Pith. sign in

REVIEW 2 major objections 4 minor 38 references

Multi-Stage Mamba-Based Architecture for Fast and Scalable Superconducting Qubit Readout

T0 review · 2 major / 4 minor · reviewed 2026-07-14 · grok-4.5

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A multi-stage Mamba discriminator reads superconducting qubits with 0.911 geometric-mean fidelity, half the parameters of the prior best method, and still works at 500 ns.

desk verdict Solid engineering win: Mamba gives better fidelity, half the params, and 500 ns MCM-ready readout on real multiplexed data; the multi-stage detector is mostly garnish. read the letter →

arxiv 2607.11668 v1 pith:SKLRXLGV submitted 2026-07-13 quant-ph

Luca Otting , Xiaorang Guo , Emmanouil Giortamis , Benjamin Lienhard , Pramod Bhatotia , Martin Schulz This is my paper · ORCID
classification quant-ph
keywords qubitreadoutMambasuperconductingqubitsrelaxationerrormid-circuitmeasurementquantumcorrectionfrequencymultiplexing
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

Readout is the slowest and most error-prone step on superconducting quantum processors, and frequency multiplexing makes it worse by introducing crosstalk. Feed-forward networks that try to fix the problem grow too large with trace length and still leave relaxation errors uncorrected. This paper replaces those networks with a two-stage Mamba pipeline: a compact selective state-space model first classifies the I/Q trace, then a second identical model corrects relaxation events that flipped the qubit from 1 to 0 during measurement. On a five-qubit multiplexed chip the lightweight version reaches 0.906 geometric-mean fidelity with 49.6 % fewer parameters than the previous best method; the full version reaches 0.911. Both stay above 0.89 even when the trace is cut to 500 ns. Monte-Carlo surface-code simulations show that the shorter, more accurate readout can cut the logical error rate by as much as 26 %. The architecture therefore offers a practical route to mid-circuit measurement that is both scalable and fast enough for real-time error correction.

What carries the argument

The multi-stage Mamba discriminator: a linear-complexity selective state-space model that expands each I/Q sample, evolves a fixed-size hidden state, and pools for classification, followed by an identical Mamba detector trained on KNN pseudo-labels that reclassifies only those traces the first stage labelled ‘0’.

What would settle it

Retrain the detector on an independent set of traces whose relaxation labels have been verified by a different physical method (for example, simultaneous photon-number monitoring) and measure whether the net fidelity gain of the second stage remains positive.

Watch

Extended reading notes

Core claim

A multi-stage Mamba architecture that first classifies superconducting-qubit I/Q traces and then post-corrects relaxation errors achieves geometric-mean five-qubit fidelity 0.911 (lightweight 0.906) while using 49.6 % fewer parameters than the prior best method and remaining usable down to 500 ns traces, yielding up to a 26 % reduction in logical error rate for a distance-7 surface code.

Load-bearing premise

The claim rests on the assumption that the KNN-derived pseudo-labels correctly mark the true relaxation events and that the second-stage detector still helps when the main model’s error distribution changes.

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

2 major / 4 minor

Summary. The paper proposes a multi-stage Mamba-based discriminator for frequency-multiplexed superconducting qubit readout. A main Mamba classifier performs initial state discrimination on I/Q traces; an optional second-stage detector, trained on KNN-derived pseudo-labels, post-corrects relaxation errors. On a public five-qubit experimental dataset the lightweight single-stage model reaches geometric-mean fidelity F5Q = 0.906 (49.6 % fewer parameters than MF-RMF-EMF) while the full multi-stage model reaches 0.911; both remain usable at 500 ns traces (F5Q = 0.893) and yield up to 26 % lower logical error rate in Monte-Carlo surface-code simulations (d = 7).

Significance. Readout latency and fidelity remain primary bottlenecks for mid-circuit measurement and surface-code QEC. Replacing multiplicative FNN parameter scaling with Mamba’s linear-time selective SSM is a well-motivated architectural choice; the reported parameter counts, short-trace fidelities, and explicit QEC logical-error reductions constitute concrete, reproducible advances over HERQULES, KLiNQ and MF-RMF-EMF on the same public traces. The linear scaling argument that follows from localized feed-line crosstalk is also practically useful for resource-constrained FPGA/ASIC deployment.

