Pith. sign in

REVIEW 4 major objections 7 minor 36 references

Fewer Histories, Faster Paths: Distributed Quantum Circuit Feynman Simulation via History Reduction, Checkpointing, and Pruning

T0 review · 4 major / 7 minor · reviewed 2026-08-06 · deepseek-v4-flash

Pith's one-line read A reduced Feynman path sum computes selected quantum amplitudes exactly, with the cost set by residual branching choices rather than qubit count.

desk verdict A sound, well-engineered sparse-output Feynman simulator with genuine history reduction; the main open question is how often the magic variable A stays small, and the paper doesn't prove it. read the letter →

arxiv 2608.01467 v1 pith:KMQLDVTK submitted 2026-08-02 cs.ET

classification cs.ET
keywords quantumcircuitsimulationFeynmansum-over-historiesartificialsourcessparse-outputcheckpointingpruningdistributedparallel
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to show that exact sparse-output quantum circuit simulation can be done by summing over classical histories, without ever building the full state vector. The key reduction is that most wire values in a history are forced by the input, the output, and the circuit's deterministic gates; only a small set of unresolved branching choices needs to be summed over. That reduces the history space from all internal-wire assignments to a sum over a few artificial sources, and the paper demonstrates the approach on QFT, amplitude amplification, QAOA, and 100-qubit quantum walks at distributed supercomputer scale. This matters because many quantum workloads only need a few output amplitudes, and exact answers at those sizes are currently out of reach of full-state simulation.

What carries the argument

The central object is the reduced-history formulation: a Feynman history is described by internal-wire values, but instead of summing over all $\{0,1\}^{|W|}$ assignments, the simulator propagates boundary constraints through deterministic gates (basis permutations) and wire-preserving gates (phase-only), introducing a binary artificial source only where branching remains unresolved. The identity $\alpha_{b_{\mathrm{in}},b_{\mathrm{out}}}=\sum_{a\in\{0,1\}^A}\chi(a)\alpha_a$ replaces the full path sum by a sum over $A$ artificial sources, and checkpointing partitions the circuit into prefix/root/suffix so shared partial amplitudes are reused. The cost expression $O(SM2^A G)$ carries the argu

What would settle it

Run the simulator on a circuit family deliberately built so that backward propagation cannot resolve most wires, e.g. many entangled Hadamards placed after layers that obscure both boundaries, and count $A$. If $A$ scales linearly with the number of qubits, the per-output cost becomes $O(2^n)$, matching full state-vector simulation and contradicting the method's advantage; the paper itself classifies random circuit sampling as unfavorable, so a concrete such family would mark the boundary of the claim.

Watch

Extended reading notes

Core claim

Exact sparse-output simulation of a quantum circuit can be done as a pure Feynman sum over histories, provided the sum is over residual branching choices rather than all internal wire assignments. For fixed input and output bitstrings, boundary constraints propagate through deterministic and wire-preserving gates, so only $A$ unresolved branching points (artificial sources) need explicit summation: $\alpha_{b_{\mathrm{in}},b_{\mathrm{out}}}=\sum_{a\in\{0,1\}^A}\chi(a)\alpha_a$. The worst-case cost is $O(SM2^A G)$, with $S$ the input support, $M$ the requested outputs, and $G$ the gate count. On the studied families this reduction is strong: $A=0$ for QFT under backward analysis, checkpointin

Load-bearing premise

The practical advantage relies on backward propagation of determinism leaving only a small number of unresolved branch choices for the circuits being simulated; the paper asserts this optimality for its benchmark families by graphical inspection but does not prove it, and if that number grows with circuit width the cost $2^A$ quickly becomes prohibitive.

