Pith. sign in

REVIEW 4 major objections 6 minor 36 references

Performance Analysis of QAOA Across Distributed Quantum Network Topologies Using SwitchQNet

T0 review · 4 major / 6 minor · reviewed 2026-07-30 · grok-4.5

Pith's one-line read QAOA’s graph-defined communication pattern gets modest but consistent latency cuts when entanglement-aware scheduling runs on switch-based quantum data-center networks.

desk verdict Solid incremental SwitchQNet extension: new QAOA routing measurements with modest ~1.4–2.2× latency gains, useful as a diagnostic sparse-graph workload if claims stay scoped to the simulator assumptions. read the letter →

arxiv 2607.23407 v1 pith:YCSISX6Z submitted 2026-07-26 quant-ph

classification quant-ph
keywords QuantumNetworkingDistributedComputingApproximateOptimizationAlgorithmDataCentersEntanglementSchedulingQAOASwitchNetworks
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 asks how a structured optimization algorithm behaves when its qubits are spread across a quantum data center whose processors share entanglement through optical switches. It turns each edge of a MaxCut-style QAOA problem graph into remote two-qubit communication requests whenever the two qubits sit on different processors, then runs those requests through a compiler that plans entanglement generation, buffering, and switch reconfiguration. Across Clos, fat-tree, and spine-leaf layouts, that network-aware schedule reduces QAOA communication latency by roughly one-and-a-half to a bit over two times relative to a standard on-demand baseline—smaller gains than for denser circuits, but steady as racks, qubits, buffers, look-ahead depth, and entanglement delays change. A sympathetic reader cares because the result shows that sparse, graph-shaped workloads stress quantum networks differently from all-to-all benchmarks, and that those differences can already be measured in simulation. QAOA therefore works both as a practical distributed workload and as a diagnostic probe for co-designing algorithms, topologies, and entanglement management.

What carries the argument

A QAOA routing generator: each problem-graph edge whose endpoints map to different processors becomes two remote-CX requests (from the usual CX–Rz–CX cost decomposition), which then drive existing look-ahead EPR scheduling, in-rack collective generation, cross-rack splitting, distillation, and retry logic.

What would settle it

Re-run the same multi-topology QAOA cases with substantially different measured in-rack and cross-rack entanglement times, or with multi-layer circuits that include local compute, and test whether the reported 1.4×–2.2× latency advantage over the baseline shrinks, vanishes, or reverses.

Watch

Extended reading notes

Core claim

When QAOA cost edges are compiled into remote-CX requests and executed under look-ahead entanglement scheduling on switch-based quantum data-center fabrics, end-to-end communication latency falls by about 1.36× to 2.17× versus a buffer-assisted on-demand baseline across Clos, fat-tree, and spine-leaf configurations. The improvement is consistent but smaller than for communication-heavy programs because QAOA’s lower interaction density limits collective entanglement opportunities; cross-rack entanglement still dominates delay. The same pattern holds under buffer, look-ahead, communication-qubit, EPR-latency, fidelity, and surface-code settings, supporting QAOA as a diagnostic benchmark for al

Load-bearing premise

The comparisons treat local gate time as negligible next to fixed entanglement and switch delays, and they score only a single QAOA layer at the routing level without full state evolution or solution quality.

Editorial extensions

If this is right

  • Sparse graph optimization workloads still gain from look-ahead entanglement scheduling even when absolute communication demand is only moderate.
  • Once path diversity exists, topology choice matters less for QAOA than for denser circuits; cross-rack entanglement availability remains the main bottleneck.
  • Most of the latency gain appears at moderate buffer sizes and shallow-to-moderate look-ahead depths; further resources yield diminishing returns.
  • Network-aware scheduling remains beneficial for QAOA under surface-code fault-tolerant communication assumptions.
  • QAOA can serve as a standard contrast workload when ranking quantum data-center compilers and interconnects.

