Pith. sign in

REVIEW 4 major objections 5 minor 76 references

StreamingQEC's certified recurrence contracts repeated QEC scheduling transitions exactly, reproducing explicit-simulation metrics with zero delta and achieving 24.0x host speedup on a 16-job anchor.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-02 05:24 UTC pith:JMEQK32N

load-bearing objection A credible systems-level QEC pipeline simulator with a useful recurrence idea, but the formal signature's metric-delta state is undefined enough to undermine the exactness contract as written. the 4 major comments →

arxiv 2607.13351 v1 pith:JMEQK32N submitted 2026-07-15 quant-ph cs.PF

StreamingQEC: Streaming Quantum Error Correction in Tightly Integrated Quantum-Classical Systems via Certified Recurrence

Panayiotis Christou , Shuwen Kan , Hao Wang , Ying Mao This is my paper
classification quant-ph cs.PF PACS 03.67.Pp
keywords quantum error correctionstreaming QECsystem-level simulationdiscrete-event simulationcertified recurrencehybrid quantum-classical systemsresource contentiondesign-space exploration
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The paper is trying to establish that the classical workload of fault-tolerant quantum error correction—readout, syndrome transfer, decoding, feedback, and control—can be simulated at the system level with three interchangeable execution modes sharing one semantics. It claims an exact accelerator: a certified recurrence that contracts repeated QEC rounds only when the full scheduling state repeats, in which case it reproduces the explicit discrete-event simulation's metrics exactly. On 35 calibrated-profile configurations and a 16-job anchor workload, the compressed runs report zero metric deltas while running up to 24x faster on the host and scaling beyond 1.22 billion decode events. A separate approximate mode, auto staged-fluid, gives 2.60% mean makespan error over 17 references, making it suitable for screening. A sympathetic reader would care because the claim makes full-system architecture exploration—where decoder, link, or controller becomes the bottleneck—practicable before the hardware exists.

Core claim

On its own terms, the paper's central discovery is that repeated streaming-QEC transitions can be contracted without loss of fidelity: if two period starts have identical signatures—resource frontiers, queue/backlog state, decoder warm/cold state, deterministic grounding state, external reservation cursor, and metric deltas—then one grouped period applies the same state and metric delta as the explicit trace, by induction over certified periods. The proposition (scoped metric preservation) states this equivalence under four preconditions, the load-bearing one being that every scheduling input that can affect the next transition appears in the signature. The claimed consequence is that certif

What carries the argument

The key mechanism is the certification signature σ(S_k) = (F_k, Q_k, B_k, W_k, G_k, E_k, M_k)—a seven-field description of scheduling state: resource frontiers, queue/order state, QEC pending/backlog, decoder warm/cold, deterministic grounding-profile state, external reservation cursor, and metric-delta state. The recurrence algorithm uses it as a hash: when the current signature matches a previously observed boundary and the stable-delta check passes, it applies the stored period delta q times instead of executing each transition. The signature carries the correctness argument: because the explicit transition observes only these fields (precondition C3), identical signatures imply identical

Load-bearing premise

The load-bearing assumption is C3: every scheduling input that can affect the next transition appears in the seven-field signature—in other words, the explicit simulator has no hidden state beyond frontiers, queues, backlog, decoder warm/cold, grounding state, reservations, and metric deltas.

What would settle it

Construct a certified configuration and introduce a hidden tie-breaking counter in the explicit engine—one that influences which of two simultaneous events is scheduled first but is not part of the signature. If the explicit trace then diverges from the recurrent contraction, the completeness assumption fails; equally, a random sample of configurations outside the author-selected 35 that yields any nonzero metric delta between explicit and certified runs would refute the universal zero-delta claim.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • If the recurrence claim holds, exact system-level QEC simulation no longer scales with round count: compressed runs preserve all reported metrics while host time depends on the number of certified contractions, not the number of events.
  • Design-space exploration of decoder placement, link choice, and QEC cadence becomes tractable: the 219-row profile matrix and 128/256-job recurrent-only rows are presented as evidence that bottleneck regimes shift with code, decoder, distance, and transport.
  • The simulation indicates that fast matching decoders can be transfer-bound while heavier decoders saturate dedicated resources, and that microsecond-scale QEC cycles push all profiled paths toward saturation—so decoder and transport capacity must be provisioned jointly.
  • The same fitted timing profiles can be charged to different resource frontiers, letting architects compare shared versus dedicated placement before hardware exists.
  • The fluid approximation, with 2.60% mean and 6.45% worst makespan error on 17 references, is claimed as adequate for screening, with promotion to certified recurrence near decision boundaries.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The signed-state-contraction idea is not QEC-specific: any real-time pipeline with a finite, stable signature—control loops in robotics, network simulation, or HPC scheduling—could reuse the same certificate machinery, provided the completeness precondition can be established for that domain.
  • The paper's zero-delta claim is scoped to configurations satisfying C1–C4; a natural extension would be fuzz testing that perturbs engine-internal state not in the signature, to empirically probe whether the completeness assumption is actually sufficient.
  • Because the 35 calibrated-profile validations are author-selected rows, the strongest version of the exactness claim would require a random or adversarially sampled validation grid, or a formal proof that the seven fields are a complete state description.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper introduces StreamingQEC, a system-level discrete-event simulator for the classical workload induced by streaming QEC in tightly integrated quantum-classical systems. Three execution modes share a common staged model: explicit DES as the reference oracle, an approximate 'auto staged-fluid' mode for design screening, and a 'certified recurrence' mode that contracts repeated QEC periods only when a signature of scheduling/metric state matches an explicitly observed period. The central exactness claim is that certified recurrence preserves explicit-DES metrics exactly under preconditions C1–C4, supported by zero-delta parity on 35 calibrated-profile configurations, a 16-job anchor preserving 59,743,936 decode events, and additional workload/cadence sentinels. The fluid mode is reported to achieve 2.60% mean and 6.45% worst-case makespan error on 17 exact references. The paper also contributes a decoder-runtime corpus of 9,998 measured rows (8,174 retained), fitted service-time profiles, and design studies of bottleneck transitions, cadence pressure, and hardware counterfactuals.

