Benchmark Study for Quantum Statevector Simulators for Scientific Computing
Abstract
Quantum circuit simulators are essential for quantum algorithm development, yet systematic runtime benchmarks comparing major frameworks across CPU and GPU backends are limited. We benchmark three open-source frameworks - Cirq (qsim), CUDA-Q, and Qiskit (Aer) - on statevector simulation of the 2D quantum acoustic wave equation, evolved via Trotterized Hamiltonian dynamics, for problem sizes from 7 to 29 qubits on a single Azure virtual machine. The evaluation spans four categories: CPU runtime versus qubit count, strong scalability on multi-core CPUs at 27 qubits, GPU runtime versus qubit count, and GPU-over-CPU speedup. Relative rankings shift from one category to the next. On CPU, Cirq is fastest up to 23 qubits (2.4x faster than Qiskit at 21 qubits), but Qiskit overtakes it at larger sizes and higher thread counts, sustaining 80% parallel efficiency at 16 threads. On GPU, CUDA-Q leads between 13 and 21 qubits, while Cirq becomes the fastest GPU backend at large problem sizes, achieving 1.5x and 2.3x speedups over CUDA-Q and Qiskit at 29 qubits. GPU-over-CPU speedup ratios should be interpreted alongside absolute GPU runtimes; CUDA-Q's 1500x at 21 qubits is largely driven by the limited scaling of its CPU backend (qpp), and the absolute GPU runtimes show that the three frameworks' GPU backends differ by less than these ratios alone suggest. These results indicate that no single framework dominates across configurations, and that backend selection for statevector simulation should be guided by both target hardware and problem size.
Related
- The spirit of independent innovation: an exploratory study of Mr. J.S. Lee's scientific research traits
- A center of expertise and scientific mediation on the lake ecosystems: Inform, train, and provide scientific support to all stakeholders
- LuKARS 3.0: a high-performance computing software to model flow and transport processes in karst aquifers
- After "nature"? Using the concept of nature in scientific and policy contexts: an "indisciplinary" approach
- Level of Scientific Literacy in Junior High School Students: Survey Research Based on Grade Level
BibTeX
@misc{yeo2026benchmark,
title = {{Benchmark Study for Quantum Statevector Simulators for Scientific Computing}},
author = {Timothy M. Yeo and Rodney W. Lessard},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.01176},
eprint = {2611.01176},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.01176},
}
RIS
TY - UNPB
TI - Benchmark Study for Quantum Statevector Simulators for Scientific Computing
AU - Timothy M. Yeo
AU - Rodney W. Lessard
PY - 2026
DA - 2026/11/01
PB - arXiv
DO - 10.48550/arxiv.2611.01176
UR - https://arxiv.org/abs/2611.01176
AB - Quantum circuit simulators are essential for quantum algorithm development, yet systematic runtime benchmarks comparing major frameworks across CPU and GPU backends are limited. We benchmark three open-source frameworks - Cirq (qsim), CUDA-Q, and Qiskit (Aer) - on statevector simulation of the 2D quantum acoustic wave equation, evolved via Trotterized Hamiltonian dynamics, for problem sizes from 7 to 29 qubits on a single Azure virtual machine. The evaluation spans four categories: CPU runtime versus qubit count, strong scalability on multi-core CPUs at 27 qubits, GPU runtime versus qubit count, and GPU-over-CPU speedup. Relative rankings shift from one category to the next. On CPU, Cirq is fastest up to 23 qubits (2.4x faster than Qiskit at 21 qubits), but Qiskit overtakes it at larger sizes and higher thread counts, sustaining 80% parallel efficiency at 16 threads. On GPU, CUDA-Q leads between 13 and 21 qubits, while Cirq becomes the fastest GPU backend at large problem sizes, achieving 1.5x and 2.3x speedups over CUDA-Q and Qiskit at 29 qubits. GPU-over-CPU speedup ratios should be interpreted alongside absolute GPU runtimes; CUDA-Q's 1500x at 21 qubits is largely driven by the limited scaling of its CPU backend (qpp), and the absolute GPU runtimes show that the three frameworks' GPU backends differ by less than these ratios alone suggest. These results indicate that no single framework dominates across configurations, and that backend selection for statevector simulation should be guided by both target hardware and problem size.
ER -