Reading between the lines

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

  • Compiler and fabric benchmarks that only use dense all-to-all circuits will overfit latency-hiding mechanisms and miss placement and partitioning effects that dominate sparse graphs.
  • If hardware narrows the gap between in-rack and cross-rack entanglement time, QAOA’s diagnostic value will shift from schedule latency hiding toward qubit placement quality.
  • Higher-degree or weighted QAOA graphs would likely expose sharper cliffs in buffer and communication-qubit demand than the 3-regular instances reported here.
  • Edge-aware placement that deliberately cuts few graph edges across racks could shrink QAOA’s cross-rack share enough for topology differences to reappear as first-order effects.
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 / 6 minor

Summary. The manuscript extends the SwitchQNet distributed-quantum-compiler framework (ISCA 2025) with a routing generator that translates QAOA MaxCut-style edge interactions (CX–Rz–CX per graph edge, p=1, routing-level only) into remote-CX requests, and evaluates the resulting workload under SwitchQNet's event-driven simulation on Clos, fat-tree, and spine-leaf quantum data-center topologies. Reported results (Table II) show end-to-end communication-latency improvements of roughly 1.36×–2.17× over a QuComm-style baseline across configurations, with sensitivity studies over buffer size, look-ahead depth, communication-qubit count, and EPR latencies (Figs. 2–3), plus one QEC-enabled (d=5 surface code) data point. The authors position QAOA as a diagnostic, graph-structured workload that stresses the scheduler differently from denser benchmarks (QFT, Grover, RCA). Scope is explicitly limited: no variational loop, no state evolution, no solution quality.

Significance. If the numbers hold, the paper makes a useful, honest contribution to QDC benchmarking: it adds an optimization-algorithm workload to a switch-based QDC simulator, ships public code (ref. [35]), and provides concrete, checkable quantitative predictions (latency, EPR overhead, wait time, retry overhead per configuration in Table II) rather than qualitative claims. The observation that QAOA's sparser communication pattern yields smaller but stable scheduler gains is a genuinely informative contrast result for algorithm–architecture co-design. However, the contribution is incremental over SwitchQNet — the scheduling machinery, latency constants, and baseline are inherited wholesale — so the paper's value rests entirely on the credibility of its empirical numbers, which currently rest on single deterministic instances and under-explained configuration differences between table rows.

