{"id":"377a45f2-894e-429b-8107-47fdd40352b4","arxiv_id":"2608.05821","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"A globally controlled qubit ring with no transport auxiliary qubits can run a cyclic stabilizer code, with simulated QEC thresholds around 0.1-0.5%.","lead":"The authors propose a quantum computer whose qubits are arranged in a ring and driven by a few global control fields, with no extra helper qubits for moving information. They simulate an error-correcting code on this ring and report error thresholds near 0.1-0.5%, which they claim are many orders higher than earlier global-control designs.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed ~10^-3 threshold is computed with a phenomenological noise model that treats all iSWAP/SWAP routing gates as perfect; circuit-level gate errors are likely to shift the threshold by orders of magnitude.","rationale":"The reader's weakest assumption identifies the same load-bearing issue: the threshold is computed under a phenomenological noise model that omits gate errors. This is indeed the most fragile premise because the architecture's QEC procedure is gate-heavy: every syndrome-extraction step and every measurement-routing SWAP involves global two-qubit gates acting on data qubits. The reported threshold of about 0.23-0.47% is within the range of current two-qubit gate error rates, so treating those gates as perfect could easily inflate the threshold by orders of magnitude. The comparison to Kay [28] makes the claim even more sensitive, since Kay's threshold was derived under a more restrictive control model. A circuit-level simulation with depolarizing noise on all gates is the natural and decisive test. I do not see an internal logical contradiction in the routing proofs; the concern is about the numerical threshold and its comparison, which the reader already flagged. Therefore the verdict remains CONDITIONAL: the paper should be accepted only after the missing circuit-level analysis is provided or the claims are appropriately qualified.","tokens_in":18376,"tokens_out":5063,"duration_ms":55536,"concrete_test":"Compile the full cyclic-code syndrome-extraction circuit from Ref [38] onto the global ring, including the M measurement rounds and the SWAP-based transport of syndrome qubits to measurement sites. Run Stim simulations for d=5,7,9 with a depolarizing channel at rate p on every global iSWAP, every SWAP (decomposed as in Eq. 7), every single-qubit gate, and every measurement, with no other error sources. Extract the logical-error-rate crossing point for each code distance and compare to the phenomenological thresholds in Fig. 2. If the circuit-level threshold is below ~10^-4, or if the crossing disappears, the paper's central claim that global-control QEC is practical at ~10^-3 error rates is not supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central numerical claim depends on a noise model that includes only idle data depolarization and syndrome measurement errors, while all global iSWAP/SWAP gates, local rotations, and transport operations are assumed noiseless. The paper states: 'We assume a phenomenological noise model based on the two major sources of error...' and inserts depolarizing noise only on data qubits between SE rounds, plus bit-flip noise before syndrome measurement. In this architecture, however, syndrome extraction and the serial measurement routing require many global two-qubit gates: each SWAP is decomposed into three iSWAPs plus global sqrt(X) pulses (Eq. 7), and syndrome qubits are carried to measurement sites by global SWAPs. These gates act on data qubits as well, so a gate error directly produces correlated data errors at a rate proportional to circuit depth rather than idle time. The claimed threshold of roughly 10^-3 is comparable to realistic two-qubit gate error rates, so omitting gate errors is not a benign simplification. Moreover, the seven-order comparison to Kay [28] is only meaningful if the noise models are aligned; Kay included costs of restricted control that this model does not. Without a circuit-level simulation, the headline threshold is untested.