Abstract Pulse-level calibration metrics — coherence times T₁ and T₂ measured through single-qubit Rabi and Ramsey sequences — are the standard by which quantum hardware providers certify device health. We demonstrate, through 191 circuit executions on three 156-qubit IBM Heron r2 processors (ibm_kingston, ibm_fez, and ibm_marrakesh), that these pulse-level metrics can give opposite conclusions to circuit-level diagnostics on the same device at the same time. Specifically, between experimental sessions on July 3 and July 5, 2026, IBM reported ibm_marrakesh T₂ improved from 303 µs to 340 µs (a 12% improvement), while the K–R scaling exponent α measured from GHZ circuit scaling increased from 1.406 to 3.811 (a 171% increase in coherent error dominance) and GHZ n = 2 infidelity rose from 0.018 to 0.204 — a 920% increase in circuit error. We further demonstrate that Ramsey circuit-context T₂ on qubit 0 is approximately 9–30 µs, compared to IBM-reported pulse-level T₂ of 32–340 µs — a discrepancy of 3–11×. We introduce five experimental demonstrations showing K–R circuit-level diagnostics detect degradation invisible to T₁/T₂, provide higher temporal sensitivity, enable algorithm failure prediction, and confirm the theoretical prediction αₜ v ₜᶜʰ + αₑₜₕₒ = 2 within 7% on real hardware. These findings establish circuit-level K–R diagnostics as a necessary complement to pulse-level calibration for operational quantum computing.
Pulse-level calibration metrics — coherence times T₁ and T₂ measured through single-qubit Rabi and Ramsey sequences — are the standard by which quantum hardware providers certify device health. We demonstrate, through 191 circuit executions on three 156-qubit IBM Heron r2 processors (ibm_kingston, ibm_fez, and ibm_marrakesh), that these pulse-level metrics can give opposite conclusions to circuit-level diagnostics on the same device at the same time. Specifically, between experimental sessions on July 3 and July 5, 2026, IBM reported ibm_marrakesh T₂ improved from 303 μs to 340 μs (a 12% improvement), while the K–R scaling exponent α measured from GHZ circuit scaling increased from 1.406 to 3.811 (a 171% increase in coherent error dominance) and GHZ n=2 infidelity rose from 0.018 to 0.204 — a 920% increase in circuit error. We further demonstrate that Ramsey circuit-context T₂ on qubit 0 is approximately 9–30 μs, compared to IBM-reported pulse-level T₂ of 32–340 μs — a discrepancy of 3–11×. We introduce five experimental demonstrations showing K–R circuit-level diagnostics detect degradation invisible to T₁/T₂, provide higher temporal sensitivity, enable algorithm failure prediction, and confirm the theoretical prediction αₜᵥₜᶜʰ + αₑₜₕₒ = 2 within 7% on real hardware. These findings establish circuit-level K–R diagnostics as a necessary complement to pulse-level calibration for operational quantum computing.
Abstract We introduce Quantum Diagnostic Intelligence (QDI), a five-layer decision-support architecture for inferring the operational state of quantum processing units from circuit-level measurements. Pulse-level calibration metrics — coherence times T₁ and T₂ measured through single-qubit Rabi and Ramsey sequences — are the dominant industry standard for certifying quantum hardware health. We demonstrate, through 191 circuit executions on three 156-qubit IBM Heron r2 processors, that circuit-level measurements contain diagnostic information that T₁/T₂ may not capture. On ibm_marrakesh between July 3 and July 5, 2026, IBM-reported T₂ improved from 303 to 340 µs (+ 12%), while circuit-level GHZ scaling yielded α = 3.811, a QDI health score of 37.8/100 [95% CI: 7.6, 67.9], and a circuit infidelity increase of 920% at n = 2 qubits. The QDI architecture formalizes this observation through four operators: a diagnostic state classifier D(χ), a health score H(χ), a recommendation generator R(χ), and a drift forecast F(χ, t). Rules trained on ibm_kingston correctly classified ibm_fez and both ibm_marrakesh sessions without retuning (3/3 held-out tests; n = 4 total — feasibility demonstration only). We identify three scientific limitations of the current framework: operators rely on expert-designed rules rather than learned statistical models; ground truth (algorithm success/failure) has not yet been collected; and diagnostic state labels have not been validated by clustering analysis. Prospective validation on n ≥ 50 independent sessions is required before statistical claims about generalization can be made. The dataset, code, and QDI system are openly available.
Numerical optimal control (GRAPE) can in principle discover pulse shapes that suppress all coherent gate error to machine precision. But when does that capability actually matter? We present a systematic comparison of Gaussian, DRAG, and GRAPE pulses for single-qubit gates on a three-level transmon model parameterized by IQM Garnet hardware (
T1=37μs,
T2=9.6μs,
α/2π=−200 MHz), with the explicit goal of identifying the regimes where numerical optimization provides genuine practical advantage over analytical methods.
Our central finding is that properly calibrated DRAG already operates near the decoherence floor. At 20 ns gate time, GRAPE eliminates all coherent error (
1−F<10−15), but DRAG achieves
1−F=4.9×10−4 in coherent error alone,and
8.4×10−4 under full decoherence -- only
1.2× above GRAPE's decoherence-limited performance. More surprisingly,DRAG is \emph{more robust} than GRAPE to qubit frequency detuning (minimum fidelity 0.990 vs.\ 0.931 over
±5 MHz), the dominant calibration uncertainty in charge-noise-limited transmons. GRAPE retains superior amplitude robustness (minimum fidelity 0.994 vs.\ 0.990) and provides the only route to guaranteed zero coherent error, which matters at short gate times (
≲15 ns) where perturbative corrections break down.
These results lead to concrete calibration guidance: (1) properly calibrated DRAG is sufficient for gate times
≳20 ns on hardware with
T2/T≳500, (2) GRAPE is necessary at short gate times or when targeting error rates well below the decoherence floor, and (3) robust optimal control incorporating frequency uncertainty should be used when detuning is the dominant noise source. We decompose the full error budget (coherent,
T1,
T2, control noise) and provide the open-source QubitPulseOpt framework for reproducing all results.