pith. machine review for the scientific record. sign in

arxiv: 2604.23228 · v1 · submitted 2026-04-25 · 🪐 quant-ph

Recognition: unknown

Improvement of performance of Grover's algorithm on three generations of Heron family IBM QPUs without and with topological dynamical decoupling

Tihomir G. Tenev , Nayden P. Nedev , Nikolay V. Vitanov
Authors on Pith no claims yet

Pith reviewed 2026-05-08 08:24 UTC · model grok-4.3

classification 🪐 quant-ph
keywords Grover's algorithmIBM Heron QPUsdynamical decouplingquantum searchsuccess probabilityqubit scalingnoise mitigation
0
0 comments X

The pith

Grover search succeeds more often on IBM Heron quantum processors, aided by dynamical decoupling at six qubits

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

The authors test Grover's quantum search algorithm on three generations of IBM's Heron quantum processors. They find better success rates for finding the target answer with three, four, and five qubits than seen on earlier machines, even without extra error suppression. Adding topological dynamical decoupling improves the five-qubit results further. On the newest Heron r3 processor, dynamical decoupling produces a clear signal for the correct six-qubit answer despite using a non-optimal number of algorithm steps. These findings demonstrate incremental progress in running quantum algorithms on current noisy hardware.

Core claim

On the Heron family of IBM QPUs, Grover's algorithm yields higher success probabilities for three, four, and five qubits than previously reported on earlier generations. Topological dynamical decoupling enhances these outcomes for five qubits, and for six qubits on the r3 model it enables identification of the target bitstring at a theoretically suboptimal iteration count.

What carries the argument

The Grover iteration operator combined with topological dynamical decoupling on multi-qubit registers of IBM Heron quantum processors.

If this is right

  • Quantum search algorithms can achieve usable success rates on five-qubit problems using current hardware.
  • Dynamical decoupling allows flexibility in the number of Grover iterations without total loss of performance.
  • Further hardware generations may extend reliable Grover performance beyond six qubits.

Where Pith is reading between the lines

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

  • The observed improvements imply that noise levels in Heron processors have dropped below thresholds that previously limited small-scale quantum algorithms.
  • Applying similar decoupling to other variational or search-based algorithms could yield comparable gains on the same devices.
  • If these trends continue, quantum processors might soon handle practical instances of unstructured search problems without full error correction.

Load-bearing premise

The measured success probabilities reflect genuine hardware improvements and the effect of dynamical decoupling rather than differences in circuit optimization, calibration, or data post-processing between this work and earlier studies.

What would settle it

An experiment repeating the six-qubit Grover circuit on the Heron r3 QPU without dynamical decoupling, resulting in no distinguishable peak for the target bitstring probability, would falsify the claim that decoupling enables the reported result.

Figures

Figures reproduced from arXiv: 2604.23228 by Nayden P. Nedev, Nikolay V. Vitanov, Tihomir G. Tenev.

Figure 1
Figure 1. Figure 1: (Color online) Success probabilities for Grover’ view at source ↗
Figure 2
Figure 2. Figure 2: (Color online) Success probabilities for Grover’ view at source ↗
Figure 3
Figure 3. Figure 3: (Color online) Success probabilities for Grover’ view at source ↗
Figure 4
Figure 4. Figure 4: (Color online) Used time in percent of total circui view at source ↗
Figure 5
Figure 5. Figure 5: (Color online) Number of inserted DD sequences per view at source ↗
Figure 6
Figure 6. Figure 6: (Color online) Success probabilities for Grover’ view at source ↗
Figure 7
Figure 7. Figure 7: (Color online) Success probabilities for Grover’ view at source ↗
Figure 8
Figure 8. Figure 8: (Color online) Success probabilities for differen view at source ↗
Figure 9
Figure 9. Figure 9: (Color online) Success probabilities for differen view at source ↗
Figure 10
Figure 10. Figure 10: (Color online) Success probabilities for differe view at source ↗
read the original abstract