Editorial extensions

If this is right

  • For QFT, backward propagation leaves $A=0$, so each requested amplitude costs a single propagated history; exact frequency-bin selection becomes linear in the number of requested outputs and input support rather than exponential in qubit count.
  • Checkpointing reuses propagated states and partial amplitudes across histories that share artificial-source assignments, reducing gate operations from $2^{A_p+A_r+A_s}(G_p+G_r+G_s)$ to $2^{A_p}(G_p+2^{A_r}(G_r+2^{A_s}G_s))$, and the autotuned partition delivers the measured speedups.
  • Threshold pruning gives a controllable fidelity-runtime knob: for QAOA, thresholds up to about $10^{-5}$ preserve near-unit fidelity, while larger thresholds cut runtime by nearly an order of magnitude as fidelity drops to roughly one half.
  • Because requested outputs are independent, the workload can be split across outputs and scheduled asynchronously; on a 100-qubit, 16-step quantum walk the scheme reconstructs exact selected-output distributions and reaches 85% parallel efficiency at 8,192 cores.
  • Against a tensor-network baseline on quantum-walk circuits, the method is more than an order of magnitude faster, with part of the gain coming from simulating large deterministic gates at low cost.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The authors do not test Hamiltonian simulation circuits, but their reduction suggests that any circuit whose branching gates are few and localized, regardless of width, would inherit the same $2^A$ advantage; that is a direct, testable extension.
  • The artificial-source placement problem on the CNOT-plus-Hadamard skeleton looks like a graph domination problem, so a formal characterization of when backward propagation is optimal, or a bounded-approximation placement algorithm, would turn the empirical $A$ values into a worst-case guarantee.
  • The pruning sweep defines an implicit error budget; a probabilistic bound on the total truncated amplitude as a function of threshold $t$ would let users set $t$ from a target fidelity instead of sweeping it empirically.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 7 minor

Summary. The paper presents an exact sparse-output quantum circuit simulator based on pure Feynman sum-over-histories. For a circuit, an input state support, and requested output bitstrings, it computes selected amplitudes by propagating boundary constraints through deterministic and wire-preserving gates, introducing artificial sources only where residual branching remains, and summing over assignments to those sources. A checkpointed three-region partition with autotuned placement reuses shared prefix/root/suffix work; threshold pruning gives a fidelity/runtime tradeoff; a server-worker MPI/OpenMP schedule distributes output batches and histories. The authors validate against Qiskit Aer on a 12-qubit QFT, demonstrate checkpointing and autotuning on amplitude amplification, show a QAOA pruning sweep, reconstruct exact 100-qubit quantum-walk distributions, report 85% parallel efficiency at 1,024 MPI processes (8,192 cores), and compare favorably with quimb. The central complexity claim is O(S M 2^A G), where A is the number of artificial sources.

Significance. If the claims hold, the paper makes a useful contribution: exact sparse-output simulation whose exponential cost is governed by residual branching (A) rather than circuit width, an open-source MPI/OpenMP implementation, external validation against a state-vector simulator, and a threshold sweep that honestly exhibits a tradeoff rather than tuning a parameter to match a target. The parallel-efficiency result, if reproducible, is noteworthy. However, the practical value hinges on A being small for the target circuits, and that is currently supported by an unproved assertion rather than a demonstrated property. The empirical section also relies on single-run timings and a possibly unfair quimb comparison. The conceptual core appears sound; the performance story needs strengthening.