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a global-control quantum processor consisting of a ring of qubits with a single local control site, global iSWAP/SWAP transport, and no dedicated addressing ancillas. It identifies a family of cyclic stabilizer codes whose syndrome extraction is claimed to be implementable using only global gates, and it reports Monte Carlo QEC thresholds of roughly 0.23%–0.47% for different numbers of measurement rounds, which the authors state are nearly seven orders of magnitude larger than a previous global-control threshold estimate. The paper also discusses how adding local measurement sites improves the thresholds and sketches a superconducting-circuit implementation.","tokens_in":18624,"tokens_out":9717,"duration_ms":91939,"significance":"If the numerical claims were established, this would be a notable advance for global-control architectures: a minimally wired ring could run a stabilizer code with a threshold close to current experimental gate fidelities, and the measurement-site trade-off is a useful design principle. The paper is careful in defining the Hamiltonian, the routing primitives, and the universality argument; the single-qubit routing theorem in the Supplemental Material is a concrete formal contribution, and the threshold simulations use the external Stim simulator with a code family from Ref. [38], so the thresholds are outputs rather than inputs. However, the central threshold claim is currently conditioned on an idealized noise model that omits all gate errors, and the comparison to Ref. [28] is made under a different set of assumptions, so the headline seven-order improvement is not yet substantiated.","major_comments":[{"comment":"The thresholds p^{(M)}_{th} reported in Fig. 2 are computed under a noise model that includes only depolarizing errors on idle data qubits and bit-flip errors before syndrome measurement; all iSWAP/SWAP gates, local rotations, and transport operations are treated as noiseless. In this architecture, syndrome extraction and measurement routing require many global two-qubit gates, each SWAP costing three iSWAPs plus three global √X pulses (Eq. (7)), and these gates act on data qubits as well. Because the claimed thresholds (0.23%–0.47%) are comparable to realistic two-qubit gate error rates, omitting gate errors is not a benign approximation. The paper should either present circuit-level simulations that include gate errors or explicitly restrict the abstract's threshold claim to the idealized phenomenological model.","section":"Quantum error correction; Fig. 2"},{"comment":"The statement that the thresholds are 'nearly seven orders of magnitude larger' than the 10^{-10} estimate of Ref. [28] is not supported by the simulations presented, because the two calculations are not performed under the same noise model. The present model excludes gate errors entirely, whereas Ref. [28]'s estimate accounts for the restricted-control costs in a global-control scheme. A fair quantitative comparison requires either running the same code under equivalent assumptions or redoing the simulation with a circuit-level model; otherwise the improvement factor is an artifact of the differing models.","section":"Quantum error correction (comparison to Ref. [28])"},{"comment":"The central claim that the cyclic code's syndrome extraction is 'realized using only nearest-neighbor iSWAP gates plus single-qubit gates' and 'comprises solely operations that act in parallel across the entire register' is stated with references [38,68] but no explicit circuit or transpilation proof is given. Since the compatibility of the SE circuit with the global alternating-bond iSWAP pattern of Eqs. (5)–(6), rather than with arbitrary local nearest-neighbor gates, is load-bearing for the threshold simulations, the authors should provide the full SE circuit and its mapping onto the global gate set, or a rigorous argument that the circuit can be executed without additional serial routing.","section":"End Matter: Cyclic codes under global control"}],"minor_comments":[{"comment":"The phrases 'zero qubit overhead' and 'every physical qubit is a computational qubit' are inconsistent with the QEC layout in the 'Quantum error correction' section, where N=2n qubits and half are syndrome qubits. Please clarify that the zero overhead refers to the absence of control/address ancillas, not to the absence of standard QEC ancillas.","section":"Abstract and QEC section"},{"comment":"Table I is misformatted: the rows for M=2,3,4,5 appear to be missing their labels, making data such as '27 11 15 ∼26' unintelligible. The table should be restructured so that each row corresponds to a single value of M with columns for d=5, d=7, d=9, and the percentage.","section":"Table I"},{"comment":"The notation for the base stabilizers, e.g., 'ZIXXIZI ⊗7' for d=5, is ambiguous; please write explicit generator strings or define the multiplication convention clearly.","section":"End Matter"},{"comment":"For reproducibility, consider providing the Stim simulation code and the data sets used