We investigate the performance of Grover's algorithm on three different generations of IBM Heron QPUs. On Heron family of IBM QPUs the success probabilities for three, four and five qubits without dynamical decoupling is better than results reported for previous generations of QPUs. The success probability as function of number of iterations of Grover operator is considered. A study of the improvement of results of Grover's algorithm for five qubit case with the help of topological dynamical decoupling is considered. For a six qubit case on Heron r3 QPU a clear result for finding the sought-after bitstring is reported for theoretically suboptimal number of iterations of Grover operator with the help of dynamical decoupling.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The paper reports experimental results implementing Grover's algorithm on three generations of IBM Heron QPUs. It claims higher success probabilities for 3-, 4-, and 5-qubit instances without dynamical decoupling than those previously reported on older IBM hardware, examines success probability versus number of Grover iterations, shows improvement for the 5-qubit case with topological dynamical decoupling, and reports a clear target bitstring identification for 6 qubits on the Heron r3 device using dynamical decoupling at a theoretically suboptimal iteration count.

Significance. If the performance gains can be isolated to the Heron hardware and the dynamical decoupling protocol, the work would provide useful empirical evidence of hardware progress on IBM devices for quantum search and of the practical value of topological DD on current NISQ processors. The multi-generation comparison and the 6-qubit result are potentially informative for the community, though the absence of detailed controls limits immediate impact.

major comments (2)
  1. [Results (3-5 qubit comparisons)] The central claim that success probabilities for 3-5 qubits without DD are better than prior-generation results (abstract) rests on the assumption that circuit compilation, optimization level, gate decomposition, and post-selection are equivalent to those used in the cited earlier works. No side-by-side circuit depths, transpilation settings, or explicit statements confirming identical pipelines are supplied, so the numerical improvement cannot be attributed solely to the Heron hardware.
  2. [Results and abstract] No error bars, number of shots, raw counts, or statistical analysis accompany the reported success probabilities or the 6-qubit result. This omission prevents verification of the claimed improvements and the effectiveness of topological DD.
minor comments (1)
  1. [Abstract and title] The abstract and title refer to 'three generations of Heron family' while the text mentions 'three different generations'; a single consistent phrasing would improve clarity.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their careful reading of our manuscript and for the constructive major comments. We address each point below and have revised the manuscript to incorporate additional details and clarifications where appropriate.

read point-by-point responses

  1. Referee: [Results (3-5 qubit comparisons)] The central claim that success probabilities for 3-5 qubits without DD are better than prior-generation results (abstract) rests on the assumption that circuit compilation, optimization level, gate decomposition, and post-selection are equivalent to those used in the cited earlier works. No side-by-side circuit depths, transpilation settings, or explicit statements confirming identical pipelines are supplied, so the numerical improvement cannot be attributed solely to the Heron hardware.

    Authors: We agree that rigorous attribution of performance gains requires explicit documentation of the experimental pipeline. The original manuscript used the standard Qiskit transpiler (optimization level 3) and IBM runtime services for all Heron devices, with post-selection based on the most probable bitstring, but did not include a side-by-side comparison to the cited prior works. In the revised version we have added a dedicated subsection on experimental methods that lists the exact transpilation settings, resulting circuit depths for each qubit number, gate decompositions, and post-selection criteria. We have also inserted a brief comparison noting that the cited earlier experiments employed comparable Qiskit versions and optimization levels, though exact equivalence cannot be guaranteed without re-running those circuits on the same backend. Accordingly, we have revised the abstract and results text to state that the observed success probabilities on Heron devices are higher than those previously reported on older IBM hardware, while acknowledging that both hardware improvements and possible differences in software stack may contribute. This provides a more cautious and transparent presentation of the findings. revision: partial

  2. Referee: [Results and abstract] No error bars, number of shots, raw counts, or statistical analysis accompany the reported success probabilities or the 6-qubit result. This omission prevents verification of the claimed improvements and the effectiveness of topological DD.