major comments (2)
  1. Section III-A and Table I: the detector is trained exclusively on KNN pseudo-labels of mean-trace-value clusters. The paper itself documents substantial overlap with the main-model errors (Fig. 3) and near-indistinguishability of early relaxations (UMAP/LDA, Fig. 6). Consequently the net fidelity gain on the Mamba main model is only 0.1–0.2 % (Table I). Because the lightweight single-stage model already exceeds all prior art, the multi-stage claim is not load-bearing for the headline numbers; the manuscript should either quantify the detector’s contribution more carefully or present the single-stage Mamba as the primary result.
  2. Section IV-E / Fig. 9: the Monte-Carlo logical-error curves keep the measured readout fidelities fixed even under the “futuristic” noise model (10× lower gate errors, 3× longer coherence). While conservative, this choice makes the reported 26–38 % reductions dependent solely on the shorter readout window rather than on any joint improvement of readout and gate errors; a short sensitivity study that also scales the readout error would strengthen the QEC claim.
minor comments (4)
  1. Fig. 4 caption and diagram contain typos (“Relexation”, “predection”).
  2. Hyper-parameter choices (linear expansion dimension = 8, Focal-loss α/γ, number of Mamba blocks, KNN neighborhood size) are stated without ablation or search ranges; a short appendix table would aid reproducibility.
  3. The geometric-mean definition F5Q is introduced only in the caption of Fig. 7; it should appear in the main text of Section IV-B.
  4. Qubit 2 is excluded from the detector pipeline because of fabrication issues; this should be stated once in the experimental-setup subsection rather than only in the detector paragraph.

Circularity Check

1 steps flagged · score 1.0 of 10

No significant circularity: empirical fidelities measured on held-out experimental traces against external baselines; only minor non-load-bearing use of KNN pseudo-labels for an optional detector whose contribution is quantified as near-zero for the Mamba main model.

  1. other [Section III-A (Integrated Error Detector) and Table I]
    "we use the KNN outputs as pseudo-labels and train the detector to replicate these assignments. ... the detector provides the greatest benefit when the main classification model struggles ... As the main model improves, the detector’s relative contribution naturally decreases. ... Mamba 0.2% 0.1% 0.1% 0.1%"

    KNN pseudo-labels are generated from the same experimental distribution used for evaluation and then used to supervise the optional second-stage detector. This is a standard pseudo-labeling practice, not a definitional reduction of the reported F5Q (which is measured against prepared-state ground truth on held-out traces). The paper quantifies the net effect as negligible for its own Mamba models, so the step does not force any headline claim.

full rationale

This is an empirical ML/systems paper whose central claims (F5Q of 0.906/0.911, 49.6% parameter reduction vs MF-RMF-EMF, robustness to 500 ns traces, and simulated 26% logical-error reduction) are obtained by training and evaluating models on a fixed experimental 5-qubit multiplexed readout dataset with a 70-15-15 split, then comparing against independently published baselines (FNN, HERQULES, KLiNQ, MF-RMF-EMF) under identical conditions. No first-principles derivation, uniqueness theorem, or free-parameter fit is presented as a prediction of the reported fidelities; model size is counted directly from architecture, and QEC Monte-Carlo results simply plug the measured per-duration fidelities into a surface-code simulator. The sole soft point is the KNN-derived pseudo-labels used to train the optional relaxation detector (Sec. III-A), but the paper itself shows the detector yields only 0.1–0.2% relative gain on the Mamba main model (Table I), the lightweight single-stage model already beats the prior SOTA, and early relaxations remain nearly inseparable (Fig. 6). That step is therefore not load-bearing for any headline number and does not reduce any claimed quantity to its own input by construction. Score 1 reflects this single minor, non-circular training detail rather than any self-definitional loop.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The central empirical claims rest on standard ML training assumptions, the experimental data set of Lienhard et al., and a small number of architectural and labeling choices that are not derived from first principles. No new physical entities are postulated; free parameters are ordinary neural-network hyper-parameters and the Focal-loss coefficients.