major comments (4)
  1. [§IV.A, Table II] Table II contains two QAOA-480 Clos rows with incompatible numbers: the "Increase #QPUs/rack" and "#qubits/QPU" groups report baseline 20,704 / ours 10,363 (2.00×, #dist. EPR 737), while the "Clos top. (unchanged)" row reports baseline 22,695 / ours 16,722 (1.36×, #dist. EPR 84). Same workload, same stated topology, yet latencies differ by up to 60% and distillation counts by ~9×. Whatever configuration differs (fidelity thresholds? distillation enabled? buffer defaults?) is never stated, yet the topology-comparison conclusion ("QAOA is relatively insensitive to topology", 1.36×–1.42×) rests on exactly these rows. This must be documented; as written the topology claim is not interpretable.
  2. [§II.A–B, Table II] Each Table II row is a single run of one deterministically generated k=3 graph (one pairing heuristic, §II.B) under one fixed placement (integer division of qubit index by qubits-per-QPU, §II.A). For a degree-3 graph with contiguous placement, the cross-QPU edge fraction — hence the entire remote-CX demand and both numerator and denominator of every improvement factor — is highly sensitive to the random pairing and to placement. No error bars, instance sweep, or placement randomization is reported anywhere. The abstract and §IV.A claim "consistent" latency reductions; consistency across one instance per configuration is asserted, not demonstrated. Please report variance over ≥ several graph instances and at least one non-trivial placement policy (or a placement-permutation sensitivity check).
  3. [§III.C, Fig. 3] Latency is normalized by reconfiguration latency (1 ms), dominated by the fixed 10 ms cross-rack EPR constant, and intra-QPU gate time is set to zero (§III.C). The improvement factor is thus a ratio of how well two schedulers overlap constant-cost EPR streams. Fig. 3b does sweep cross-rack/reconfig over 5–30 and the ~2× factor survives — that is a real robustness check and should be stated as such. But no experiment includes nonzero local compute time, and only p=1 is evaluated; if local gates or multi-layer depth contribute comparably to cross-rack EPR, the reported factors need not hold. Please add either one sensitivity point with nonzero intra-QPU gate cost or an explicit analytic statement of when the factor degrades.
  4. [§III.C, Table II] Table II shows "Ours Wait" of 12.6–19.9 vs baseline waits of 0.11–0.78 — roughly a 30–100× increase in average EPR buffer residence time under SwitchQNet scheduling. §III.C sets this aside as hardware-dependent, and §V concludes overhead is not "prohibitive". But with cross-rack generation at 10 ms, pairs waiting ~17 normalized ms are stored longer than their own generation time; under any finite coherence time this is a fidelity cost that directly trades against the reported latency win, and it is currently unquantified. At minimum, report wait-time in absolute units alongside an assumed coherence-time bound under which the latency gain remains a net fidelity win, or soften the conclusion accordingly.
minor comments (6)
  1. [Figs. 2–3] Figs. 2–3: the inline "1.1x/2.1x" annotations presumably mark improvement-factor range endpoints, but this is not stated in captions; Fig. 3 x-axis labels ("cross-rack / reconfig") are ambiguous — state that the axis is the EPR-generation-to-reconfiguration latency ratio.
  2. [§III.A] §III.A states "Scheduling times wherever mentioned are in seconds (s)", but no scheduling/wall-clock times appear anywhere in the paper; either report scheduler runtime per Table II row (relevant for a compiler paper) or remove the sentence.
  3. [§I–IV] Typesetting artifacts throughout: "Ferrariet al." (missing space, §I), "forplayers" and "2pparameters" (missing spaces/math mode around p, §I), headings "SOLUTIONMETHODOLOGY", "FRAMEWORKSETUP", "EXPERIMENTALRESULTS" lack interword spacing.
  4. [§III.C, Eq. (1)] Eq. (1): retry overhead is defined over "time step numbers" t_attempts and t_compilation, but the distinction (attempted vs. actual steps) is only clear after reading the retry-logic description in §II.C; please define the terms at first use. Also, retry overhead is 1.00 in nearly all Table II rows — state explicitly that retries essentially never fired, and under what fidelity assumptions that holds.
  5. [§IV.C] The QEC experiment (§IV.C) is a single configuration (QAOA-480, d=5). A sentence on how the 1.51× factor and 12% EPR overhead scale with code distance, or an explicit statement that this is out of scope, would set reader expectations.
  6. [§IV.B] §IV.B: "saturation near 11 buffer qubits, ~26.8% of total qubits per QPU" — verify: 11/30 = 36.7%, not 26.8%; if the denominator includes communication qubits or a different total, state it.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: QAOA latency ratios are forward simulation outputs under fixed SwitchQNet assumptions, not quantities forced by definition or self-fit.

full rationale

This is an empirical systems paper. The load-bearing chain is: (1) map 3-regular MaxCut/QAOA edges to remote-CX requests via placement; (2) feed that request list into SwitchQNet’s existing scheduler and a QuComm-style baseline under stated latency/topology constants; (3) report simulated communication latency, EPR counts, wait time, and retry overhead, plus sensitivity sweeps. Improvement factors in Table II are ratios of two independently scheduled runtimes on the same request stream—they are not algebraic restatements of fitted targets, nor are they defined in terms of the claimed speedup. Retry overhead is explicitly definitional (t_attempts/t_compilation) and is reported as a metric, not sold as a derived prediction. Dependence on SwitchQNet’s prior mechanisms and fixed EPR/reconfiguration latencies is framework inheritance and modeling assumption, not a self-citation uniqueness chain or a fit-then-predict loop. No uniqueness theorem, ansatz smuggled via overlapping-author citation, or renaming of a known empirical law appears as a load-bearing step. Honest finding: no significant circularity.

