Predicting noisy quantum circuit simulation runtime after transpilation
Transpilation-Aware Runtime Prediction for Noisy Quantum Circuit Simulation
Performance
Summary
Figuring out how long it takes to simulate quantum circuits with noise is tricky because the circuits often change after being prepared for a specific quantum computer backend. The authors studied different ways to predict simulation time, comparing graph neural networks that look at circuit structure before and after these changes with traditional regression methods. They found that using information from after the changes (transpilation) helps make better predictions, but traditional methods can sometimes match or beat neural networks depending on the simulation setup. This work helps improve planning and resource use in noisy quantum computing simulations.
What this means in practice
- •For quantum software developers: Estimate simulation runtime after transpilation to optimize scheduling on quantum simulator platforms.
- •For quantum hardware operators: Allocate computational resources efficiently by predicting noisy quantum simulation costs specific to backend settings and optimization levels.
Authors
Davud Azizov, Javier Vela-Tambo, Tian Guo
Abstract
Predicting the runtime of noisy quantum circuit simulations is important for scheduling, resource allocation, and performance optimization. However, accurate prediction is challenging because backend-aware transpilation can substantially alter the original circuit structure, while the backend-derived noise model and simulator execution behavior can introduce additional runtime variation. We study the effectiveness of graph neural networks (GNNs) and conventional regression methods in predicting Qiskit Aer simulation runtime measured after transpilation. We construct a dataset from a benchmark pool of 1,402 unique circuits spanning 22 circuit families, two Qiskit fake-backend configurations, and four transpiler optimization levels. Specifically, we compare a source GNN using original circuit information, hybrid GNN combining source-level graph with post-transpilation features, and transpiled GNN using only transpiled circuit information, along with five regression models. In the overall-model setting, the transpiled GNN achieves the strongest performance among the graph-based representations at all four optimization levels, obtaining $R^2$ values of 0.974, 0.713, 0.745, and 0.605 for optimization levels 0 through 3, respectively. However, under backend-specific evaluation, the advantage of GNN decreases, with conventional regression models matching or outperforming the GNNs in several settings. These results indicate that post-transpilation information is useful, while the value of explicit graph modeling depends on the backend and optimization level.