Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

neural information processing systems

Classical Simulation of Quantum Circuits: Parallel Environments and Benchmark

Xiao-Yang Liu, Zeliang Zhang

Published Jan 1, 2023• 1 citation
Publisher
DOI
10.52202/075280-2931

Related papers

Improving Decision Diagram-Based Quantum Circuit Simulation Using Static Variable Ordering and Multinode Ring Communication

Yusuke Kimura, Shaowen Li, Hiroyuki Satō +2 more

Currently the development of quantum computers is active; however, large-scale machines remain limited and noisy. Furthermore, such quantum computers do not allow direct access to state vectors, posing challenges for quantum algorithm development. Quantum circuit simulators on classical computers offer a solution, with decision diagram (DD)-based simulators being particularly memory-efficient for representing quantum states. However, DD-based simulation still requires optimization, especially concerning variable ordering and multi-node parallelization. Processing time for DDs heavily depends on variable order, but existing static variable ordering methods did not have general applicability. The prior multi-threaded DD simulations showed slowdowns. This study proposes two techniques to address these issues. The first is a scoring-based heuristic static variable ordering method that analyzes an input quantum circuit, such as the distribution of parameterized rotations and multi-bit gates, to suppress the growth of graph nodes and amount of computation. The second is a ring communication approach that reduces communication overhead when performing parallel simulations across multiple computing nodes. Parallel computation requires data exchange between nodes, which results in additional communication time. The proposed ring communication method can eliminate broadcast communication, thereby reducing the waiting time until communication is completed. Evaluations using various benchmark circuits demonstrate that the proposed ordering achieves up to a 4.5 speedup in a single-node environment. Furthermore, the ring communication exhibits superior scalability, achieving up to an 11x speedup with 16 computing nodes. The proposed variable ordering method generally reduces the overall simulation time in multi-node environments.

IEEE Transactions on Quantum Engineering•Jan 1, 2026•0 citations

Classical Simulation of Quantum Computation

Yongshan Ding, Frederic T. Chong

Synthesis lectures on computer architecture•Jan 1, 2020•0 citations

Jet: Fast quantum circuit simulations with parallel task-based tensor-network contraction

Trevor J. Vincent, Lee J. O apos Riordan, Mikhail Andrenkov +4 more

We introduce a new open-source software library Jet , which uses task-based parallelism to obtain speed-ups in classical tensor-network simulations of quantum circuits. These speed-ups result from i) the increased parallelism introduced by mapping the tensor-network simulation to a task-based framework, ii) a novel method of reusing shared work between tensor-network contraction tasks, and iii) the concurrent contraction of tensor networks on all available hardware. We demonstrate the advantages of our method by benchmarking our code on several Sycamore-53 and Gaussian boson sampling (GBS) supremacy circuits against other simulators. We also provide and compare theoretical performance estimates for tensor-network simulations of Sycamore-53 and GBS supremacy circuits for the first time.

arXiv:2107.09793•Quantum•May 9, 2022•48 citations

Massively Parallel Quantum Computer Simulations: Towards Realistic Systems

Mark Richter, Binh Trieu, G. W. Arnold +1 more

•Jan 1, 2007•5 citations

A Parallelized Qubit Mapping Algorithm for Large-scale Quantum Circuits

Dongmin Kim, Sengthai Heng, Youngsun Han

Quantum computers are superior to existing classical supercomputers in terms of run time for specific tasks, and various studies on quantum computers are in progress. Despite this, there is a significant gap in technology between available technologies and actual quantum computers on the noisy intermediate-scale quantum-scale. Currently, various quantum algorithms have been implemented and verified using quantum simulations. Therefore, the runtime performance of a quantum circuit can be verified in the same environment as a real quantum machine. A quantum compiler compiles quantum circuits. When a quantum circuit is compiled, logical qubits in a quantum circuit should be mapped to physical qubits, called qubit mapping. As qubit mapping is an NP-Complete problem, the compilation time increases exponentially according to the number of qubits. Therefore, various studies on effective qubit mapping techniques have been developed to accelerate the execution time. This paper proposes a new parallelized Qubit mapping algorithm. The proposed method consists of three-step processes: one circuit is divided into several circuits; the SABRE enhanced algorithm is executed on each circuit in parallel and then merged. The performance was evaluated using 10 benchmarks. The experimental results revealed an acceleration of up to 6.42x and an average speed up of 4.8x, with an average overhead of 1.03x in terms of the number of gates obtained. The proposed method outperformed the well-known modern qubit-mapping algorithm in terms of the execution time with exponential speedup improvement.