Assumptions & free parameters 7 free parameters · 6 assumptions · 1 invented entities

The central latency-comparison claim rests on SwitchQNet’s inherited physical/scheduling model, a simplified QAOA communication abstraction (p=1, CX–Rz–CX edges only), fixed numerical latencies and resource counts, and a QuComm-like baseline—not on new physics or fitted universal constants. Free parameters are mostly configuration knobs chosen for the study; axioms are standard QAOA/DQC modeling choices plus SwitchQNet domain assumptions.

free parameters (7)
  • in-rack EPR generation latency = 0.1 ms
    Fixed at 0.1 ms following SwitchQNet; directly drives normalized latency and which operations dominate.
  • cross-rack EPR generation latency = 10 ms
    Fixed at 10 ms; paper states cross-rack EPR dominates QAOA latency, so this choice heavily shapes improvement factors.
  • switch reconfiguration latency = 1 ms
    Fixed at 1 ms and used as the normalization unit for reported latency and wait time.
  • look-ahead depth (default) = 10
    Default 10 in main experiments; sensitivity shows most gains at shallow-to-moderate depth—hand-chosen compiler knob.
  • communication qubits per QPU = 2 (default)
    Default two reserved communication qubits; varied in sensitivity and affects concurrency of entanglement.
  • QAOA graph degree k and layer count p = k=3, p=1
    Experiments use k=3 regular graphs and p=1 routing only; communication density and depth are chosen, not derived.
  • surface-code distance d for QEC experiment = d=5
    Single-point QEC result uses d=5 on Program-480; not swept.
assumptions (6)
  • domain assumption After qubit placement, only nonlocal CX legs of each edge’s CX–Rz–CX cost term generate inter-QPU routing requests; mixers and Rz are local.
    Section II.A; defines the entire communication workload extracted from QAOA.
  • domain assumption Intra-QPU computation time is negligible relative to inter-QPU communication and can be omitted from latency.
    Metrics section III.C; without this, end-to-end rankings could change.
  • domain assumption Clos/fat-tree/spine-leaf switch QDCs with stated BSM, buffer, and cross-rack weight model adequately represent future quantum data centers for relative scheduler comparison.
    Architecture setup III.A and inheritance from SwitchQNet/QDC literature.
  • domain assumption Buffer-aware QuComm-style compilation is an appropriate baseline for claiming SwitchQNet improvement factors.
    Comparison baseline III.D; improvement factors are only meaningful relative to this choice.
  • ad hoc to paper Evaluating a single QAOA layer’s routing list (no variational loop, no statevector/shot outcomes) is sufficient to characterize network interaction for this study’s claims.
    Explicitly scoped in II.A; solution quality and multi-layer accumulation are out of scope by choice.
  • standard math Standard linear-algebraic QAOA/MaxCut cost Hamiltonian and gate model (Farhi et al.).
    Introduction recall of HC and UC/UM alternation; not re-derived.
invented entities (1)
  • QAOA routing generator for SwitchQNet independent evidence
    purpose: Convert problem-graph edges into SwitchQNet remote-CX request lists after QPU placement.
    Software adapter, not a new physical entity; independent evidence is the linked implementation and reported runs.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Performance Analysis of QAOA Across Distributed Quantum Network Topologies Using SwitchQNet." pith.science (2026). https://pith.science/paper/YCSISX6Z

@misc{pith2026260723407,
  author       = {Pith},
  title        = {Pith review of: Performance Analysis of QAOA Across Distributed Quantum Network Topologies Using SwitchQNet},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YCSISX6Z}},
  note         = {Machine review of arXiv:2607.23407}
}
read the original abstract