for Fig. 2 in an ancillary file or public repository.","section":"General reproducibility"}],"recommendation":"major_revision","confidential_remarks":"The manuscript fits the scope of a quantum information journal, and the architecture proposal is original. The main concern is that the headline threshold claim is presented as a property of the full architecture while the simulations omit all gate errors. I would not rule out acceptance after a major revision that either adds circuit-level simulations or clearly reframes the result as a phenomenological threshold and removes the unfair comparison to Ref. [28]. The reference list includes several self-citations, but this is not unusual for a group developing a specific architecture and does not affect my assessment."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague —\n\nThe key thing to know: the architecture is new and interesting, but the headline thresholds (~10^-3) are computed with a noise model that treats every routing iSWAP/SWAP gate as perfect. That means the central numerical claim is untested, not that the architecture is wrong.\n\nWhat is actually new and good: the all-qubit ring with a single locally controlled site, global iSWAP transport, and the cyclic stabilizer code from Simakov & Besedin is a real combination that I haven't seen elsewhere. The routing theorem in the SM is constructive and correct; the keyhole geometry with two ports and all-to-all connectivity is a nice twist; and the proposed superconducting implementation with always-on XY interactions and global flux lines is plausible. The paper is also honest about using a phenomenological noise model — the problem is what that model omits.\n\nThe soft spots, in proportion. The largest: syndrome extraction and measurement routing require many global two-qubit gates. Each SWAP decomposes into three iSWAPs plus global sqrt(X) pulses, and syndrome qubits are carried to the local measurement site by exactly those SWAPs. The simulation only inserts depolarizing noise on idle data qubits between SE rounds and bit-flip before syndrome measurement; all iSWAP/SWAP gates, local rotations, and transport are noiseless. Since the claimed threshold (~0.2–0.5%) is comparable to realistic two-qubit gate error rates, omitting gate errors is not a benign simplification. A circuit-level simulation with uniform depolarizing noise on every iSWAP could shift the threshold by orders of magnitude. Second, the seven-order comparison to Kay [28] mixes different noise models: Kay included restricted-control overheads that this model does not. The paper notes that Kay did not optimize the threshold, but that doesn't bridge the model gap. Third, the \"zero qubit overhead\" label is misleading: half the ring is syndrome qubits, which function as ancillas for the code. Protecting all qubits under one scheme is a real advantage, but \"zero overhead\" overstates it.\n\nWho should read this: anyone working on global control architectures or on code–hardware matching for QEC. The architectural ideas and the code mapping deserve serious attention; the threshold numbers do not, until they are re-derived with a circuit-level noise model. I would send this to peer review with the expectation of substantial revision. I would not cite the thresholds as they stand, but I would bring the architecture to a reading group.","headline":"The architecture is new and worth serious attention, but the headline ~10^-3 thresholds are computed with a noise model that treats every routing iSWAP/SWAP gate as perfect, so the central quantitative claim is untested.","tokens_in":19134,"tokens_out":3069,"would_cite":false,"duration_ms":29943,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81P70","81P68"],"pacs":["03.67.Pp","03.67.Lx"],"model":"deepseek-v4-flash","headline":"A ring of qubits controlled by global fields can run quantum error correction at a practical threshold.","keywords":["global control","quantum error correction","cyclic stabilizer code","iSWAP","fault tolerance threshold","superconducting qubits","syndrome extraction","zero-qubit-overhead architecture"],"falsifier":"Run the same Monte Carlo threshold extraction on the cyclic code with a depolarizing channel applied to every global iSWAP/SWAP and single-qubit rotation at the same strength as the data-idle error; if the crossing points vanish or fall by orders of magnitude, the reported $10^{-3}$ thresholds depend on the perfect-gate assumption.","tokens_in":18203,"feed_emoji":"🛡️","tokens_out":6066,"duration_ms":60123,"temperature":0.7,"pith_summary":"This paper claims that quantum error