Significance. If the certified-recurrence contract is made precise and validated independently, this is a valuable contribution to system-level QEC simulation: it attacks a real gap by modeling resource contention, queueing, and backpressure that circuit-level simulators omit, and it explicitly separates simulator correctness from timing-profile accuracy. The paper deserves credit for shipping an unusually detailed evidence ledger, for distinguishing simulated time from host wall time, for fail-closed behavior on unsupported configurations, and for grounding decoder service times in a substantial measurement campaign. The design insights (transfer-bound fast matching, decoder-bound heavier decoders, microsecond-cycle saturation) are plausible and follow from the presented experiments. However, the formal inconsistency in the recurrence signature definition (Eq. 12) undermines the central exactness claim as written, and the fluid-mode error is reported on the same reference set used to calibrate its correction constants. These issues are fixable but require substantive revision.

major comments (4)
  1. [§5.2, Eq. (12); §5.4 Algorithm 1] The signature σ(S_k) includes M_k as 'metric-delta state,' while §5.1 defines m_k as the accumulated metric vector. If M_k equals or contains m_k, then after any nonzero metric delta (e.g., the anchor's 2,065.85 s decode wait), σ(S_k) strictly changes, so the lookup in Algorithm 1 line 4 can never succeed. This would make the certified recurrence vacuous and contradict every reported zero-delta hit. If M_k is instead a stable per-period descriptor, it must be defined formally and proven constant under C1–C4. As written, the Proposition in Eq. (13) either has an unsatisfiable precondition or is circular, because C4's 'stable metric deltas' are not connected to the signature definition.
  2. [§5.3, precondition C3] C3 asserts that every scheduling input affecting the next transition appears in the seven signed fields of Eq. (12). The paper supports this only by enumeration and by zero-delta agreement with explicit DES from the same implementation. A shared implementation can hide missing state: event-queue ordering, profile-cache details, or native-engine bookkeeping could make both paths agree without semantic completeness. The induction needs a formal argument that the transition F reads only the signed fields, or an independent validation method (e.g., mutation testing, an independent reference engine, or fault-injection on candidate hidden state). Without that, the scoped exactness claim is only as strong as the completeness assumption the paper asserts but does not verify.
  3. [§8.3; Appendix D.1] The auto staged-fluid correction constants (θ=1.5, κ=6.4, λ=1.1, ω=0.10/0.28) appear in Appendix D.1, and the 2.60% mean / 6.45% worst error is measured on the 17 references listed in Table 11. The paper gives no evidence of a train/test split or held-out evaluation. As reported, the error is in-sample goodness of fit, not predictive accuracy for the stated 'screening' claim. Please either re-evaluate on held-out configurations or explicitly relabel these numbers as calibration error and provide an out-of-sample estimate before claiming the 10% screening threshold is met.
  4. [§8.2, Table 2] The zero-delta parity claim would be more convincing if the paper reported how often the certificate rejected a candidate or fell back to explicit execution in the validated runs. The current text reports only successful contractions. Rejection counts, fallback rates, and the fraction of rounds actually contracted would demonstrate that the certificate is non-trivially exercised and that the zero-delta result is not an artifact of a signature that only matches trivially repeated configurations.
minor comments (5)
  1. [Abstract] Typo: 'automatics taged-fluid' should be 'auto staged-fluid'.
  2. [Table 2] Column headers 'Total diff (s)' and 'Decode diff' are ambiguous; the zero values are clear only after reading the text. Add units or a footnote for each diff column.
  3. [§5.2] The term 'metric-delta state' is used without a formal definition. A one-sentence definition or a pointer to Appendix C.4 would remove the ambiguity discussed in the major comment.
  4. [§8.3] The phrase 'The model remains below the 10% operational screening threshold on the current reference set' should explicitly note that this is an in-sample statement if no validation split is added.
  5. [Figure 4] The y-axis lists repeating placement labels and error values; consider sorting by error and using distinct colors for placement to improve readability.

Circularity Check

2 steps flagged

Staged-fluid error is computed on the same references used to learn its correction terms, and the certified-recurrence signature leaves M_k undefined, making the central equivalence claim either vacuous or dependent on an unstated definition.

specific steps
  1. fitted input called prediction [Appendix D.1 (fluid model equations); Section 8.3 (E3)]
    "The final auto staged-fluid model adds regime information learned from explicit/certified telemetry. ... The current artifact uses θ=1.5, κ=6.4, and λ=1.1. ... By separating light shared-resource decode from heavy decoder-service regimes, auto staged-fluid reduces mean absolute makespan error to 2.60%, median error to 2.03%, and worst error to 6.45% across 17 references."

    The fluid correction parameters (θ, κ, λ, ω, regime predicates) are described as 'learned from explicit/certified telemetry,' i.e., from the same 17 exact references used to compute the reported 2.60% mean and 6.45% worst error. The headline approximation error is therefore a training-set residual, not an out-of-sample prediction; the reported accuracy is forced by the calibration rather than independently measured.

  2. self definitional [Section 5.1 (Eq. 11); Section 5.2 (Eq. 12); Section 5.3 (Proposition); Algorithm 1]
    "Let S_k be the complete scheduling state and m_k the accumulated metric vector at event boundary k. ... σ(S_k)= (F_k,Q_k,B_k,W_k,G_k,E_k,M_k), where F_k is the vector of resource frontiers, Q_k is the relevant queue/order state, B_k is QEC pending and backlog state, W_k is decoder warm/cold state, G_k is the deterministic grounding-profile state, E_k is the external reservation cursor, and M_k is metric-delta state."