free parameters (4)
  • Focal-loss alpha and gamma
    Chosen by search to emphasize hard-to-classify relaxation traces; values not reported but affect detector performance.
  • Linear expansion dimension (8)
    Hand-chosen input expansion before the Mamba block; directly sets model capacity.
  • Number of Mamba blocks and hidden-state size
    Architectural hyper-parameters that determine the 769-parameter lightweight footprint.
  • KNN neighborhood size for pseudo-label generation
    Controls which traces are labeled as relaxation events and therefore trains the detector.
assumptions (4)
  • domain assumption I/Q traces after 50 ns binning retain sufficient information for state discrimination
    Standard preprocessing step taken from prior readout literature; not re-validated here.
  • ad hoc to paper KNN clustering on mean-trace values yields reliable pseudo-labels for relaxation events
    Section III-A; the detector is trained exclusively on these labels.
  • domain assumption Crosstalk is sufficiently local that a small fixed set of helper traces captures the dominant errors
    Justified by the measured conditional-error matrix (Fig. 5) and used to keep model size linear.
  • domain assumption Mamba’s selective SSM can focus on or ignore individual time bins analogously to matched-filter weights
    Motivates replacing FNNs; taken from the original Mamba paper.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Multi-Stage Mamba-Based Architecture for Fast and Scalable Superconducting Qubit Readout." pith.science (2026). https://pith.science/paper/SKLRXLGV

@misc{pith2026260711668,
  author       = {Pith},
  title        = {Pith review of: Multi-Stage Mamba-Based Architecture for Fast and Scalable Superconducting Qubit Readout},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SKLRXLGV}},
  note         = {Machine review of arXiv:2607.11668}
}
abstract

Reliable qubit readout is a critical bottleneck toward fault-tolerant quantum computing (FTQC). In superconducting quantum processors, readout operations are both error-prone and high-latency. These challenges become more severe in frequency-multiplexed architectures, where signal crosstalk among neighboring qubits significantly degrades readout fidelity. Existing machine learning (ML)-based approaches rely on feed-forward neural networks (FNNs) that suffer from large parameter sizes and lack an end-to-end network that jointly addresses relaxation errors and discriminates qubit states. In this work, we present a multi-stage qubit state discriminator based on the Mamba model, which enables efficient sequence modeling with linear complexity. The first stage performs initial state discrimination, followed by a refinement stage that identifies and mitigates relaxation-induced errors. Our lightweight model achieves a geometric mean readout fidelity of 0.906, outperforming the best-reported state-of-the-art method while reducing parameter size by 49.6%; our optimal model further reaches 0.911. Both models remain robust across varying input trace lengths, maintaining a high fidelity of 0.893 at readout durations as short as 500 $ns$, achieving up to a 26% reduction in logical error rate over prior work in quantum error correction (QEC).

Figures

Figures reproduced from arXiv: 2607.11668 by the authors.

Figure 1
Figure 1. Readout pipeline of superconducting qubits. (a) On the QPUs side, [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 3
Figure 3. Comparison between the main model’s predicted trace clusters [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figure 4
Figure 4. Architecture of the proposed multi-stage discriminator. The multi [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figures from the paper (4 more)
Figure 5
Figure 5. Figure 5: The heatmap illustrates the difference in error probability on a target [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: The UMAP clustering (left side) and the LDA results with a random [PITH_FULL_IMAGE:figures/full_fig_p006_6.png]
Figure 8
Figure 8. Figure 8: The cumulative readout accuracy (geometric mean) for different [PITH_FULL_IMAGE:figures/full_fig_p007_8.png]
Figure 9
Figure 9. Figure 9: Simulated logical error rates of our proposed approach compared to MF-RMF-EMF, HERQULES, and KLiNQ (results available up to 500 [PITH_FULL_IMAGE:figures/full_fig_p009_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

38 extracted references · 1 canonical work pages

  1. [1]

    Quantum computing and chemistry,

    J. D. Weidman, M. Sajjan, C. Mikolas, Z. J. Stewart, J. Pollanen, S. Kais, and A. K. Wilson, “Quantum computing and chemistry,”Cell Reports Physical Science, vol. 5, no. 9, p. 102105, 2024. [Online]. Available: https://www.sciencedirect.com/science/article/pii/S2666386424003837

    2024

  2. [2]

    Efficient combinatorial opti- mization by quantum-inspired parallel annealing in analogue memristor crossbar,

    M. Jiang, K. Shan, C. He, and C. Li, “Efficient combinatorial opti- mization by quantum-inspired parallel annealing in analogue memristor crossbar,”Nature Communications, vol. 14, 09 2023

    2023

  3. [3]

    Integration of quantum accelerators with high performance computing—a review of quantum programming tools,

    A. Elsharkawy, X.-T. M. To, P. Seitz, Y . Chen, Y . Stade, M. Geiger, Q. Huang, X. Guo, M. A. Ansari, C. B. Mendl, D. Kranzlm ¨uller, and M. Schulz, “Integration of quantum accelerators with high performance computing—a review of quantum programming tools,” ACM Transactions on Quantum Computing, vol. 6, no. 3, Jul. 2025. [Online]. Available: https://doi.o...

    doi:10.1145/3743149 2025

  4. [4]

    Superconducting quantum computers: who is leading the future?

