaluating its efficiency and potential as an alternative to existing graph-based strategies in quantum computation.

Contribution

The paper presents Etch, a novel circuit-etching tool, and assesses its effectiveness compared to traditional graph strategies in quantum circuit compilation.

Findings

Etch can transpile IQP circuits to graph states with higher Pauli to non-Pauli qubit ratios.

Circuit etching shows promise as a resource-efficient strategy for magic state distillation.

Future research will explore optimizing both graph and circuit-etching strategies.

Abstract

The algorithm-specific graph and circuit etching are two strategies for compiling a graph state to implement quantum computation. Benchmark testing exposed limitations to the proto-compiler, Jabalizer giving rise to Etch (https://github.com/QSI-BAQS/Etch), an open-source, circuit-etching tool for transpiling a quantum circuit to a graph state. The viability of circuit etching is evaluated, both as a resource allocation strategy for distilling magic states and as an alternative to the algorithm-specific graph strategy as realised in Jabalizer. Experiments using Etch to transpile IQP circuits to an equivalent graph state resulted in higher ratios of Pauli qubits to non-Pauli qubit than required for efficient magic state distillation. Future research directions for the algorithm-specific graph and circuit-etching strategies are proposed.