    The recurrence proof requires repeated signatures σ(S_k)=σ(S_j), and then reuses the observed metric delta. But the only metric state defined is m_k, the accumulated metric vector. If M_k is that vector, any nonzero metric delta makes it strictly increase, so no signature can ever repeat and Proposition (13) is vacuous; the reported 59,743,936-event zero-delta rows would be impossible under the stated definition. Since M_k is never defined or proved periodic, the certificate's central precondition is either unsatisfiable or rests on an unstated redefinition, so the exactness claim reduces to an assumption about M_k.

full rationale

The clearest input-to-output reduction is in the aggressive staged-fluid mode: its correction terms are learned from explicit/certified telemetry and then evaluated on the same 17 references, so the reported 2.60%/6.45% error is an in-sample fit statistic, not an independent prediction. The certified-recurrence theorem is not itself circular—it is a deterministic-equivalence induction over identical signatures—but its formal contract is under-specified: M_k ('metric-delta state') is never defined separately from m_k, and under the literal reading the recurrence precondition is unsatisfiable, making the zero-delta validation vacuous or dependent on hidden state. There are no load-bearing self-citations; the decoder profiles are externally grounded and the recurrence correctness is checked against the explicit DES oracle. Because the fluid overfitting affects a secondary contribution and the recurrence gap is a definitional/formal weakness rather than a deliberate fit, the overall score is 5.

Axiom & Free-Parameter Ledger

5 free parameters · 4 axioms · 0 invented entities

The central recurrence result depends mainly on C1-C4 and the completeness of the signature; the fitted timing profiles and fluid constants are the main 'paid-for-upstream' numerical inputs. No new physical entities are introduced.

free parameters (5)
  • Auto staged-fluid correction constants theta, kappa, lambda, omega, w = theta=1.5, kappa=6.4, lambda=1.1, omega=0.10/0.28 CPU/GPU, w=0.5 or 0.25*gamma
    Chosen constants in Appendix D.1 shape Delta_chunk, Delta_l, Delta_h; no independent calibration split is reported, and they affect the 2.60%/6.45% fluid error claim.
  • Fitted decoder service-time surfaces (alpha, beta_d, beta_r, beta_b, beta_p) for 48 profiles = e.g., neural forward t=7.4372e-6 * b^0.5119 * d^0.9781 * r^0.5365
    Log-linear least-squares fits to the author-generated 8,174-row corpus (Appendix C.3); these supply every decode service time and therefore drive all design conclusions.
  • Link latency/bandwidth nominal parameters = Table 25, e.g., NVQLink host 3.84 us / 50 GB/s; Ethernet 30 us / 12.5 GB/s
    Modeled scenario parameters for the 219-row bottleneck matrix; no provider guarantee, acknowledged in Appendix G.2.
  • Payload bit densities b_s, b_f = not explicitly given
    Used to derive syndrome/feedback payloads when explicit byte counts are unavailable (Eqs. 7-8); assumed rather than measured.
  • Validation hyperparameters (backlog limit, circuit units per job) = at most two outstanding rounds; 5-10 materialized circuit units
    Chosen experiment constraints (Section 7.2) that bound the validation scope.
axioms (4)
  • domain assumption Five-stage canonical chain readout->syndrome transfer->decode->feedback transfer->control apply represents every protected logical interval.
    Section 3.3; StreamingQEC maps all gadgets to this chain; lattice surgery, feedforward steps, etc. are assumed to induce the same class of systems work.
  • domain assumption Stage durations are deterministic functions of signed state and event identity (C2).
    Section 5.3; required for the recurrence induction; real decoder runtimes and noise are stochastic, so this holds only for the simulator's fitted deterministic profiles.
  • domain assumption Decoder service-time profiles measured in isolation remain valid when placed on shared or contended frontiers.
    Sections 3.5 and 9; the simulator charges the same fitted service time to CPU/GPU/dedicated lanes without modeling interference; the paper admits unmeasured concurrent execution requires extra lanes or a detailed scheduler.
  • ad hoc to paper Log-linear power-law form is an adequate interpolation model for decoder runtime within the measured grid.
    Appendix C.3 Eq. 25; the fitted surface b_t(pi) is an empirical modeling choice; outside-grid uses are labeled extrapolative, but some design sweeps use distances above the measured grid.

pith-pipeline@v1.3.0-alltime-deepseek · 39096 in / 15387 out tokens · 163207 ms · 2026-08-02T05:24:48.721475+00:00 · methodology

0 comments
read the original abstract

Fault-tolerant quantum computing requires a continuous hybrid quantum error correction (QEC) pipeline comprising measurement readout, syndrome transport, decoding, feedback, and control. Existing QEC simulators primarily evaluate circuits, noise models, decoders, and protocol-level outcomes. System architects, however, must also understand how these workloads contend for and queue across controller, compute, accelerator, and communication resources during protected logical execution. We introduce StreamingQEC, a system-level simulator that translates fault-tolerant logical workloads into resource-constrained streaming-QEC pipelines. An explicit discrete-event simulation provides the reference execution semantics. An automatic staged-fluid mode enables faster approximate design-space exploration, while a certified recurrence mechanism compresses repeated transitions only when their scheduling state and metric contributions match those of the explicit execution trace. We assemble a decoder-runtime dataset containing 9,998 measurements, of which 8,174 are used to fit performance profiles. Recurrence reproduces the reported explicit-simulation metrics across 35 calibrated-profile configurations, as well as additional workload and cadence validation cases. For a 16-job anchor workload, it preserves 59,743,936 decoding events while achieving a 24.0x host-side speedup, and recurrent simulations scale beyond 1.22 billion events. Across 17 reference configurations, the automatics taged-fluid mode yields a mean makespan error of 2.60% and a worst-case error of 6.45%. Design-space studies reveal transfer-limited resource matching,decoder-driven pipeline stalls, and saturation of dedicated resources under microsecond-scale QEC cycles.

Figures

Figures reproduced from arXiv: 2607.13351 by Hao Wang, Panayiotis Christou, Shuwen Kan, Ying Mao.