Quantum data-center (QDC) architectures aim to scale distributed quantum computing (DQC) by interconnecting multiple quantum processing units (QPUs), but their performance depends strongly on how algorithmic communication patterns interact with entanglement generation, switch reconfiguration, and network topology. This paper studies the Quantum Approximate Optimization Algorithm (QAOA) as a graph-structured optimization workload for QDC-based distributed quantum computing. We adapt QAOA to SwitchQNet, a distributed quantum compiler framework that schedules communication and entanglement generation over switch-based QDC networks, by adding a routing generator that converts graph-dependent two-qubit cost interactions into remote-CX communication requests across QPUs. Using this extension, we evaluate QAOA instances across Clos, fat-tree, and spine-leaf topologies, measuring communication latency, EPR-pair overhead, EPR wait time, retry overhead, and sensitivity to buffer size, look-ahead depth, communication-qubit count, EPR latency, and EPR fidelity assumptions. The results show that QAOA obtains modest but consistent latency reductions, highlighting its value as a diagnostic benchmark for studying the interaction between algorithm structure, entanglement management, and quantum-network architecture.

Figures

Figures reproduced from arXiv: 2607.23407 by the authors.

Figure 1
Figure 1. Clos topology for n = 4 generated from gen_clos_conn function. Each rack contains one ToR switch and four QPUs. TABLE I: QAOA Workload Configurations Workload Racks QPUs/Rack Qubits/QPU QAOA-480 4 4 30 QAOA-360 4 3 30 QAOA-540 9 3 20 QAOA-600 4 5 30 QAOA-608 4 4 38 QAOA-720 4 6 30 QAOA-960 16 3 20 Fat-tree QAOA-960 8 3 30 Spine-leaf QAOA-720 6 4 30 destination ports. Fat-trees in SwitchQNet follow a standard impleme… view at source ↗
Figure 2
Figure 2. QAOA performance improvement of SwitchQNet varying with (a) buffer-size and (b) look-ahead depth [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. QAOA sensitivity varying with (a) communication qubits per QPU, (b) cross-rack EPR latency normalized by [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

36 extracted references · 6 linked inside Pith

  1. [35]

    Quantum Network Compiler Minimal Codes (SwitchQNet + QAOA Extension),

    S. Sharma, “Quantum Network Compiler Minimal Codes (SwitchQNet + QAOA Extension),” GitHub repository, 2026. [Online]. Available: https: //github.com/starktech23/quantum network compiler-minimal codes

  2. [1]

    M. A. Nielsen and I. L. Chuang,Quantum Computation and Quantum Information: 10th Anniversary Edition. Cambridge, U.K.: Cambridge University Press, 2011

  3. [2]

    Quantum computation,

    D. P. DiVincenzo, “Quantum computation,”Science, vol. 270, no. 5234, pp. 255-261, 1995

  4. [3]

    The physical implementation of quantum computa- tion,

    D. P. DiVincenzo, “The physical implementation of quantum computa- tion,”F ortschritte der Physik, vol. 48, no. 9-11, pp. 771-783, 2000

  5. [4]

    Quantum computing in the NISQ era and beyond,

    J. Preskill, “Quantum computing in the NISQ era and beyond,”Quan- tum, vol. 2, p. 79, 2018

  6. [5]

    How parallel circuit execution can be useful for NISQ computing?,

    S. Niu and A. Todri-Sanial, “How parallel circuit execution can be useful for NISQ computing?,” inProc. Design, Automation & Test in Europe Conf. (DATE), Antwerp, Belgium, 2022, pp. 1065-1070

  7. [6]

    Enabling multi-programming mechanism for quantum computing in the NISQ era,

    S. Niu and A. Todri-Sanial, “Enabling multi-programming mechanism for quantum computing in the NISQ era,”Quantum, vol. 7, p. 925, 2023

  8. [7]

    Simultaneous execution of quantum circuits on current and near-future NISQ systems,

    Y . Ohkura, T. Satoh, and R. Van Meter, “Simultaneous execution of quantum circuits on current and near-future NISQ systems,”IEEE Transactions on Quantum Engineering, vol. 3, pp. 1-10, 2022

Show all 36 references
  1. [8]

    Quantum error correction for beginners,

    S. J. Devitt, W. J. Munro, and K. Nemoto, “Quantum error correction for beginners,”Reports on Progress in Physics, vol. 76, no. 7, p. 076001, 2013