major comments (4)
  1. [§IV-B; Eq. (7); §IV-A complexity bound] The assertion 'For the benchmark families in this paper, backward propagation attains the optimum' is not proved; it is justified only by inspecting graphical representations. Since Eq. (7) and the O(SM2^A G) bound make the method's advantage depend exponentially on A, and the Discussion (Section VI) admits that source placement is central to performance, this is a load-bearing claim. If A can be as large as O(|W|), the method degenerates to full-width Feynman summation. Please provide a proof for each family (or a formal characterization of the minimum A), or a systematic exact computation of the optimum for the actual instances, rather than a heuristic justification.
  2. [§V, Figs. 7, 8, 10, 11, 12] All reported runtimes are single measurements without variance or confidence intervals. The speedups in the checkpointing ablation, the pruning sweep, the 93%/85% parallel efficiency values, and the 'more than one order of magnitude' advantage over quimb are point estimates. Performance measurements on shared HPC nodes need multiple repetitions with median/min and spread, plus a statement of the timing methodology. Without this, the empirical central claims are not quantitatively established.
  3. [§V, Fig. 12] The quimb comparison is not stated to be apples-to-apples: quimb runtime is plotted against the number of Qiskit transpiled operations, while the Feynman simulator appears to use the native circuit representation. The peaks in transpiled operations coincide with quimb timeouts, suggesting the comparison may include an artificially inflated operation count or unaccounted preprocessing. Please specify the circuit representation given to each tool, include transpilation/preprocessing time for both, or otherwise control for this factor before claiming a speedup.
  4. [§V, 'Quantum Walk' paragraph and Fig. 9] The text says the simulation targets '2,000 output bitstrings corresponding to 1,000 walker positions to the left and right of the initial position,' but a 16-step quantum walk has support of at most 33 positions, and Fig. 9's x-axis is -15..15. This is internally inconsistent and overstates the number of nonzero outputs actually computed. Please correct the description and state exactly how many requested bitstrings were nonzero and whether the 73 s run includes preprocessing (source placement and autotuning).
minor comments (7)
  1. [§IV-B, Fig. 3] The graph-simplification rule 'remove the source, remove the CNOT it is connected to, and merge the two nodes' needs a formal definition of 'reached' and a precise explanation of why the merged node captures the equivalence; currently the rule is intuitive but not fully defined.
  2. [Table I] Define E in the QAOA row (presumably the number of edges) and state the exact circuit parameterization; also define s, n, and the gate set in the caption.
  3. [§V, Fig. 6] The QFT validation is visual only. Report the maximum absolute difference between the Feynman and Qiskit amplitudes for the requested bins, or add a numeric error column.
  4. [§V, QAOA] The QAOA instance is not fully specified: graph size, number of edges, number of layers, and the exact definition of fidelity are missing. Provide these for reproducibility.
  5. [§IV-A, Observation 2] 'Most of the traditional quantum algorithms' is vague; name the specific circuit families or soften the claim.
  6. [§V, Fig. 5] The caption should state that the spread at fixed A is caused by exact pruning, threshold pruning, and checkpointing; the text says this, but a caption note would improve readability.
  7. [§II, Eq. (3)] Define m as the number of gates and clarify that i(t) denotes an intermediate computational-basis state; the notation can be confused with an index.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central derivations are explicit rewrites of the Feynman path sum, and the experimental claims are validated against external simulators without fitted parameters being relabeled as predictions.

full rationale

The paper's core derivation is a direct reformulation of the standard Feynman sum-over-histories amplitude: Eq. (7) rewrites the sum over all internal-wire assignments as a sum over artificial-source assignments after determinism propagation. This is a computational reorganization, not a definition of the target in terms of the output. The artificial-source count A is not fitted to reproduce reported amplitudes; it is a structural quantity determined by circuit analysis, and the experiments vary circuit families to expose different A regimes. Exactness is validated against the external Qiskit Aer simulator (Figure 6), and the large-scale quantum walk results are checked by probability normalization and compared against the external tensor-network simulator quimb. The autotuned checkpoint partition is selected by minimizing the paper's own cost model, Eq. (8), but the resulting speedups are measured directly in ablation experiments, not asserted from the model. The pruning threshold sweep in QAOA is presented as a tradeoff, not as a tuned parameter chosen to force a target fidelity. Self-citations [20]–[22] are background and implementation references and are not load-bearing for the central claim. The unproven assertion that backward propagation attains optimal artificial-source placement for the benchmark families is a performance-evidence gap, not circularity: the method remains exact and the claim is not used to derive the measured results from itself.

