{"id":"0f46b033-8176-4145-bb7e-4b5f225a984b","arxiv_id":"2412.15629","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":11,"one_line_summary":"A simulated three-transmon, single-resonator architecture performs pairwise CNOT gates with average fidelities between 0.96 and 0.99.","lead":"This paper simulates a design where three transmon qubits share one fixed-frequency resonator, reporting CNOT fidelities around 0.96 to 0.99. If the design holds up in hardware, it could let future superconducting processors connect more qubits with fewer couplers.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Closed-system fidelities do not establish the hardware-level claim; decoherence and spectator crosstalk could push the two sub-0.98 gates and the 0.984 average below the 0.98 threshold.","rationale":"The six-gate mean in Table 2 is 0.9839, so the literal 'average exceeds 0.98' claim survives despite two individual gates being below 0.98. The load-bearing weakness is nonetheless the same one the reader identified: the fidelities are closed-system values, while the paper's abstract and discussion make claims about preserving gate performance in real transmon hardware. The paper's own Methods acknowledge that relaxation and dephasing are present in real devices, but no such terms enter the fidelity calculation. Since the two weakest gates sit near 0.96-0.97, realistic noise could push both them and the average below the stated threshold. A Lindblad re-simulation with typical coherence times is a concrete, feasible check. The printed fidelity formula also appears mistyped, but that is secondary to the missing noise model; the conditional verdict remains appropriate.","tokens_in":12137,"tokens_out":10549,"duration_ms":98587,"concrete_test":"Re-simulate the six CNOT gates in Table 2 using a Lindblad master equation with the same Hamiltonian and pulse parameters, adding transmon relaxation and dephasing at T1=T2=100 microseconds and resonator decay at kappa=1/T1, plus a spectator Stark-shift or ZZ term of the size expected from the Gi=70 MHz couplings. Recompute the mean fidelity over the six gates; if it remains above 0.98, the concern is resolved, and if it falls below, the hardware claim should be weakened to a coherent-design claim.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that a single fixed-frequency resonator coupler can mediate three transmons while preserving gate performance rests on Table 2 fidelities computed from the closed Hamiltonian in Eqs. (1)-(4). The gates run 110-190 ns, and the paper itself notes that real hardware has relaxation and dephasing, yet no Lindblad, dephasing, or crosstalk terms are included. Typical transmon T1/T2 values of 50-150 microsecond contribute an estimated decoherence error of roughly 0.1-0.4% per gate. This is comparable to the margin by which the six-gate average (0.9839) exceeds 0.98 and is larger than the gap for CNOT12 (0.9640) and CNOT02 (0.9713). Furthermore, during every two-qubit gate the third qubit remains strongly coupled to the shared resonator; its residual interaction, Purcell decay through the resonator, and drive-line crosstalk are not modeled. Because the abstract and Discussion frame the result as a path to practical transmon machines, the coherent simulation alone is not sufficient to support the hardware-oriented claim.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a three-transmon architecture in which a single fixed-frequency resonator coupler mediates the interaction among all three qubits. The authors design cross-resonance based microwave pulses for CNOT gates on each qubit pair, optimize the pulse parameters numerically against an average fidelity objective, and report average fidelities in Table 2. The central claim is that a single resonator coupler can support more than two transmons while maintaining average two-qubit CNOT fidelities above 0.98.","tokens_in":12485,"tokens_out":6986,"duration_ms":61970,"significance":"The proposed architecture is practically motivated: fixed-frequency transmons with microwave-only control avoid flux-noise-induced dephasing, and a single resonator coupling several qubits could increase connectivity compared with pairwise transmon-resonator-transmon structures. The simulations use a multilevel Hamiltonian (including transmon levels up to |3> and resonator states up to |3>), which is more credible than a purely two-level treatment, and the authors publish their pulse parameters in a repository. If the fidelity metric is corrected and the closed-system limitation is made explicit, the work could provide a useful data point for experimental groups designing fixed-frequency processors. However, the paper's central quantitative claims currently rest on a fidelity formula that is incorrect as written, and the hardware-oriented conclusions are drawn from noise-free simulations.","major_comments":[{"comment":"The fidelity definition F_ψ = |<ψ|U†U|ψ>| is identically 1 for a unitary target gate U, so it cannot produce the values reported in Table 2. The manuscript needs to define the actual time-evolution operator (e.g., V or U_pulse) and use a standard expression such as F_ψ = |<ψ|U_target† U_pulse|ψ>|^2, with an explicit projection onto the computational three-qubit subspace. Without this correction, the numerical results are not reproducible as stated and the central claim is unverifiable.","section":"Quantum