  2. [9]

    Surface codes: Towards practical large-scale quantum computation,

    A. G. Fowler, M. Matteo, J. M. Martinis, and A. N. Cleland, “Surface codes: Towards practical large-scale quantum computation,”Physical Review A, vol. 86, p. 032324, 2012

  3. [10]

    Distributed arithmetic on a quantum multicomputer,

    R. Van Meter, K. Nemoto, W. J. Munro, and K. M. Itoh, “Distributed arithmetic on a quantum multicomputer,”SIGARCH Computer Archi- tecture News, vol. 34, no. 2, pp. 354-365, 2006

  4. [11]

    Architecture of a quantum multicomputer opti- mized for Shor’s factoring algorithm,

    R. D. Van Meter III, “Architecture of a quantum multicomputer opti- mized for Shor’s factoring algorithm,” arXiv:quant-ph/0607065, 2006

  5. [12]

    Van Meter,Quantum Networking

    R. Van Meter,Quantum Networking. John Wiley & Sons, 2014

  6. [13]

    Quantum Internet: Networking challenges in distributed quantum computing,

    A. S. Cacciapuoti, M. Caleffi, F. Tafuri, F. S. Cataliotti, S. Gherardini, and G. Bianchi, “Quantum Internet: Networking challenges in distributed quantum computing,”IEEE Network, vol. 34, no. 1, pp. 137-143, 2020

  7. [14]

    Quantum networking and internetworking,

    R. Van Meter, “Quantum networking and internetworking,”IEEE Net- work, vol. 26, no. 4, pp. 59-64, 2012

  8. [15]

    A quantum Internet architecture,

    R. Van Meter, R. Satoh, N. Benchasattabuse, K. Teramoto, T. Matsuo, M. Hajdusek, T. Satoh, S. Nagayama, and S. Suzuki, “A quantum Internet architecture,” inProc. IEEE International Conference on Quantum Computing and Engineering (QCE), 2022, pp. 341-352

  9. [16]

    Q-Fly: An Optical Interconnect for Modular Quantum Computers,

    D. Sakumaet al., “Q-Fly: An Optical Interconnect for Modular Quantum Computers,” arXiv:2412.09299v4, 2025

  10. [17]

    Quantum data center infrastructures: A scalable architectural design perspective,

    H. Shapourian, E. Kaur, T. Sewell, J. Zhao, M. Kilzer, R. Kompella, and R. Nejabati, “Quantum data center infrastructures: A scalable architectural design perspective,” arXiv:2501.05598, 2025

  11. [18]

    SwitchQNet: Optimizing distributed quantum computing for quantum data centers with switch networks,

    H. Zhang, Y . Xu, H. Hu, K. Yin, H. Shapourian, J. Zhao, R. R. Kom- pella, R. Nejabati, and Y . Ding, “SwitchQNet: Optimizing distributed quantum computing for quantum data centers with switch networks,” in Proc. 52nd Annual International Symposium on Computer Architecture (IS...

  12. [19]

    A study of non-blocking switching networks,

    C. Clos, “A study of non-blocking switching networks,”The Bell System Technical Journal, vol. 32, no. 2, pp. 406-424, 1953

  13. [20]

    Fat-trees: Universal networks for hardware-efficient supercomputing,

    C. E. Leiserson, “Fat-trees: Universal networks for hardware-efficient supercomputing,”IEEE Transactions on Computers, vol. C-34, no. 10, pp. 892-901, 1985

  14. [21]

    A scalable, commodity data center network architecture,

    M. Al-Fares, A. Loukissas, and A. Vahdat, “A scalable, commodity data center network architecture,”SIGCOMM Computer Communication Review, vol. 38, no. 4, pp. 63-74, 2008

  15. [22]

    Revolutionizing datacenter networks via reconfigurable topologies,

    C. Avin and S. Schmid, “Revolutionizing datacenter networks via reconfigurable topologies,”Communications of the ACM, vol. 68, no. 6, pp. 44–53, 2025