Assumptions & free parameters 2 free parameters · 5 assumptions · 1 invented entities

The method's correctness relies on the completeness of boundary propagation and the soundness of the history enumeration; its practical value relies on the unproven assumption that backward propagation yields few artificial sources for structured circuits. The autotuned checkpointing uses the paper's own cost model, so speedup claims are not independent predictions.

free parameters (2)
  • pruning threshold t = 0 in exact mode; up to 1e-4 in the QAOA sweep
    User-specified constant controlling early truncation of histories. It is not fitted to reproduce results but is varied to show a runtime-fidelity tradeoff.
  • checkpoint partition (prefix/root/suffix split) = chosen by autotuning for each circuit
    Selected to minimize the cost estimate in Eq. (8). It affects runtime but not correctness; it is an optimization parameter rather than a fitted constant.
assumptions (5)
  • standard math Deterministic gates are basis permutations and propagate wire values uniquely in both directions.
    Used for the core propagation rules in Section IV-A; follows directly from the definition of gates such as X, CNOT, and Toffoli.
  • standard math Wire-preserving gates leave the computational-basis value on each wire unchanged and contribute only phase or activation factors.
    Defines the wire-preserving class in Section IV-A; true for diagonal and controlled-phase gates.
  • domain assumption Every valid history for a fixed input-output pair is uniquely specified by an assignment to the artificial sources together with all propagated wire values.
    This is the central reduction, stated as Observations 1 and 2 in Section IV-A; the paper argues it but does not provide a formal completeness proof.
  • ad hoc to paper Backward propagation attains the minimum number of artificial sources for all benchmark circuit families.
    Asserted in Section IV-B and used to justify the backward-propagation choice throughout the evaluation; no proof or exhaustive search is given.
  • domain assumption The gate-operation count in Eq. (8) is a faithful proxy for runtime and guides the autotuned partition selection.
    The autotuner minimizes this estimate; the paper does not validate the proxy against measured runtimes beyond the ablation in Figure 7.
invented entities (1)
  • artificial sources
    purpose: Binary variables introduced at unresolved branching points to index the reduced history space.
    Algorithmic constructs only; they have no direct experimental observable, so independent evidence is absent.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Fewer Histories, Faster Paths: Distributed Quantum Circuit Feynman Simulation via History Reduction, Checkpointing, and Pruning." pith.science (2026). https://pith.science/paper/KMQLDVTK

@misc{pith2026260801467,
  author       = {Pith},
  title        = {Pith review of: Fewer Histories, Faster Paths: Distributed Quantum Circuit Feynman Simulation via History Reduction, Checkpointing, and Pruning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KMQLDVTK}},
  note         = {Machine review of arXiv:2608.01467}
}
read the original abstract

We present a distributed method for exact sparse-output quantum circuit simulation based on the pure Feynman sum-over-histories formulation. The method computes selected computational-basis amplitudes exactly and addresses the exponential growth of the path sum through a reduced history formulation based on internal-wire assignments, determinism propagation, artificial sources, pruning, and checkpointed reuse. Boundary constraints are propagated through deterministic and wire-preserving gates, and explicit branching variables are introduced only where residual ambiguity remains. Shared work across related histories is captured via an autotuned checkpointed partition. The parallel execution model combines decomposition over requested outputs with concurrent history evaluation, while a dynamic server-worker architecture mitigates load imbalance from irregular branching and pruning. Across the circuit families studied, the method adapts to different structural regimes of the reduced history space: zero artificial sources for QFT under backward analysis, substantial speedups from checkpointing and autotuning for amplitude amplification, and a runtime-fidelity tradeoff from threshold pruning for QAOA. On quantum walk circuits, it reconstructs exact selected-output distributions up to 100 qubits and achieves 85% parallel efficiency on 8,192 CPU cores of a supercomputer.

Figures

Figures reproduced from arXiv: 2608.01467 by the authors.