Figure 1
Figure 1. Figure 1: StreamingQEC maps protected logical computation to grounded, resource-bound QEC stages evaluated [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Certified recurrence contracts a repeated QEC region only when its signed scheduling state and [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Host runtime for calibrated code/decoder profiles. Squares are explicit references, filled circles are [PITH_FULL_IMAGE:figures/full_fig_p015_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Auto staged-fluid makespan error over 17 exact references. All rows remain below the 10% screening [PITH_FULL_IMAGE:figures/full_fig_p016_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Representative accumulated QEC stage service on the mixed workload. Fast matching exposes [PITH_FULL_IMAGE:figures/full_fig_p017_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: QEC-cycle pressure under fixed workload and dedicated placement. Millisecond cycles retain headroom, [PITH_FULL_IMAGE:figures/full_fig_p018_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Bottleneck transitions in the 219-row profile matrix. Panel (a) shows dominant pressure by link, code, [PITH_FULL_IMAGE:figures/full_fig_p019_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Simulated completion-time reduction from independent hardware-axis counterfactuals on the 4-job [PITH_FULL_IMAGE:figures/full_fig_p019_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Resource-frontier representation of streaming QEC. An explicit staged trace is summarized by the [PITH_FULL_IMAGE:figures/full_fig_p025_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Resource-frontier recurrence for staged QEC rounds. The same per-stage reservation rule governs [PITH_FULL_IMAGE:figures/full_fig_p026_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Streaming-QEC workload coverage across four workload families. Recurrence-scaling evidence is [PITH_FULL_IMAGE:figures/full_fig_p029_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Base calibrated-profile validation speedup on the real mixed workload at [PITH_FULL_IMAGE:figures/full_fig_p031_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: Metric-mode ablation. Makespan-only recurrence can be useful for very large sweeps, but full-metric [PITH_FULL_IMAGE:figures/full_fig_p034_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: Supplementary distance-21 streaming-QEC cycle-pressure sweep on the mixed workload with [PITH_FULL_IMAGE:figures/full_fig_p040_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: Distribution of directly measured decoder service-time artifacts used by the grounding analysis. [PITH_FULL_IMAGE:figures/full_fig_p047_15.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

76 extracted references · 10 canonical work pages

  1. [1]

    Humble, Ryan Landfield, Ketan Maheshwari, Sarp Oral, Michael A

    Thomas Beck, Alessandro Baroni, Ryan Bennink, Gilles Buchs, Eduardo Antonio Coello Pérez, Markus Eisenbach, Rafael Ferreira da Silva, Muralikrishnan Gopalakrishnan Meena, Kalyan Gottiparthi, Peter Groszkowski, Travis S. Humble, Ryan Landfield, Ketan Maheshwari, Sarp Oral, Michael A. Sandoval, Amir Shehata, In-Saeng Suh, and Christopher Zimmer. 2024. Integ...

    work page doi:10.1016/j.futur 2024

  2. [2]

    Reinhardt, Ali Saidi, Arkaprava Basu, Joel Hestness, Derek R

    Nathan Binkert, Bradford Beckmann, Gabriel Black, Steven K. Reinhardt, Ali Saidi, Arkaprava Basu, Joel Hestness, Derek R. Hower, Tushar Krishna, Somayeh Sardashti, Rathijit Sen, Korey Sewell, Muhammad Shoaib, Nilay Vaish, Mark D. Hill, and David A. Wood. 2011. The gem5 Simulator.ACM SIGARCH Computer Architecture News39, 2 (2011), 1–7. doi:10.1145/2024716.2024718

    arXiv 2011

  3. [3]

    Bombin and M

    H. Bombin and M. A. Martin-Delgado. 2006. Topological Quantum Distillation.Physical Review Letters97, 18 (2006), 180501. doi:10.1103/PhysRevLett.97.180501

    doi:10.1103/physrevlett.97.180501 2006

  4. [4]

    Sergey Bravyi and Alexei Kitaev. 2005. Universal Quantum Computation with Ideal Clifford Gates and Noisy Ancillas. Physical Review A71, 2 (2005), 022316. doi:10.1103/PhysRevA.71.022316

    doi:10.1103/physreva.71.022316 2005

  5. [5]

    Brennan, João Barbosa, Chong Li, Meraj Ahmad, Fiheon Imroze, Calum Rose, Wridhdhisom Karar, Manoj Stanley, Hadi Heidari, Nick M

    Jack C. Brennan, João Barbosa, Chong Li, Meraj Ahmad, Fiheon Imroze, Calum Rose, Wridhdhisom Karar, Manoj Stanley, Hadi Heidari, Nick M. Ridler, and Martin Weides. 2025. Classical interfaces for controlling cryogenic quantum computing technologies.APL Quantum2, 4 (12 2025), 041501. doi:10.1063/5.0273490

    work page doi:10.1063/5.0273490 2025

  6. [6]

    Fall, Sally Floyd, John S

    Lee Breslau, Deborah Estrin, Kevin R. Fall, Sally Floyd, John S. Heidemann, Ahmed Helmy, Polly Huang, Steven McCanne, Kannan Varadhan, Ya Xu, and Haobo Yu. 2000. Advances in Network Simulation.Computer33, 5 (2000), 59–67. doi:10.1109/2.841785

    work page doi:10.1109/2.841785 2000

  7. [7]

    Randy Brown. 1988. Calendar queues: A fast O(1) priority queue implementation for the simulation event set problem. Commun. ACM31, 10 (1988), 1220–1227. doi:10.1145/63039.63045

    arXiv 1988

  8. [8]

    Robert Calderbank and Peter W

    A. Robert Calderbank and Peter W. Shor. 1996. Good quantum error-correcting codes exist.Physical Review A54, 2 (1996), 1098–1105. doi:10.1103/PhysRevA.54.1098

    doi:10.1103/physreva.54.1098 1996

  9. [9]

    Shane Caldwell, Moein Khazraee, Ben Howe, Muyuan Li, Justin Lietz, Tom Lubowe, and Alex McCaskey. 2025. NVIDIA NVQLink Architecture Integrates Accelerated Computing with Quantum Processors. https://developer.nvidia.com/blo g/nvidia-nvqlink-architecture-integrates-accelerated-computing-with-quantum-processors/. Accessed 2026