correction can run on a globally controlled ring of qubits with no auxiliary qubits, using only parallel global gates plus one locally controlled site. The proposed cyclic stabilizer code makes every physical qubit a data or syndrome qubit, so a single error-correcting scheme protects the whole device. Monte Carlo simulations under a noise model that keeps gates perfect but includes idle-data depolarization and measurement errors give QEC thresholds around $10^{-3}$, roughly seven orders of magnitude higher than earlier global-control estimates. If true, this would mean the wiring bottleneck does not fundamentally block fault tolerance, and global-control processors could operate at physical error rates comparable to current devices.","feed_headline":"Qubit ring lifts global-control QEC threshold to ~10^-3","feed_subtitle":"No auxiliary qubits: every qubit is protected by one cyclic code, so syndrome extraction is fully parallel.","key_machinery":"The machinery is the pairing of global SWAP transport with a cyclic stabilizer code. The ring alternates two global coupling patterns that act on even and odd bonds, producing parallel iSWAP layers; correcting each iSWAP to a SWAP with global rotations lets any qubit be moved to the single local-control site, while the code's stabilizer generators are cyclic translations of a base operator. Because the code is translation-invariant, syndrome extraction uses only spatially uniform parallel operations, so stabilizer information accumulates in syndrome qubits through the same global dynamics; only the final readout of syndrome qubits is serial, with global SWAPs carrying each syndrome qubit to the local measurement site.","core_discovery":"The paper's central discovery is a zero-overhead global-control architecture: a homogeneous ring of XY-coupled qubits driven by two global coupling patterns and one locally addressed site, on which a cyclic stabilizer code $[\\![n,1,d]\\!]$ has syndrome extraction built entirely from global iSWAP gates and global single-qubit rotations. Qubit states are transported around the ring by decomposing the native iSWAP into SWAP using three iSWAPs and global $\\sqrt{X}$ pulses, and syndrome qubits are then shuttled to the local measurement site. Simulating the code with a stabilizer-circuit simulator and decoding with a look-up table augmented by a memory, the paper reports logical-error thresholds of $p_{\\mathrm{th}}^{(2)}=0.47\\%$, $p_{\\mathrm{th}}^{(3)}=0.35\\%$, and $p_{\\mathrm{th}}^{(5)}=0.23\\%$ for different numbers of measurement rounds, placing the thresholds near $10^{-3}$ and nearly seven orders of magnitude above a prior global-control threshold estimate near $10^{-10}$. The thresholds improve as local measurement sites are added, revealing a tunable trade-off between wiring simplicity and fault-tolerant performance.","pith_inferences":["The paper computes thresholds with all gates assumed perfect; a natural next test is to repeat the threshold extraction with a depolarizing rate applied to every global iSWAP/SWAP and local rotation, since finite gate errors could substantially move or erase the reported crossing points.","The systematic improvement of thresholds with added measurement sites suggests a design rule that the paper does not spell out: for a fixed wiring budget, choose the smallest number of measurement rounds that keeps the logical error rate below the physical error rate.","The comparison to the earlier $10^{-10}$ global-control threshold may depend on different noise models; re-simulating that earlier scheme with the same phenomenological noise would give a cleaner benchmark for the claimed seven-order improvement.","Because the cyclic code is not a quantum low-density parity-check code, the architecture may be best suited to near-term small logical qubits; the keyhole connectivity mentioned in the paper could be explored as a route to global-control qLDPC implementations."],"forward_implications":["A globally controlled processor with zero auxiliary qubits could in principle run fault-tolerant quantum error correction at physical error rates near $10^{-3}$, a regime comparable to current hardware, rather than the much lower rates suggested by earlier global-control schemes.","Because every physical qubit is protected by the same error-correcting scheme, the dual computational/auxiliary correction overhead that forced two interleaved correction procedures in previous global-control proposals is eliminated.","Increasing the number of local measurement sites systematically raises the threshold, giving designers a quantitative trade-off