Figure 1
Figure 1. Three-qubit example in the internal-wire formulation. The Hadamard [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Internal-wire view of a four-qubit circuit. The figure shows a standard [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Top: Circuit we want to find optimal artificial source placement of. [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (7 more)
Figure 4
Figure 4. Figure 4: Implementation workflow and two-level parallel execution of the [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Time consumption versus number of artificial sources for quantum [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: Validation of sparse-output amplitudes for a 12-qubit QFT circuit. [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 7
Figure 7. Figure 7: Checkpointing ablation for amplitude amplification with 4 qubits [PITH_FULL_IMAGE:figures/full_fig_p009_7.png]
Figure 9
Figure 9. Figure 9: Exact sparse-output reconstruction for a 100-qubit quantum walk of 16 [PITH_FULL_IMAGE:figures/full_fig_p009_9.png]
Figure 11
Figure 11. Figure 11: Strong scaling test on AMD EPYC 7742 and Fujitsu A64FX CPUs [PITH_FULL_IMAGE:figures/full_fig_p010_11.png]
Figure 12
Figure 12. Figure 12: Left axis: Time consumption of our Feynman simulator and quimb. [PITH_FULL_IMAGE:figures/full_fig_p010_12.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

36 extracted references · 26 canonical work pages

  1. [1]

    M. A. Nielsen and I. L. Chuang,Quantum Computation and Quantum Information, 10th ed. Cambridge: Cambridge University Press, 2010

  2. [2]

    5 petabyte simulation of a 45-qubit quantum circuit,

    T. H ¨aner and D. S. Steiger, “5 petabyte simulation of a 45-qubit quantum circuit,” inProceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis, 2017, pp. 1–10

  3. [3]

    qHiPSTER: The quantum high performance software testing environment,

    V . N. Smelyanskiy, N. P. D. Sawaya, and A. Aspuru-Guzik, “qHiPSTER: The quantum high performance software testing environment,” 2016

  4. [4]

    QuEST and high performance simulation of quantum computers,

    T. Jones, A. Brown, I. Bush, and S. C. Benjamin, “QuEST and high performance simulation of quantum computers,” 2019

  5. [5]

    ProjectQ: An open source software framework for quantum computing,

    D. S. Steiger, T. H ¨aner, and M. Troyer, “ProjectQ: An open source software framework for quantum computing,”Quantum, vol. 2, p. 49, 2018

  6. [6]

    Simulating quantum computation by contract- ing tensor networks,

    I. L. Markov and Y . Shi, “Simulating quantum computation by contract- ing tensor networks,”SIAM Journal on Computing, vol. 38, no. 3, pp. 963–981, 2008

  7. [7]

    A practical introduction to tensor networks: Matrix product states and projected entangled pair states,

    R. Or ´us, “A practical introduction to tensor networks: Matrix product states and projected entangled pair states,”Annals of Physics, vol. 349, p. 117–158, Oct. 2014. [Online]. Available: http://dx.doi.org/10.1016/j.aop.2014.06.013

  8. [8]

    quimb: a python library for quantum information and many- body calculations,

    J. Gray, “quimb: a python library for quantum information and many- body calculations,”Journal of Open Source Software, vol. 3, no. 29, p. 819, 2018

Show all 36 references
  1. [9]

    Quantum supremacy using a programmable supercon- ducting processor,

    F. Aruteet al., “Quantum supremacy using a programmable supercon- ducting processor,”Nature, vol. 574, no. 7779, pp. 505–510, 2019

  2. [10]

    Quantum supremacy is both closer and farther than it appears

    I. L. Markov, A. Fatima, S. V . Isakov, and S. Boixo, “Quantum supremacy is both closer and farther than it appears.” [Online]. Available: http://arxiv.org/abs/1807.10749

  3. [11]

    Unbiased simulation of near-Clifford quantum circuits,

    R. S. Bennink, E. M. Ferragut, T. S. Humble, J. A. Laska, J. J. Nutaro, M. G. Pleszkoch, and R. C. Pooser, “Unbiased simulation of near-Clifford quantum circuits,”Physical Review A, vol. 95, no. 6, p. 062337, Jun. 2017. [Online]. Available: http://link.aps.org/doi/10.1103/Phys...