Gate Optimization (Eq. (12))"},{"comment":"The sentence \"CNOT gates with a fidelity of at least 0.988 can be achieved\" is contradicted by the paper's own Table 2, which lists CNOT12 = 0.9640 and CNOT02 = 0.9713, and by the corresponding basis-state success probabilities of 0.9681 and 0.9720 in Figure 3. If the intended claim is only that for each qubit pair at least one CNOT direction achieves a fidelity above 0.988, that should be stated explicitly; otherwise the statement must be revised.","section":"Discussion"},{"comment":"All fidelities are computed from the closed-system Hamiltonian in Eqs. (1)-(4), with no relaxation, dephasing, or parasitic crosstalk terms. The authors themselves note in Methods that real hardware has relaxation and dephasing, and the gate times are 110-190 ns. With typical transmon T1/T2 values of tens to hundreds of microseconds, the expected decoherence error is on the order of 0.1-0.4%, which is comparable to the margin by which the reported average (0.9839) exceeds 0.98 and is larger than the shortfall of the two sub-0.98 gates. The abstract and Discussion frame the design as \"preserving gate performance\" in a transmon-based quantum computer; that hardware-level claim requires at least an order-of-magnitude estimate of open-system effects or a Lindblad/dephasing simulation.","section":"Results and Discussion (closed-system simulation)"}],"minor_comments":[{"comment":"There are several typos, including \"It is appearant\" and \"SW AP gates\"; a careful proofreading pass is needed.","section":"Introduction"},{"comment":"The column header \"ΩX Sq, ΩS\" is ambiguous; please clarify the meaning of the two entries per gate (pulse amplitude and envelope shape) and their units.","section":"Table 2"},{"comment":"The two rows both labeled \"CNOT(sym.) 01\" should be distinguished, for example by indicating the control-target direction.","section":"Table 4"},{"comment":"The caption says \"The average of the gate success probability of (a)CNOT01...\" but the values listed are averages over the eight computational basis states; please make this explicit.","section":"Figure 3 caption"},{"comment":"The sentence \"The shape index q is not variable for the pulse optimization but fixes the shape of the envelope\" is unclear; it should specify that q is fixed beforehand and not optimized.","section":"Methods (pulse optimization)"},{"comment":"Some references have formatting errors, e.g., \"Nakamura, Y .,\" and inconsistent capitalization of journal names.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The core architecture idea is interesting, and the multilevel simulation approach is a strength. However, the fidelity formula in Eq. (12) is a fundamental error that undermines confidence in all reported numbers until corrected. The internal inconsistency about the 0.988 threshold and the absence of any decoherence model also need to be addressed before the paper can support its hardware-oriented conclusions. The authors should also consider stating more carefully that the reported fidelities are optimized best-case values for the chosen pulse family, not robust predictions under realistic noise."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper's core idea is a three-transmon architecture with a single fixed-frequency resonator coupler, driven by cross-resonance pulses to implement pairwise CNOT gates. That is a legitimate extension of the standard transmon-resonator-transmon setup, and the authors do real work: they include four transmon levels, check basis-state success probabilities, show Bloch trajectories, and give concrete pulse parameters. The comparison with symmetric ECR gates is a useful sanity check. The average CNOT fidelity across the six gates is 0.9839, so the headline claim \"average fidelity exceeds 0.98\" is literally true, and the success probabilities average 0.9859. I think the reader's take overstates the contradiction with Table 2; the issue is not that the average claim is false, but that two gates (CNOT12 at 0.9640 and CNOT02 at 0.9713) fall below 0.98, so the phrasing invites a misreading that every gate meets the threshold.\n\nTwo soft spots matter. First, Eq. (12) defines fidelity as |⟨ψ|U†U|ψ⟩|, which equals 1 by unitarity. That is a clear typo; they presumably meant the overlap with the actual time-evolution operator. It needs fixing, but it does not invalidate the numeric results. Second, the simulations are closed-system. There is no Lindblad, no dephasing, no spectator crosstalk, even though the gates run 110–190 ns and typical transmon coherence times would add roughly 0.1–0.4% error per gate. That is comparable to the margin above 0.98, so the claim that a real machine would preserve gate performance is not established by this paper. The authors acknowledge noise in passing, but they do not model it.\n\nThe pulse parameters are optimized with Nelder-Mead, so the fidelities are fitted, not independent predictions. That is normal for pulse design, and the direct simulation does demonstrate that the architecture is controllable—not every system reaches these numbers with CR pulses.