Qompose: A Technique to Select Optimal Algorithm- Specific Layout for Neutral Atom Quantum Architectures

TL;DR

Qompose tackles the problem of topology selection for neutral-atom quantum architectures by predicting the most effective 2-D atom topology for a given circuit. It combines quantum descriptors with PageRank-inspired features in a neural predictor to choose among square, S-Triangle, and T-Triangle layouts, then maps, inserts SWAPs, and schedules the circuit accordingly. Empirical results show improvements in critical-path pulse count (up to 8.4% on random circuits and 5.4% on real-world benchmarks) and near-Oracle total pulses, along with fidelity considerations under realistic noise. The work demonstrates that topology-aware compilation can yield substantial performance gains in neutral-atom systems and provides an open-source framework for exploring this design space across algorithms and benchmarks.

Abstract

As quantum computing architecture matures, it is important to investigate new technologies that lend unique advantages. In this work, we propose, Qompose, a neutral atom quantum computing framework for efficiently composing quantum circuits on 2-D topologies of neutral atoms. Qompose selects an efficient topology for any given circuit in order to optimize for length of execution through efficient parallelism and for overall fidelity. our extensive evaluation demonstrates the Qompose is effective for a large collection of randomly-generated quantum circuits and a range of real-world benchmarks including VQE, ISING, and QAOA.

Figures (10)

• Figure 1: As a Rydberg interaction is created around qubit q2, all adjacent qubits are blocked in the restriction zone and cannot be entangled with any other qubits using the same frequency at the same time.
• Figure 2: Topology arrangement options. The three topologies pictured each pose unique advantages for certain circuit types. For example, lower circuit width accurately predicts the s-triangular topology for reducing the critical pulse path.
• Figure 3: Demonstration of activation radius blocking. Each topology blocks nearby qubits differently, which is what leads to unique advantages offered by each.
• Figure 4: An overview of all of the steps of Qompose: model training, topology prediction, mapping, scheduling, and execution.
• Figure 5: The logical circuit passed into Qompose is already parallelized. This transfer of the parallelism for neutral atom circuits is not trivial and must be reconstructed with a best-effort approach. However, some topologies are more suited to the parallelism of certain logical circuits than others.