    2025

  10. [10]

    Shane A. Caldwell, Moein Khazraee, Elena Agostini, Tom Lassiter, Corey Simpson, Omri Kahalon, Mrudula Kanuri, Jin-Sung Kim, Sam Stanwyck, Muyuan Li, Jan Olle, Christopher Chamberland, Ben Howe, Bruno Schmitt, Justin G. Lietz, Alex McCaskey, Jun Ye, Ang Li, Alicia B. Magann, Corey I. Ostrove, Kenneth Rudinger, Robin Blume-Kohout, Kevin Young, Nathan E. Mil...

    arXiv 2025

  11. [11]

    Calheiros, Rajiv Ranjan, Anton Beloglazov, Cesar A

    Rodrigo N. Calheiros, Rajiv Ranjan, Anton Beloglazov, Cesar A. F. De Rose, and Rajkumar Buyya. 2011. CloudSim: A Toolkit for Modeling and Simulation of Cloud Computing Environments and Evaluation of Resource Provisioning Algorithms.Software: Practice and Experience41, 1 (2011), 23–50. doi:10.1002/spe.995

    doi:10.1002/spe.995 2011

  12. [12]

    Christopher Chamberland, Jan Olle, Muyuan Li, Scott Thornton, and Igor Baratta. 2026. Fast and accurate AI-based pre-decoders for surface codes. arXiv:2604.12841 [quant-ph] https://arxiv.org/abs/2604.12841

    Pith/arXiv arXiv 2026

  13. [13]

    Romero, Alejandro Gomez Cadavid, Anton Simen, Enrique Solano, and Narendra N

    Pranav Chandarana, Sebastián V. Romero, Alejandro Gomez Cadavid, Anton Simen, Enrique Solano, and Narendra N. Hegade. 2025. Hybrid Sequential Quantum Computing. arXiv:2510.05851 [quant-ph] https://arxiv.org/abs/2510.05851

    arXiv 2025

  14. [14]

    Yaojian Chen, Chun-Yang Luan, Peilin Zheng, Xianghong Zeng, Jia-Yi Hou, Zhuo Fu, Yirong Jin, Fei Wang, Guangwen Yang, and Dingshun Lv. 2026. Real-Time Quantum Error Correction System Stack: Architecture, Algorithms, and Engineering Practice. arXiv:2605.30765 [quant-ph] https://arxiv.org/abs/2605.30765 , Vol. 1, No. 1, Article . Publication date: July 2026...

    Pith/arXiv arXiv 2026

  15. [15]

    Tyler Christeson, Md Habib Ullah, Ali Arabnya, Amin Khodaei, and Rui Fan. 2025. Hybrid Quantum-Classical Optimization of the Resource Scheduling Problem. arXiv:2511.00733 [eess.SY] https://arxiv.org/abs/2511.00733

    arXiv 2025

  16. [17]

    Tim Coopmans, Robert Knegjens, Axel Dahlberg, David Maier, Loek Nijsten, Julio de Oliveira Filho, Martijn Papendrecht, Julian Rabbie, Filip Rozpędek, Matthew Skrzypczyk, Leon Wubben, Walter de Jong, Damian Podareanu, Ariana Torres- Knoop, David Elkouss, and Stephanie Wehner. 2021. NetSquid, a NETwork Simulator for QUantum Information using Discrete Events...

    doi:10.1038/s42005-021-00647-8 2021

  17. [18]

    Kalbarczyk, and Ravishankar K

    Shengkun Cui, Archit Patke, Hung Nguyen, Aditya Ranjan, Ziheng Chen, Phuong Cao, Gregory Bauer, Brett Bode, Catello Di Martino, Saurabh Jha, Chandra Narayanaswami, Daby Sow, Zbigniew T. Kalbarczyk, and Ravishankar K. Iyer. 2025. Story of Two GPUs: Characterizing the Resilience of Hopper H100 and Ampere A100 GPUs. InProceedings of the International Confere...

    arXiv 2025

  18. [19]

    Axel Dahlberg and Stephanie Wehner. 2018. SimulaQron: A simulator for developing quantum internet software. Quantum Science and Technology4, 1 (2018), 015001. doi:10.1088/2058-9565/aad56e

    work page doi:10.1088/2058-9565/aad56e 2018

  19. [20]

    David Tolnay. 2026. thiserror: derive macros for the standard Rust error trait. https://crates.io/crates/thiserror. Software documentation. Accessed 2026-07-10

    2026

  20. [21]

    McMahon, Duncan Roweth, and Torsten Hoefler

    Daniele De Sensi, Salvatore Di Girolamo, Kim H. McMahon, Duncan Roweth, and Torsten Hoefler. 2020. An In-Depth Analysis of the Slingshot Interconnect. InProceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis (SC). IEEE, Piscataway, NJ, USA, 1–14. doi:10.1109/SC41405.2020.00039

    Pith/arXiv arXiv 2020

  21. [22]

    Eric Dennis, Alexei Kitaev, Andrew Landahl, and John Preskill. 2002. Topological quantum memory.J. Math. Phys.43, 9 (2002), 4452–4505. doi:10.1063/1.1499754

    doi:10.1063/1.1499754 2002

  22. [23]

    Dmitry Duplyakin, Alexandru Uta, Aleksander Maricq, and Robert Ricci. 2020. In Datacenter Performance, The Only Constant Is Change. InIEEE/ACM International Symposium on Cluster, Cloud and Internet Computing (CCGrid). IEEE, Piscataway, NJ, USA, 370–379. doi:10.1109/CCGrid49817.2020.00-56

    arXiv 2020

  23. [24]

    Philip Döbler and Manpreet Singh Jattana. 2025. A Survey on Integrating Quantum Computers into High Performance Computing Systems. arXiv:2507.03540 [cs.ET] https://arxiv.org/abs/2507.03540

    Pith/arXiv arXiv 2025

  24. [25]