between measurement wiring density and error-correction performance.","The architecture's effective connectivity can approach all-to-all in a keyhole geometry, which may ease future implementation of qLDPC codes under global control.","The ring can act as a building block for a scalable logical architecture, with multiple rings tiled and coupled through shared physical qubits so that transport and entangling gates are lifted to the logical level."],"supporting_citations":[{"why":"Supplies the cyclic stabilizer-code family and the syndrome-extraction construction that the architecture executes with parallel global gates.","marker":"[38]"},{"why":"Provides the earlier global-control threshold near $10^{-10}$ that the paper's $10^{-3}$ thresholds are compared against.","marker":"[28]"},{"why":"Provides the stabilizer-circuit simulator used for the Monte Carlo logical-error-rate simulations and threshold extraction.","marker":"[48]"},{"why":"Demonstrates the iSWAP gate from XY interaction and its decomposition, which underlies the SWAP transport used in the ring.","marker":"[37]"},{"why":"Supplies synthesis results for energy-conserving XY circuits used in compiling iSWAP-based SWAP operations.","marker":"[40]"},{"why":"Provides the standard formalisms for stabilizer codes, depolarizing noise, and the universal gate set used in the argument.","marker":"[47]"},{"why":"Supports the tunable-coupler superconducting implementation used to realize the CZ-plus-iSWAP form of the SWAP gate.","marker":"[42]"},{"why":"Shows that globally optimized pulses can compensate disorder, supporting the claimed compatibility with realistic superconducting hardware.","marker":"[50]"}],"fun_headline_variants":["Zero-overhead global QEC on a ring hits 10^-3 threshold","Global iSWAP ring runs cyclic code, threshold near 10^-3","All qubits computational: global control QEC at 10^-3","Homogeneous ring: zero aux qubits, threshold jumps to 10^-3","Cyclic codes on global ring: fault tolerance without wiring"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The claimed thresholds assume all two-qubit gates, swaps, and rotations are perfect and only idle data-qubit depolarization plus syndrome measurement errors are included; if actual gate errors are comparable to the threshold, the reported numbers do not follow.","fun_headline_variants_meta":{"raw":{"variants":["Zero-overhead global QEC on a ring hits 10^-3 threshold","Global iSWAP ring runs cyclic code, threshold near 10^-3","All qubits computational: global control QEC at 10^-3","Homogeneous ring: zero aux qubits, threshold jumps to 10^-3","Cyclic codes on global ring: fault tolerance without wiring"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.0002,"raw_usage":{"total_tokens":1382,"prompt_tokens":960,"completion_tokens":422,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":576,"completion_tokens_details":{"reasoning_tokens":325}},"tokens_in":576,"tokens_out":422,"duration_ms":4617,"temperature":1.0,"reasoning_tokens":325,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T22:52:11.289608+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the same Monte Carlo threshold extraction on the cyclic code with a depolarizing channel applied to every global iSWAP/SWAP and single-qubit rotation at the same strength as the data-idle error; if the crossing points vanish or fall by orders of magnitude, the reported $10^{-3}$ thresholds depend on the perfect-gate assumption.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the cyclic stabilizer-code family and the syndrome-extraction construction that the architecture executes with parallel global gates."},{"cited_title":"Deriving a Fault-Tolerant Threshold for a Global Control Scheme","cited_arxiv_id":"quant-ph/0702239","evidence_quote":"Provides the earlier global-control threshold near $10^{-10}$ that the paper's $10^{-3}$ thresholds are compared against."},{"cited_title":"Bai and I","cited_arxiv_id":null,"evidence_quote":"Supplies synthesis results for energy-conserving XY circuits used in compiling iSWAP-based SWAP operations."},{"cited_title":"Sunget al., Realization of high-fidelity cz and zz-free iswap gates with a tunable coupler, Physical Review X11, 021058 (2021)","cited_arxiv_id":null,"evidence_quote":"Supports the tunable-coupler superconducting implementation used to realize the CZ-plus-iSWAP form of the SWAP gate."},{"cited_title":"Aiudi, J","cited_arxiv_id":null,"evidence_quote":"Shows that globally optimized pulses can compensate disorder, supporting the claimed compatibility with realistic superconducting hardware."}],"review_version":1}