    M. AbuGhanem, “Superconducting quantum computers: who is leading the future?”EPJ Quantum Technology, vol. 12, no. 1, p. 102, 2025

    2025

  5. [5]

    Realization of an error-correcting surface code with superconducting qubits,

    Y . Zhao, Y . Ye, H.-L. Huang, Y . Zhang, D. Wu, H. Guan, Q. Zhu, Z. Wei, T. He, S. Caoet al., “Realization of an error-correcting surface code with superconducting qubits,”Physical Review Letters, vol. 129, no. 3, p. 030501, 2022

    2022

  6. [6]

    Astrea: Accurate quantum error-decoding via practical minimum-weight perfect-matching,

    S. Vittal, P. Das, and M. Qureshi, “Astrea: Accurate quantum error-decoding via practical minimum-weight perfect-matching,” in Proceedings of the 50th Annual International Symposium on Computer Architecture, ser. ISCA ’23. New York, NY , USA: Association for Computing Machinery, 2023. [Online]. Available: https://doi.org/10. 1145/3579371.3589037

    arXiv 2023

  7. [7]

    Efficient and scalable architectures for multi-level superconducting qubit readout,

    C. N. Mude, S. Maurya, B. Lienhard, and S. Tannu, “Efficient and scalable architectures for multi-level superconducting qubit readout,” in 2025 62nd ACM/IEEE Design Automation Conference (DAC). San Francisco, CA, USA: IEEE, 2025, pp. 1–7

    2025

  8. [8]

    Deep-neural- network discrimination of multiplexed superconducting-qubit states,

    B. Lienhard, A. Veps ¨al¨ainen, L. C. Govia, C. R. Hoffer, J. Y . Qiu, D. Rist `e, M. Ware, D. Kim, R. Winik, A. Melville, B. Niedzielski, J. Yoder, G. J. Ribeill, T. A. Ohki, H. K. Krovi, T. P. Orlando, S. Gustavsson, and W. D. Oliver, “Deep-neural- network discrimination of multiplexed superconducting-qubit states,” Phys. Rev. Appl., vol. 17, p. 014024,...

    doi:10.1103/physrevapplied.17.014024 2022

Show all 38 references
  1. [9]

    Ibm quantum cloud compute resources,

    “Ibm quantum cloud compute resources,” https://quantum.ibm.com/ services/resources, accessed: 2025-20-08

    2025

  2. [10]

    Characterizing midcircuit measurements on a superconducting qubit using gate set tomography,

    K. Rudinger, G. J. Ribeill, L. C. Govia, M. Ware, E. Nielsen, K. Young, T. A. Ohki, R. Blume-Kohout, and T. Proctor, “Characterizing midcircuit measurements on a superconducting qubit using gate set tomography,” Physical Review Applied, vol. 17, no. 1, p. 014014, 2022

    2022

  3. [11]

    Decoherence benchmarking of superconducting qubits,

    J. J. Burnett, A. Bengtsson, M. Scigliuzzo, D. Niepce, M. Kudra, P. Dels- ing, and J. Bylander, “Decoherence benchmarking of superconducting qubits,”npj Quantum Information, vol. 5, no. 1, p. 54, 2019

    2019

  4. [12]

    Multiplexed dispersive readout of superconducting phase qubits,

    Y . Chen, D. Sank, P. O’Malley, T. White, R. Barends, B. Chiaro, J. Kelly, E. Lucero, M. Mariantoni, A. Megrantet al., “Multiplexed dispersive readout of superconducting phase qubits,”Applied Physics Letters, vol. 101, no. 18, 2012

    2012

  5. [13]

    Multiplexed readout of superconducting qubits using a three-dimensional reentrant- cavity filter,

    M. Bakr, S. D. Fasciati, S. Cao, G. Campanaro, J. Wills, M. Alghadeer, M. Piscitelli, B. Shteynas, V . Chidambaram, and P. J. Leek, “Multiplexed readout of superconducting qubits using a three-dimensional reentrant- cavity filter,”Phys. Rev. Appl., vol. 23, p. 054089, May 2025...