IEIE Transactions on Smart Processing and Computing•Feb 28, 2022•1 citation

Parallelizing quantum circuits

Anne Broadbent, Elham Kashefi

arXiv:0704.1736•Theoretical Computer Science•Jan 7, 2009•115 citations
Currently the development of quantum computers is active; however, large-scale machines remain limited and noisy. Furthermore, such quantum computers do not allow direct access to state vectors, posing challenges for quantum algorithm development. Quantum circuit simulators on classical computers offer a solution, with decision diagram (DD)-based simulators being particularly memory-efficient for representing quantum states. However, DD-based simulation still requires optimization, especially concerning variable ordering and multi-node parallelization. Processing time for DDs heavily depends on variable order, but existing static variable ordering methods did not have general applicability. The prior multi-threaded DD simulations showed slowdowns. This study proposes two techniques to address these issues. The first is a scoring-based heuristic static variable ordering method that analyzes an input quantum circuit, such as the distribution of parameterized rotations and multi-bit gates, to suppress the growth of graph nodes and amount of computation. The second is a ring communication approach that reduces communication overhead when performing parallel simulations across multiple computing nodes. Parallel computation requires data exchange between nodes, which results in additional communication time. The proposed ring communication method can eliminate broadcast communication, thereby reducing the waiting time until communication is completed. Evaluations using various benchmark circuits demonstrate that the proposed ordering achieves up to a 4.5 speedup in a single-node environment. Furthermore, the ring communication exhibits superior scalability, achieving up to an 11x speedup with 16 computing nodes. The proposed variable ordering method generally reduces the overall simulation time in multi-node environments.
We introduce a new open-source software library Jet , which uses task-based parallelism to obtain speed-ups in classical tensor-network simulations of quantum circuits. These speed-ups result from i) the increased parallelism introduced by mapping the tensor-network simulation to a task-based framework, ii) a novel method of reusing shared work between tensor-network contraction tasks, and iii) the concurrent contraction of tensor networks on all available hardware. We demonstrate the advantages of our method by benchmarking our code on several Sycamore-53 and Gaussian boson sampling (GBS) supremacy circuits against other simulators. We also provide and compare theoretical performance estimates for tensor-network simulations of Sycamore-53 and GBS supremacy circuits for the first time.
Quantum computers are superior to existing classical supercomputers in terms of run time for specific tasks, and various studies on quantum computers are in progress. Despite this, there is a significant gap in technology between available technologies and actual quantum computers on the noisy intermediate-scale quantum-scale. Currently, various quantum algorithms have been implemented and verified using quantum simulations. Therefore, the runtime performance of a quantum circuit can be verified in the same environment as a real quantum machine. A quantum compiler compiles quantum circuits. When a quantum circuit is compiled, logical qubits in a quantum circuit should be mapped to physical qubits, called qubit mapping. As qubit mapping is an NP-Complete problem, the compilation time increases exponentially according to the number of qubits. Therefore, various studies on effective qubit mapping techniques have been developed to accelerate the execution time. This paper proposes a new parallelized Qubit mapping algorithm. The proposed method consists of three-step processes: one circuit is divided into several circuits; the SABRE enhanced algorithm is executed on each circuit in parallel and then merged. The performance was evaluated using 10 benchmarks. The experimental results revealed an acceleration of up to 6.42x and an average speed up of 4.8x, with an average overhead of 1.03x in terms of the number of gates obtained. The proposed method outperformed the well-known modern qubit-mapping algorithm in terms of the execution time with exponential speedup improvement.