    Bryan Eastin and Emanuel Knill. 2009. Restrictions on Transversal Encoded Quantum Gate Sets.Physical Review Letters102, 11 (2009), 110502. doi:10.1103/PhysRevLett.102.110502

    doi:10.1103/physrevlett.102.110502 2009

  25. [26]

    Bo Fang, Daoce Wang, Sian Jin, Quincey Koziol, Zhao Zhang, Qiang Guan, Suren Byna, Sriram Krishnamoorthy, and Dingwen Tao. 2021. Characterizing Impacts of Storage Faults on HPC Applications: A Methodology and Insights. In IEEE International Conference on Cluster Computing (CLUSTER). IEEE, Piscataway, NJ, USA, 409–420. doi:10.1109/Clus ter48925.2021.00048

    arXiv 2021

  26. [27]

    Fowler, Matteo Mariantoni, John M

    Austin G. Fowler, Matteo Mariantoni, John M. Martinis, and Andrew N. Cleland. 2012. Surface codes: Towards practical large-scale quantum computation.Physical Review A86, 3 (2012), 032324. doi:10.1103/PhysRevA.86.032324

    doi:10.1103/physreva.86.032324 2012

  27. [28]

    Fujimoto

    Richard M. Fujimoto. 2000.Parallel and Distributed Simulation Systems. Wiley, New York, NY, USA

    2000

  28. [29]

    Craig Gidney. 2021. Stim: a fast stabilizer circuit simulator.Quantum5 (2021), 497. doi:10.22331/q-2021-07-06-497

    doi:10.22331/q-2021-07-06-497 2021

  29. [30]

    Google Quantum AI. 2023. Suppressing quantum errors by scaling a surface code logical qubit.Nature614 (2023), 676–681. doi:10.1038/s41586-022-05434-1

    doi:10.1038/s41586-022-05434-1 2023

  30. [31]

    1997.Stabilizer codes and quantum error correction

    Daniel Gottesman. 1997.Stabilizer codes and quantum error correction. Ph. D. Dissertation. California Institute of Technology. doi:10.7907/97V4-YE48

    work page doi:10.7907/97v4-ye48 1997

  31. [32]

    Harris, K

    Charles R. Harris, K. Jarrod Millman, Stéfan J. van der Walt, Ralf Gommers, Pauli Virtanen, David Cournapeau, Eric Wieser, Julian Taylor, Sebastian Berg, Nathaniel J. Smith, Robert Kern, Matti Picus, Stephan Hoyer, Marten H. van Kerkwijk, Matthew Brett, Allan Haldane, Jaime Fernández del Río, Mark Wiebe, Pearu Peterson, Pierre Gérard- Marchant, Kevin Shep...

  32. [33]

    Sascha Heußen, Don Winter, Manuel Rispler, and Markus Müller. 2024. Dynamical subset sampling of quantum error-correcting protocols.Physical Review Research6, 1 (2024), 013177. doi:10.1103/PhysRevResearch.6.013177

    work page doi:10.1103/physrevresearch.6.013177 2024

  33. [34]

    Oscar Higgott. 2022. PyMatching: A Python package for decoding quantum codes with minimum-weight perfect matching.ACM Transactions on Quantum Computing3, 3 (2022), 16:1–16:16. doi:10.1145/3505637

    doi:10.1145/3505637 2022

  34. [35]

    Oscar Higgott and Craig Gidney. 2025. Sparse Blossom: Correcting a Million Errors per Core Second with Minimum- Weight Matching.Quantum9 (2025), 1600. doi:10.22331/q-2025-01-20-1600

    doi:10.22331/q-2025-01-20-1600 2025

  35. [36]

    Jefferson

    David R. Jefferson. 1985. Virtual time.ACM Transactions on Programming Languages and Systems7, 3 (1985), 404–425. doi:10.1145/3916.3988 , Vol. 1, No. 1, Article . Publication date: July 2026. StreamingQEC: Streaming Quantum Error Correction in Tightly Integrated Quantum-Classical Systems via Certified Recurrence 23

    arXiv 1985

  36. [37]

    Jaehoon Kang. 2026. Edge-Inference Governors Need Memory-Clock State. arXiv:2606.16106 [cs.DC] https://arxiv.or g/abs/2606.16106

    Pith/arXiv arXiv 2026

  37. [38]

    Apostolos Kokolis, Michael Kuchnik, John Hoffman, Adithya Kumar, Parth Malani, Faye Ma, Zachary DeVito, Shubho Sengupta, Kalyan Saladi, and Carole-Jean Wu. 2025. Revisiting Reliability in Large-Scale Machine Learning Research Clusters. In2025 IEEE International Symposium on High Performance Computer Architecture (HPCA). IEEE, Piscataway, NJ, USA, 1259–127...

    arXiv 2025

  38. [39]

    Yaniv Kurman, Lior Ella, Ramon Szmuk, Oded Wertheim, Benedikt Dorschner, Sam Stanwyck, and Yonatan Cohen

  39. [40]

    Henderson

    Mathieu Lacage and Thomas R. Henderson. 2006. Yet Another Network Simulator. InProceedings of the 2006 Workshop on ns-2: The IP Network Simulator. Association for Computing Machinery, New York, NY, USA, 12–es. doi:10.1145/1190 455.1190467

    work page doi:10.1145/1190 2006

  40. [41]

    Tallent, and Kevin J

    Ang Li, Shuaiwen Leon Song, Jieyang Chen, Jiajia Li, Xu Liu, Nathan R. Tallent, and Kevin J. Barker. 2020. Evaluating Modern GPU Interconnect: PCIe, NVLink, NV-SLI, NVSwitch and GPUDirect.IEEE Transactions on Parallel and Distributed Systems31, 1 (2020), 94–110. doi:10.1109/TPDS.2019.2928289

    arXiv 2020

  41. [42]

    Santiago Lopez, Jonathan Andrade Plascencia, and Gabriel N. Perdue. 2024. qec_code_sim: An open-source Python framework for estimating the effectiveness of quantum-error correcting codes on superconducting qubits. arXiv:2402.06745 [quant-ph] https://arxiv.org/abs/2402.06745