    2025 doi

  6. [14]

    Tomography via correlation of noisy measurement records,

    C. A. Ryan, B. R. Johnson, J. M. Gambetta, J. M. Chow, M. P. da Silva, O. E. Dial, and T. A. Ohki, “Tomography via correlation of noisy measurement records,”Phys. Rev. A, vol. 91, p. 022118, Feb

  7. [15]

    Available: https://link.aps.org/doi/10.1103/PhysRevA

    [Online]. Available: https://link.aps.org/doi/10.1103/PhysRevA. 91.022118

    doi

  8. [16]

    Machine learning for discriminating quantum measurement trajectories and improving readout,

    E. Magesan, J. M. Gambetta, A. D. C ´orcoles, and J. M. Chow, “Machine learning for discriminating quantum measurement trajectories and improving readout,”Phys. Rev. Lett., vol. 114, p. 200501, May

  9. [17]

    Available: https://link.aps.org/doi/10.1103/PhysRevLett

    [Online]. Available: https://link.aps.org/doi/10.1103/PhysRevLett. 114.200501

    doi

  10. [18]

    Low-latency machine learning fpga accelerator for multi-qubit-state discrimination,

    P. K. Gautam, S. Kalipatnapu, S. H, U. Singhal, B. Lienhard, V . Singh, and C. S. Thakur, “Low-latency machine learning fpga accelerator for multi-qubit-state discrimination,” 2024. [Online]. Available: https://arxiv.org/abs/2407.03852

    2024 arXiv

  11. [19]

    Qubit-state discrimination using neural networks with rapid and energy-efficient compute arrays,

    Y . Liu, Y . S. Chong, B. Lienhard, M. Fan, W. L. Goh, V . P. Nambiar, and A. T. Do, “Qubit-state discrimination using neural networks with rapid and energy-efficient compute arrays,” in2025 IEEE International Sym- posium on Circuits and Systems (ISCAS). London, United Kingdom...

    2025

  12. [20]

    Scaling qubit readout with hardware efficient machine learning architectures,

    S. Maurya, C. N. Mude, W. D. Oliver, B. Lienhard, and S. Tannu, “Scaling qubit readout with hardware efficient machine learning architectures,” inProceedings of the 50th Annual International Symposium on Computer Architecture, ser. ISCA ’23. New York, NY , USA: Association for...

    2023 doi

  13. [21]

    Qubicml: Ml-powered real-time quantum state discrimination enabling mid-circuit measurements,

    N. R. V ora, Y . Xu, A. Hasim, N. Fruitwala, N. Nguyen, H. Liao, J. Balewski, A. Rajagopala, K. Nowrouzi, Q. Ji, K. B. Whaley, I. Siddiqi, P. Nguyen, and G. Huang, “Qubicml: Ml-powered real-time quantum state discrimination enabling mid-circuit measurements,” in2024 IEEE Inter...

    2024

  14. [22]

    Klinq: Knowledge distillation-assisted lightweight neural network for qubit readout on fpga,

    X. Guo, T. Bunarjyan, D. Liu, B. Lienhard, and M. Schulz, “Klinq: Knowledge distillation-assisted lightweight neural network for qubit readout on fpga,” in2025 62nd ACM/IEEE Design Automation Con- ference (DAC). San Francisco, CA, USA: IEEE, 2025, pp. 1–7

    2025

  15. [23]

    Hardware-efficient neural networks for low-latency multiplexed superconducting qubit readout,

    X. Guo and M. Schulz, “Hardware-efficient neural networks for low-latency multiplexed superconducting qubit readout,” inSE2026. Gesellschaft f ¨ur Informatik, Bonn, 2026, pp. 10–18 420

    2026

  16. [24]

    Mamba: Linear-time sequence modeling with selective state spaces,

    A. Gu and T. Dao, “Mamba: Linear-time sequence modeling with selective state spaces,” inFirst conference on language modeling. Philadelphia, Pennsylvania, USA: OpenReview, 2024

    2024

  17. [25]

    Lightmamba: Efficient mamba acceleration on fpga with quantization and hardware co-design,

    R. Wei, S. Xu, L. Zhong, Z. Yang, Q. Guo, Y . Wang, R. Wang, and M. Li, “Lightmamba: Efficient mamba acceleration on fpga with quantization and hardware co-design,” in2025 Design, Automation & Test in Europe Conference (DATE). Lyon, France: IEEE, 2025, pp. 1–7

    2025

  18. [26]

    Frequency division multiplexing readout and simultaneous manipulation of an array of flux qubits,