  4. [12]

    Simulation of quantum circuits by low-rank stabilizer decompositions,

    S. Bravyi, D. Browne, P. Calpin, E. Campbell, D. Gosset, and M. Howard, “Simulation of quantum circuits by low-rank stabilizer decompositions,”Quantum, vol. 3, p. 181, 2019

  5. [13]

    Simulation of qubit quantum circuits via Pauli propagation,

    P. Rall, D. Liang, J. Cook, and W. Kretschmer, “Simulation of qubit quantum circuits via Pauli propagation,”Physical Review A, vol. 99, no. 6, p. 062337, 2019

  6. [14]

    Pauli Propagation: A Computational Framework for Simulating Quantum Systems,

    M. S. Rudolph, T. Jones, Y . Teng, A. Angrisani, and Z. Holmes, “Pauli Propagation: A Computational Framework for Simulating Quantum Systems,” May 2025, arXiv:2505.21606 [quant-ph]. [Online]. Available: http://arxiv.org/abs/2505.21606

  7. [15]

    Simulating physics with computers,

    R. P. Feynman, “Simulating physics with computers,”International Journal of Theoretical Physics, vol. 21, no. 6–7, pp. 467–488, 1982

  8. [16]

    A sum-over-paths description of quantum circuits,

    B. Rudiak-Gould, “A sum-over-paths description of quantum circuits,” 2006

  9. [17]

    Feynman path-sum quantum computer simulator,

    D. A. C. Ferreira, “Feynman path-sum quantum computer simulator,” Master’s thesis, Universidade do Minho (Portugal), 2023

  10. [18]

    Simulation of low-depth quantum circuits as complex undirected graphical models,

    S. Boixo, S. V . Isakov, V . N. Smelyanskiy, and H. Neven, “Simulation of low-depth quantum circuits as complex undirected graphical models,” 2017

  11. [19]

    N. D. Mermin,Quantum Computer Science: An Introduction. Cam- bridge: Cambridge University Press, 2007

  12. [20]

    What is quantum parallelism, anyhow?

    S. Markidis, “What is quantum parallelism, anyhow?” 2024

  13. [21]

    Quantum computer simulations at warp speed: Assessing the impact of GPU acceleration: A case study with IBM Qiskit Aer, Nvidia Thrust & cuQuantum,

    J. Faj, I. B. Peng, J. Wahlgren, and S. Markidis, “Quantum computer simulations at warp speed: Assessing the impact of GPU acceleration: A case study with IBM Qiskit Aer, Nvidia Thrust & cuQuantum,” in2023 IEEE 19th International Conference on e-Science (e-Science), 2023, pp. 1–10

  14. [22]

    Enabling quantum computer simulations on AMD GPUs: a HIP backend for Google’s qsim,

    S. Markidis, “Enabling quantum computer simulations on AMD GPUs: a HIP backend for Google’s qsim,” inProceedings of the SC’23 Workshops of The International Conference on High Performance Computing, Network, Storage, and Analysis, 2023, pp. 1478–1486

  15. [23]

    A flexible high-performance simulator for the veri- fication and benchmarking of quantum circuits implemented on real hardware,

    B. Villalonga, S. Boixo, B. Nelson, C. Henze, E. G. Rieffel, R. Biswas, and S. Mandr `a, “A flexible high-performance simulator for the veri- fication and benchmarking of quantum circuits implemented on real hardware,”arXiv preprint arXiv:1811.09599, 2018

  16. [24]