    Pith/arXiv arXiv 2024

  42. [43]

    Waylon Luo, Cheng-Chang Lu, Tong Zhan, and Qiang Guan. 2025. A Simulation Framework for Workload Management in Hybrid Quantum-HPC Cloud System. InProceedings of the SC ’25 Workshops. Association for Computing Machinery, New York, NY, USA, 1851–1859. doi:10.1145/3731599.3767548

    arXiv 2025

  43. [44]

    National Energy Research Scientific Computing Center (NERSC). 2026. Perlmutter Architecture. https://docs.nersc.g ov/systems/perlmutter/architecture/. Official system documentation. Accessed 2026-07-11

    2026

  44. [45]

    Nguyen, Muhammad Usman, and Rajkumar Buyya

    Hoa T. Nguyen, Muhammad Usman, and Rajkumar Buyya. 2024. iQuantum: A toolkit for modeling and simulation of quantum computing environments.Software: Practice and Experience54 (2024), 1141–1171. doi:10.1002/spe.3331

    doi:10.1002/spe.3331 2024

  45. [46]

    Nguyen, Muhammad Usman, and Rajkumar Buyya

    Hoa T. Nguyen, Muhammad Usman, and Rajkumar Buyya. 2025. QSimPy: A learning-centric simulation framework for quantum cloud resource management. InQuantum Computing, Rajkumar Buyya and Sukhpal Singh Gill (Eds.). Elsevier, Amsterdam, The Netherlands, 165–183. doi:10.1016/B978-0-443-29096-1.00012-X

    work page doi:10.1016/b978-0-443-29096-1.00012-x 2025

  46. [47]

    NVIDIA. 2025. NVIDIA NVQLink. https://www.nvidia.com/en-us/solutions/quantum-computing/nvqlink/. Accessed 2026

    2025

  47. [48]

    NVIDIA Corporation. 2026. CUDA Programming Guide. https://docs.nvidia.com/cuda/cuda-programming-guide/. Accessed 2026-07-10

    2026

  48. [49]

    NVIDIA Corporation. 2026. NVIDIA Ising-Decoding. https://github.com/NVIDIA/Ising-Decoding. Public inference software and checkpoints. Accessed 2026-07-10

    2026

  49. [50]

    Adam Paszke, Sam Gross, Francisco Massa, Adam Lerer, James Bradbury, Gregory Chanan, Trevor Killeen, Zeming Lin, Natalia Gimelshein, Luca Antiga, Alban Desmaison, Andreas Kopf, Edward Yang, Zachary DeVito, Martin Raison, Alykhan Tejani, Sasank Chilamkurthy, Benoit Steiner, Lu Fang, Junjie Bai, and Soumith Chintala. 2019. PyTorch: An Imperative Style, High...

    2019

  50. [51]

    Alberto Peruzzo, Jarrod McClean, Peter Shadbolt, Man-Hong Yung, Xiao-Qi Zhou, Peter J Love, Alán Aspuru-Guzik, and Jeremy L O’Brien. 2014. A variational eigenvalue solver on a photonic quantum processor.Nature Communications 5, 1 (2014), 4213. doi:10.1038/ncomms5213

    doi:10.1038/ncomms5213 2014

  51. [52]

    John Preskill. 1998. Reliable quantum computers.Proceedings of the Royal Society of London. Series A454, 1969 (1998), 385–410. doi:10.1098/rspa.1998.0167

    arXiv 1998

  52. [53]

    PyO3 Project. 2026. PyO3: Rust bindings for Python. https://pyo3.rs/. Software documentation. Accessed 2026-07-10

    2026

  53. [54]

    Rodrigues, K

    Arun F. Rodrigues, K. Scott Hemmert, Brian W. Barrett, Chad D. Kersey, Ron A. Oldfield, Marlo Weston, Rolf Riesen, Jeanine Cook, Paul Rosenfeld, Elliott Cooper-Balis, and Bruce L. Jacob. 2011. The Structural Simulation Toolkit.ACM SIGMETRICS Performance Evaluation Review38, 4 (2011), 37–42. doi:10.1145/1964218.1964225

    arXiv 2011

  54. [55]

    Joschka Roffe. 2022. LDPC: Python tools for low density parity check codes. https://pypi.org/project/ldpc/. Software package. Accessed 2026

    2022

  55. [56]

    Romero, Alejandro Gomez Cadavid, Enrique Solano, and Narendra N

    Sebastián V. Romero, Alejandro Gomez Cadavid, Enrique Solano, and Narendra N. Hegade. 2025. Sequential Quantum Computing. arXiv:2506.20655 [quant-ph] https://arxiv.org/abs/2506.20655

    Pith/arXiv arXiv 2025

  56. [57]

    Ryosuke Satoh, Michal Hajdušek, Naphan Benchasattabuse, Shota Nagayama, Kentaro Teramoto, Takaaki Matsuo, Sara Ayman Metwalli, Poramet Pathumsoot, Takahiko Satoh, Shigeya Suzuki, and Rodney Van Meter. 2022. QuISP: A Quantum Internet Simulation Package. In2022 IEEE International Conference on Quantum Computing and Engineering (QCE). IEEE, Piscataway, NJ, U...

    arXiv 2022

  57. [58]

    Kenta Sawamura, Kensuke Araki, Naoki Maruyama, Renichiro Haba, and Masayuki Ohzeki. 2025. Quantum-classical hybrid algorithm using quantum annealing for multi-objective job shop scheduling. arXiv:2511.03257 [quant-ph] https://arxiv.org/abs/2511.03257

    arXiv 2025

  58. [59]

    Bianca Schroeder, Eduardo Pinheiro, and Wolf-Dietrich Weber. 2009. DRAM Errors in the Wild: A Large-Scale Field Study. InProceedings of the ACM SIGMETRICS International Conference on Measurement and Modeling of Computer Systems. Association for Computing Machinery, New York, NY, USA, 193–204. doi:10.1145/1555349.1555372

    arXiv 2009

  59. [60]