    M. Jerger, S. Poletto, P. Macha, U. H ¨ubner, E. Il’ichev, and A. V . Ustinov, “Frequency division multiplexing readout and simultaneous manipulation of an array of flux qubits,”Applied Physics Letters, vol. 101, no. 4, 2012

    2012

  19. [27]

    Benchmarking the quality of multiplexed qubit readout beyond assignment fidelity,

    A. Di Giovanni, A. S. Aasen, J. Lisenfeld, M. G ¨arttner, H. Rotzinger, and A. V . Ustinov, “Benchmarking the quality of multiplexed qubit readout beyond assignment fidelity,”Physical Review Applied, vol. 24, no. 4, Oct. 2025. [Online]. Available: http://dx.doi.org/10.1103/dpft-lxtx

    2025 doi

  20. [28]

    Attention is all you need,

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, L. u. Kaiser, and I. Polosukhin, “Attention is all you need,” in Advances in Neural Information Processing Systems, I. Guyon, U. V . Luxburg, S. Bengio, H. Wallach, R. Fergus, S. Vishwanathan, and R. Garne...

    2017

  21. [29]

    Marca: Mamba accelerator with reconfigurable architecture,

    J. Li, S. Huang, J. Xu, J. Liu, L. Ding, N. Xu, and G. Dai, “Marca: Mamba accelerator with reconfigurable architecture,” inProceedings of the 43rd IEEE/ACM International Conference on Computer-Aided Design. New York,NY , USA: ACM, 2024, pp. 1–9

    2024

  22. [30]

    Surface codes: Towards practical large-scale quantum computation,

    A. G. Fowler, M. Mariantoni, J. M. Martinis, and A. N. Cleland, “Surface codes: Towards practical large-scale quantum computation,” Phys. Rev. A, vol. 86, p. 032324, Sep 2012. [Online]. Available: https://link.aps.org/doi/10.1103/PhysRevA.86.032324

    2012 doi

  23. [31]

    How to factor 2048 bit rsa integers in 8 hours using 20 million noisy qubits,

    C. Gidney and M. Eker ˚a, “How to factor 2048 bit rsa integers in 8 hours using 20 million noisy qubits,”Quantum, vol. 5, 2021. [Online]. Available: http://dx.doi.org/10.22331/q-2021-04-15-433

    doi

  24. [32]

    Fault-tolerant quantum computation,

    J. Preskill, “Fault-tolerant quantum computation,” 1997. [Online]. Available: https://arxiv.org/abs/quant-ph/9712048

    1997 arXiv

  25. [33]

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

    2010

  26. [34]

    Quantum error correction: an introductory guide,

    J. Roffe, “Quantum error correction: an introductory guide,” Contemporary Physics, vol. 60, no. 3, p. 226–245, Jul. 2019. [Online]. Available: http://dx.doi.org/10.1080/00107514.2019.1667078

    2019 doi

  27. [35]

    Real-time decoding for fault-tolerant quantum computing: progress, challenges and outlook,

    F. Battistel, C. Chamberland, K. Johar, R. W. J. Overwater, F. Sebastiano, L. Skoric, Y . Ueno, and M. Usman, “Real-time decoding for fault-tolerant quantum computing: progress, challenges and outlook,”Nano Futures, vol. 7, no. 3, p. 032003, Aug. 2023. [Online]. Available: htt...

    2023 doi

  28. [36]

    Characterising the failure mechanisms of error-corrected quantum logic gates,

    R. Harper, C. Lain ´e, E. Hockings, C. McLauchlan, G. M. Nixon, B. J. Brown, and S. D. Bartlett, “Characterising the failure mechanisms of error-corrected quantum logic gates,” 2025. [Online]. Available: https://arxiv.org/abs/2504.07258

    2025 arXiv

  29. [37]

    Focal loss for dense object detection,

    T.-Y . Lin, P. Goyal, R. Girshick, K. He, and P. Doll ´ar, “Focal loss for dense object detection,” 2018. [Online]. Available: https: //arxiv.org/abs/1708.02002

    2018 arXiv

  30. [38]

    Nvidia h100 tensor core gpu, extraordinary performance, scalability, and security for every data center

    Nvidia, “Nvidia h100 tensor core gpu, extraordinary performance, scalability, and security for every data center.” 2024.9. [Online]. Available: https://www.nvidia.com/en-us/data-center/h100/

    2024

Pith tools

Reviewed July 14, 2026 · model on record in the stance chip above.