  16. [23]

    Optical switching for data center net- works,

    N. Parsons and N. Calabretta, “Optical switching for data center net- works,” inSpringer Handbook of Optical Networks. Cham, Switzerland: Springer, 2020, pp. 795-825

  17. [24]

    New structural quantum circuit simulating a Toffoli gate,

    M. Asano and C. Ishii, “New structural quantum circuit simulating a Toffoli gate,” arXiv:quant-ph/0512016, 2005

  18. [25]

    Elementary gates for quantum computation,

    A. Barenco, C. H. Bennett, R. Cleve, D. P. DiVincenzo, N. Margolus, P. Shor, T. Sleator, J. A. Smolin, and H. Weinfurter, “Elementary gates for quantum computation,”Physical Review A, vol. 52, pp. 3457-3467, 1995

  19. [26]

    An approximate Fourier transform useful in quantum factoring,

    D. Coppersmith, “An approximate Fourier transform useful in quantum factoring,” arXiv:quant-ph/0201067, 2002

  20. [27]

    A fast quantum mechanical algorithm for database search,

    L. K. Grover, “A fast quantum mechanical algorithm for database search,” inProc. 28th Annual ACM Symposium on Theory of Computing (STOC), Philadelphia, PA, USA, 1996, pp. 212-219

  21. [28]

    A new quantum ripple-carry addition circuit,

    S. A. Cuccaro, T. G. Draper, S. A. Kutin, and D. P. Moulton, “A new quantum ripple-carry addition circuit,” arXiv:quant-ph/0410184, 2004

  22. [29]

    A quantum approximate optimization algorithm,

    E. Farhi, J. Goldstone, and S. Gutmann, “A quantum approximate optimization algorithm,” arXiv:1411.4028, 2014

  23. [30]

    Benchmarking quantum data center architectures: A performance and scalability perspective,

    S. Pouryousef, E. Kaur, H. Shapourian, D. Towsley, R. Kompella, and R. Nejabati, “Benchmarking quantum data center architectures: A performance and scalability perspective,” arXiv:2601.01353, 2026

  24. [31]

    Compiler Design for Distributed Quantum Computing,

    D. Ferrari, A. S. Cacciapuoti, M. Amoretti and M. Caleffi, “Compiler Design for Distributed Quantum Computing,” inIEEE Transactions on Quantum Engineering, vol. 2, pp. 1-20, 2021

  25. [32]

    DQC-QR: Distributing and routing quantum circuits with minimum execution time,

    R. Sundaram, H. Gupta, and C. Ramakrishnan, “DQC-QR: Distributing and routing quantum circuits with minimum execution time,”ACM Transactions on Quantum Computing, vol. 6, no. 4, pp 1-26, 2025

  26. [33]

    AutoComm: A Framework for Enabling Efficient Communication in Distributed Quantum Programs,

    A. Wu, H. Zhang, G. Li, A. Shabani, Y . Xie and Y . Ding, “AutoComm: A Framework for Enabling Efficient Communication in Distributed Quantum Programs,” inProc. 55th IEEE/ACM International Symposium on Microarchitecture (MICRO), Chicago, IL, USA, pp. 1027-1041, 2022

  27. [34]

    QuComm: Optimizing collective com- munication for distributed quantum computing,

    A. Wu, Y . Ding, and A. Li, “QuComm: Optimizing collective com- munication for distributed quantum computing,” inProc. 56th Annual IEEE/ACM International Symposium on Microarchitecture (MICRO), Toronto, ON, Canada, pp. 479-493, 2023

  28. [36]

    SwitchQNet: Optimizing Distributed Quantum Computing for Quantum Data Centers with Switch Networks,

    H. Zhang, “SwitchQNet: Optimizing Distributed Quantum Computing for Quantum Data Centers with Switch Networks,” Zenodo, 2025. [Online]. Available: https://zenodo.org/records/15377656

Pith tools

Reviewed July 30, 2026 · model on record in the stance chip above.