\n\nWho is this for? Researchers working on superconducting qubit connectivity who want a concrete simulation study of a three-qubit-per-resonator layout. The paper is not a hardware demonstration, and the title is a bit grand for a closed-system study, but it is a legitimate design proposal. With the fidelity formula fixed and the claims reworded to explicitly say \"average\" and \"coherent simulation,\" it deserves peer review. I would send it out, and I'd flag the open-system question as the main thing the referees should push on.","headline":"A plausible three-transmon single-resonator design with solid closed-system simulations, but the fidelity formula is a typo and the hardware-level claim outruns the evidence.","tokens_in":12990,"tokens_out":3478,"would_cite":false,"duration_ms":28730,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A single fixed-frequency resonator can run CNOT gates among three transmon qubits with average fidelity above 0.98.","keywords":["three-transmon system","resonator coupler","cross-resonance gate","CNOT gate","average gate fidelity","transmon qubit","fixed-frequency qubit","all-microwave control"],"falsifier":"Add energy relaxation and dephasing to the Hamiltonian of Eqs. (1)-(4) and re-optimize the pulse parameters; if the average CNOT fidelity drops below 0.98, the closed-system claim does not transfer to real hardware. A direct device measurement of CNOT12, the lowest-fidelity gate in Table 2, would give a sharp test of the model.","tokens_in":11964,"feed_emoji":"⚛️","tokens_out":9704,"duration_ms":79949,"temperature":0.7,"pith_summary":"The paper proposes a wiring alternative to the usual one-resonator-per-qubit-pair layout: three fixed-frequency transmon qubits all coupled through a single resonator coupler. Using cross-resonance microwave pulses, the authors design CNOT gates for every ordered pair of qubits and report an average gate fidelity above 0.98 and gate times under 200 ns in a closed-system simulation. If this holds in hardware, a single resonator could fan out to more than two qubits, increasing connectivity in transmon processors without adding flux-tunable elements or extra couplers. The claim matters because connectivity and gate fidelity are two main bottlenecks in scaling superconducting quantum computers.","feed_headline":"One resonator can couple three transmon qubits at 98% gate fidelity","feed_subtitle":"The average over all six CNOT directions stays above 0.98 in a closed-system simulation.","key_machinery":"The mechanism that carries the design is the cross-resonance effect on a shared resonator bus. A microwave pulse applied to one transmon at the frequency of another transmon makes the second qubit rotate in a direction that depends on the first qubit's state; the resonator mediates the interaction without needing a flux-tunable coupler. The paper implements this with a sinusoidal flat-top envelope for the cross-resonance pulse, a Gaussian auxiliary pulse on the target qubit, DRAG correction to suppress population of the transmon's higher levels, and virtual Z gates to fix the frame. The simulation includes transmon levels up to the third excited state and treats the resonator staying in its ground state as the success condition.","core_discovery":"The paper's central claim is that the transmon-resonator-transmon building block is not the only viable wiring choice for a superconducting processor: a single fixed-frequency resonator coupler can mediate interactions among three transmon qubits at once. Concretely, the authors specify qubit frequencies, charging and Josephson energies, and qubit-resonator couplings (Table 1) and give optimized pulse parameters for six CNOT directions (Table 2). The average fidelity over these six gates exceeds 0.98, and the average basis-state success probability also exceeds 0.98. The protocol uses only local microwave drives: a cross-resonance pulse on the control qubit, an auxiliary Gaussian pulse on the target, DRAG shaping to suppress leakage, and virtual Z rotations. The authors position this as a connectivity improvement over two-qubit-per-coupler layouts and a gate-time improvement over echoed cross-resonance gates.","pith_inferences":["A natural next check is whether the idle qubit's state stays intact for all six gates and all input states; the paper shows the idle Bloch vector static only for CNOT01, and the shared bus may still cause spectator errors in other configurations.","The 0.98 headline is an average: CNOT12 (0.964) and CNOT02 (0.971) fall below it, so fault-tolerance thresholds should be evaluated against the worst pair rather than the average.","The same optimization pipeline could be applied to four or more transmons per resonator, but mode crowding and parasitic interactions would likely grow; the paper does not quantify that scaling limit.","A direct hardware test would be to fabricate the specified device and measure the six CNOT fidelities; the design's viability in a real processor depends on whether the omitted relaxation, dephasing, and crosstalk terms preserve the ordering of the simulated fidelities."],"forward_implications":["A single resonator can serve more than two qubits, so transmon processors could increase qubit connectivity without adding a coupler per pair.","Because all gates are driven by local microwave pulses with no flux bias, the design avoids flux-noise dephasing that limits tunable-frequency architectures.","CNOT gate times under 200 ns are shorter than echoed cross-resonance gates, which could reduce accumulated error in deep circuits.","Every ordered qubit pair has a direct CNOT, potentially reducing the number of SWAP gates needed to