    HybridQ: A hybrid simulator for quantum circuits,

    S. Mandr `a, J. Marshall, E. G. Rieffel, and R. Biswas, “HybridQ: A hybrid simulator for quantum circuits,” in2021 IEEE/ACM Second International Workshop on Quantum Computing Software (QCS), 2021, pp. 99–105

  17. [25]

    Hyper-optimized tensor network contraction,

    J. Gray and S. Kourtis, “Hyper-optimized tensor network contraction,” Quantum, vol. 5, p. 410, 2021

  18. [26]

    Hybrid Schr ¨odinger- Feynman simulation of quantum circuits with decision diagrams,

    L. Burgholzer, H. Bauer, and R. Wille, “Hybrid Schr ¨odinger- Feynman simulation of quantum circuits with decision diagrams,” in2021 IEEE International Conference on Quantum Computing and Engineering (QCE), 2021, pp. 199–206. [Online]. Available: http://arxiv.org/abs/2105.07045

  19. [27]

    Graph-based simulation of quantum computation in the density matrix representation,

    G. F. Viamontes, I. L. Markov, and J. P. Hayes, “Graph-based simulation of quantum computation in the density matrix representation,”Quantum Information & Computation, vol. 7, no. 1, pp. 1–41, 2007

  20. [28]

    A polynomial-time classical algorithm for noisy random circuit sampling,

    D. Aharonov, X. Gao, Z. Landau, Y . Liu, and U. Vazirani, “A polynomial-time classical algorithm for noisy random circuit sampling,” inProceedings of the 55th Annual ACM Symposium on Theory of Computing, 2023, pp. 945–957

  21. [29]

    Gropp, E

    W. Gropp, E. Lusk, and A. Skjellum,Using MPI: Portable Parallel Programming with the Message-Passing Interface, 2nd ed. MIT Press, 1999

  22. [30]

    Chapman, G

    B. Chapman, G. Jost, and R. van der Pas,Using OpenMP: Portable Shared Memory Parallel Programming. MIT Press, 2007

  23. [31]

    E. R. Johnston, N. Harrigan, and M. Gimeno-Segovia,Programming quantum computers: essential algorithms and code samples. O’Reilly Media, 2019

  24. [32]

    A Quantum Approximate Optimization Algorithm,

    E. Farhi, J. Goldstone, and S. Gutmann, “A Quantum Approximate Optimization Algorithm,” Nov. 2014, arXiv:1411.4028 [quant-ph]. [Online]. Available: http://arxiv.org/abs/1411.4028

  25. [33]

    Quantum random walks: an introductory overview,

    J. Kempe, “Quantum random walks: an introductory overview,”Con- temporary Physics, vol. 44, no. 4, pp. 307–327, 2003

  26. [34]

    Simulations of quantum circuits with approximate noise using qsim and cirq,

    S. V . Isakov, D. Kafri, O. Martin, C. V . Heidweiller, W. Mruczkiewicz, M. P. Harrigan, N. C. Rubin, R. Thomson, M. Broughton, K. Kissell, E. Peters, E. Gustafson, A. C. Y . Li, H. Lamm, G. Perdue, A. K. Ho, D. Strain, and S. Boixo, “Simulations of quantum circuits with appro...

  27. [35]

    Pulse-level noisy quantum circuits with QuTiP,

    B. Li, S. Ahmed, S. Saraogi, N. Lambert, F. Nori, A. Pitchford, and N. Shammah, “Pulse-level noisy quantum circuits with QuTiP,” Quantum, vol. 6, p. 630, Jan. 2022. [Online]. Available: http: //dx.doi.org/10.22331/q-2022-01-24-630

  28. [36]

    Simulating noisy quantum protocols with quantum trajectories,

    G. G. Carlo, G. Benenti, G. Casati, and C. Mej ´ıa-Monasterio, “Simulating noisy quantum protocols with quantum trajectories,” Phys. Rev. A, vol. 69, p. 062317, Jun 2004. [Online]. Available: https://link.aps.org/doi/10.1103/PhysRevA.69.062317

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.