    Serde Project. 2026. Serde: a serialization framework for Rust. https://serde.rs/. Software documentation. Accessed 2026-07-10

    2026

  60. [61]

    Amir Shehata, Peter Groszkowski, Thomas Naughton, Muralikrishnan Gopalakrishnan Meena, Elaine Wong, Daniel Claudino, Rafael Ferreira da Silva, and Thomas Beck. 2026. Bridging paradigms: Designing for HPC-Quantum convergence.Future Generation Computer Systems174 (Jan. 2026), 107980. doi:10.1016/j.future.2025.107980

    arXiv 2026

  61. [62]

    Peter W. Shor. 1995. Scheme for reducing decoherence in quantum computer memory.Physical Review A52, 4 (1995), R2493–R2496. doi:10.1103/PhysRevA.52.R2493

    doi:10.1103/physreva.52.r2493 1995

  62. [63]

    Akshitha Sriraman, Sihang Liu, Sinan Gunbay, Shan Su, and Thomas F. Wenisch. 2016. Deconstructing the Tail at Scale Effect Across Network Protocols. Workshop on Duplicating, Deconstructing and Debunking (WDDD). https://arxiv.org/abs/1701.03100

    Pith/arXiv arXiv 2016

  63. [64]

    Andrew M. Steane. 1996. Error correcting codes in quantum theory.Physical Review Letters77, 5 (1996), 793–797. doi:10.1103/PhysRevLett.77.793

    doi:10.1103/physrevlett.77.793 1996

  64. [65]

    Aleksandra Świerkowska, Jannik Pflieger, Emmanouil Giortamis, and Pramod Bhatotia. 2026. ECCentric: An Empirical Analysis of Quantum Error Correction Codes.Proceedings of the ACM on Measurement and Analysis of Computing Systems10, 2 (2026), 37:1–37:33. doi:10.1145/3805635

    work page doi:10.1145/3805635 2026

  65. [66]

    Runzhou Tao, Hongzheng Zhu, Jason Nieh, Jianan Yao, and Ronghui Gu. 2025. Quantum Virtual Machines. In Proceedings of the 19th USENIX Symposium on Operating Systems Design and Implementation (OSDI ’25). USENIX Association, Berkeley, CA, USA, 411–428. https://www.usenix.org/conference/osdi25/presentation/tao

    2025

  66. [67]

    Barbara M. Terhal. 2015. Quantum error correction for quantum memories.Reviews of Modern Physics87, 2 (2015), 307–346. doi:10.1103/RevModPhys.87.307

    doi:10.1103/revmodphys.87.307 2015

  67. [68]

    Jean-Pierre Tillich and Gilles Zémor. 2014. Quantum LDPC codes with positive rate and minimum distance proportional to the square root of the blocklength.IEEE Transactions on Information Theory60, 2 (2014), 1193–1202. doi:10.1109/TIT. 2013.2292061

    arXiv 2014

  68. [69]

    David K. Tuckett. 2021. qecsim: A Python quantum error correction simulator. https://pypi.org/project/qecsim/. Software package. Accessed 2026

    2021

  69. [70]

    van Dijk, E

    J.P.G. van Dijk, E. Kawakami, R.N. Schouten, M. Veldhorst, L.M.K. Vandersypen, M. Babaie, E. Charbon, and F. Sebastiano. 2019. Impact of Classical Control Electronics on Qubit Fidelity.Phys. Rev. Appl.12, 4 (Oct. 2019), 044054. doi:10.1103/PhysRevApplied.12.044054

    work page doi:10.1103/physrevapplied.12.044054 2019

  70. [71]

    Oliphant, et al

    Pauli Virtanen, Ralf Gommers, Travis E. Oliphant, et al. 2020. SciPy 1.0: fundamental algorithms for scientific computing in Python.Nature Methods17 (2020), 261–272. doi:10.1038/s41592-019-0686-2

    doi:10.1038/s41592-019-0686-2 2020

  71. [72]

    Xiaoliang Wu, Alexander Kolar, Joaquin Chung, Dong Jin, Tian Zhong, Rajkumar Kettimuthu, and Martin Suchara

  72. [73]

    Yue Wu and Lin Zhong. 2023. Fusion Blossom: Fast MWPM Decoders for QEC. In2023 IEEE International Conference on Quantum Computing and Engineering (QCE). IEEE, Piscataway, NJ, USA, 928–938. doi:10.1109/QCE57702.2023.00107

    arXiv 2023

  73. [74]

    Bauer, Brett Bode, Ian T

    Zhu Zhu, Yu Sun, Dhatri Parakal, Bo Fang, Steven Farrell, Gregory H. Bauer, Brett Bode, Ian T. Foster, Michael E. Papka, William Gropp, Zhao Zhang, and Lishan Yang. 2025. Understanding the Landscape of Ampere GPU Memory Errors. arXiv:2508.03513 [cs.DC] https://arxiv.org/abs/2508.03513 , Vol. 1, No. 1, Article . Publication date: July 2026. StreamingQEC: S...

    Pith/arXiv arXiv 2025

  74. [2020]

    doi:10.1038/s41586-020-2649-2

    Array programming with NumPy.Nature585 (2020), 357–362. doi:10.1038/s41586-020-2649-2

    doi:10.1038/s41586-020-2649-2 2020

  75. [2021]

    doi:10.1088/2058-9565/ac22f6

    SeQUeNCe: A Customizable Discrete-Event Simulator of Quantum Networks.Quantum Science and Technology 6, 4 (2021), 045027. doi:10.1088/2058-9565/ac22f6

    doi:10.1088/2058-9565/ac22f6 2021

  76. [2025]

    doi:10.1109/TQE.2025.3608053

    Benchmarking the Ability of a Controller to Execute Quantum Error Corrected Non-Clifford Circuits.IEEE Transactions on Quantum Engineering6 (2025), 1–14. doi:10.1109/TQE.2025.3608053

    arXiv 2025