route algorithms on near-neighbor hardware.","The reported average fidelities above 0.98 and basis-state success probabilities provide a concrete benchmark that future hardware implementations can be compared against."],"supporting_citations":[{"why":"Defines the transmon Hamiltonian and its anharmonicity, which the three-qubit simulation is built on.","marker":"[8]"},{"why":"Introduces the all-microwave cross-resonance entangling gate for fixed-frequency qubits, the mechanism behind the CNOT protocol.","marker":"[13]"},{"why":"Establishes that linear couplings with fixed transition frequencies can implement universal microwave-only gates, supporting the fixed-frequency design.","marker":"[15]"},{"why":"Provides the echoed cross-resonance (ECR) gate whose gate time and fidelity the asymmetric CNOT is compared against.","marker":"[16]"},{"why":"Models the effective cross-resonance Hamiltonian and explains the role of higher transmon levels in CR gates.","marker":"[41]"},{"why":"Supplies the DRAG pulse shaping used to suppress leakage out of the computational subspace.","marker":"[43]"},{"why":"Provides the product-formula decomposition used to approximate the driven time evolution in simulation.","marker":"[45]"},{"why":"Defines the average gate fidelity metric used to score each CNOT gate.","marker":"[47]"},{"why":"Supplies the derivative-free simplex optimization routine used to minimize infidelity over the pulse parameters.","marker":"[48]"}],"fun_headline_variants":["Three transmons, one coupler: CNOT fidelity above 0.98","Single resonator links three transmons at >98% gate fidelity","Triple-transmon wiring: one coupler, six CNOTs above 0.98","Beyond two-qubit coupling: resonator serves three transmons","One fixed-frequency resonator drives three-transmon CNOTs"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the unmodeled effects of relaxation, dephasing, and parasitic crosstalk will not substantially lower the simulated fidelities when the design is built in real hardware.","fun_headline_variants_meta":{"raw":{"variants":["Three transmons, one coupler: CNOT fidelity above 0.98","Single resonator links three transmons at >98% gate fidelity","Triple-transmon wiring: one coupler, six CNOTs above 0.98","Beyond two-qubit coupling: resonator serves three transmons","One fixed-frequency resonator drives three-transmon CNOTs"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000536,"raw_usage":{"total_tokens":2568,"prompt_tokens":928,"completion_tokens":1640,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":544,"completion_tokens_details":{"reasoning_tokens":1545}},"tokens_in":544,"tokens_out":1640,"duration_ms":10737,"temperature":1.0,"reasoning_tokens":1545,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T11:14:38.065285+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Add energy relaxation and dephasing to the Hamiltonian of Eqs. (1)-(4) and re-optimize the pulse parameters; if the average CNOT fidelity drops below 0.98, the closed-system claim does not transfer to real hardware. A direct device measurement of CNOT12, the lowest-fidelity gate in Table 2, would give a sharp test of the model.","supporting_citations":[{"cited_title":"Charge-insensitive qubit design derived from the Cooper pair box","cited_arxiv_id":null,"evidence_quote":"Defines the transmon Hamiltonian and its anharmonicity, which the three-qubit simulation is built on."},{"cited_title":"Simple all-microwave entangling gate for fixed-frequency superconducting qubits","cited_arxiv_id":null,"evidence_quote":"Introduces the all-microwave cross-resonance entangling gate for fixed-frequency qubits, the mechanism behind the CNOT protocol."},{"cited_title":"Fully microwave-tunable universal gates in superconducting qubits with linear couplings and fixed transition frequencies","cited_arxiv_id":null,"evidence_quote":"Establishes that linear couplings with fixed transition frequencies can implement universal microwave-only gates, supporting the fixed-frequency design."},{"cited_title":"Procedure for systematically tuning up cross-talk in the cross-resonance gate","cited_arxiv_id":null,"evidence_quote":"Provides the echoed cross-resonance (ECR) gate whose gate time and fidelity the asymmetric CNOT is compared against."},{"cited_title":"Gambetta","cited_arxiv_id":null,"evidence_quote":"Models the effective cross-resonance Hamiltonian and explains the role of higher transmon levels in CR gates."},{"cited_title":"Simple pulses for elimination of leakage in weakly nonlinear qubits","cited_arxiv_id":null,"evidence_quote":"Supplies the DRAG pulse shaping used to suppress leakage out of the computational subspace."},{"cited_title":"Decomposition formulas of exponential operators and Lie exponentials with some applications to quantum mechanics and statistical physics","cited_arxiv_id":null,"evidence_quote":"Provides the product-formula decomposition used to approximate the driven time evolution in simulation."},{"cited_title":"A simple formula for the average gate fidelity of a quantum dynamical operation","cited_arxiv_id":null,"evidence_quote":"Defines the average gate fidelity metric used to score each CNOT gate."},{"cited_title":"A simplex method for function minimization","cited_arxiv_id":null,"evidence_quote":"Supplies the derivative-free simplex optimization routine used to minimize infidelity over the pulse parameters."}],"review_version":1}