What is it about?
Classical quantm simulators are essential for developing quantum algorithms, but traditional full-state simulators quickly hit memory and runtime limits because they must track all possible states. Sparse simulators are much faster when only a small fraction of states are populated, yet they can become inefficient if the circuit actually explores a large part of the state space. Our work proposes an adaptive technique that, before running the circuit, quickly predicts how many basis states can become active by modeling gates as linear constraints on bit strings and applying Gaussian elimination, then automatically chooses between full-state and sparse-state backends.
Featured Image
Photo by Markus Winkler on Unsplash
Why is it important?
Today, picking the “right” simulator often relies on expert intuition or trial-and-error, which is slow and unreliable as circuits grow larger and more complex. Our method provides a fast, polynomial-time pre-analysis that typically runs in milliseconds and predicts whether a circuit will remain sparse or become dense. Across 24 benchmark circuits, the adaptive approach delivered 10–1000× speedups on sparse circuits while matching leading full-state simulators on dense ones, enabling practical simulation of many circuits beyond 50 qubits.
Perspectives
This project showed us that two circuits with the same size can be radically different in how much of the state space they actually explore. By explicitly tracking how gates expand or restrict an affine subspace of reachable basis states, we turned that intuition into a concrete signal for choosing the most efficient simulator. It was striking that a simple idea—Gaussian elimination over bit strings—could yield large, practical gains across diverse benchmarks, and we see our work as a foundation for future hybrid simulation methods that reason more deeply about interference and structure.
Kisung Jin
Electronics and Telecommunications Research Institute
Read the Original
This page is a summary of: Tracking Affine Subspace with Gaussian Elimination for Adaptive Quantum Circuit Simulation, ACM Transactions on Quantum Computing, May 2026, ACM (Association for Computing Machinery),
DOI: 10.1145/3815191.
You can read the full text:
Contributors
The following have contributed to this page