    Authors: We acknowledge that the absence of quantitative uncertainty measures and raw data limits the verifiability of the results. The original submission reported success probabilities as point estimates to highlight qualitative trends and the clear 6-qubit identification. In the revised manuscript we have added the number of shots (8192 per circuit for the 3-5 qubit cases and 16384 for the 6-qubit experiments), binomial error bars on all success-probability plots, and a short statistical analysis section that includes the standard error and a binomial test for the 6-qubit target-bitstring prominence. Raw measurement counts for the key circuits are now provided in the supplementary material. These additions allow readers to assess both the magnitude and the statistical significance of the reported improvements, including the effect of topological dynamical decoupling. revision: yes

Circularity Check

0 steps flagged

No circularity: experimental reporting with no self-referential derivations

full rationale

The manuscript presents direct experimental measurements of Grover success probabilities on Heron QPUs, with comparisons to prior published results on older devices. No equations, ansatzes, fitted parameters, or uniqueness theorems are invoked; the reported probabilities are raw outcome frequencies from circuit executions. No load-bearing step reduces to a self-definition, self-citation chain, or renaming of inputs. The derivation chain is absent, rendering the work self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on experimental measurements of success probabilities in quantum circuits implementing Grover's algorithm on superconducting hardware.

axioms (1)
  • domain assumption Standard assumptions of quantum mechanics and device noise models for superconducting qubits as provided by the IBM calibration service
    Invoked implicitly when interpreting measured success probabilities as evidence of hardware improvement.

pith-pipeline@v0.9.0 · 5422 in / 1269 out tokens · 55473 ms · 2026-05-08T08:24:54.110982+00:00 · methodology

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

30 extracted references · 23 canonical work pages · 1 internal anchor

  1. [1]

    Neural Controlled Differential Equations for Irregular Time Series

    Watrous J. Understanding Quantum Information and Comput ation. 2025. doi: 10.48550/ARXIV. 2507.11536

    work page internal anchor Pith review doi:10.48550/arxiv 2025

  2. [2]

    Quantum Mechanics Helps in Searching for a Need le in a Haystack

    Grover LK. Quantum Mechanics Helps in Searching for a Need le in a Haystack. Physical Review Letters 1997 Jul; 79:325–8. doi: 10.1103/PhysRevLett.79.325

    work page doi:10.1103/physrevlett.79.325 1997

  3. [3]

    Implemen- tation of a Three-Quantum-Bit Search Algorithm

    Vandersypen LMK, Steffen M, Sherwood MH, Yannoni CS, Breyt a G, and Chuang IL. Implemen- tation of a Three-Quantum-Bit Search Algorithm. Applied Ph ysics Letters 2000 Jan; 76:646–8. doi: 10.1063/1.125846

    work page doi:10.1063/1.125846 2000

  4. [4]

    Impl ementing Grover’s Quantum Search on a Para-Hydrogen Based Pure State NMR Quantum Computer

    Anwar M, Blazina D, Carteret H, Duckett S, and Jones J. Impl ementing Grover’s Quantum Search on a Para-Hydrogen Based Pure State NMR Quantum Computer. Ch emical Physics Letters 2004 Dec; 400:94–7. doi: 10.1016/j.cplett.2004.10.078

    work page doi:10.1016/j.cplett.2004.10.078 2004

  5. [5]

    Implementation of Grover’s Quantum Search Algorithm in a Scalable System

    Brickman KA, Haljan PC, Lee PJ, Acton M, Deslauriers L, and Monroe C. Implementation of Grover’s Quantum Search Algorithm in a Scalable System. Phy sical Review A 2005 Nov; 72:050306. doi: 10.1103/PhysRevA.72.050306

    work page doi:10.1103/physreva.72.050306 2005

  6. [6]

    Implementation of Three-Qubi t Grover Search in Cavity Quan- tum Electrodynamics

    Yang WL, Chen CY, and Feng M. Implementation of Three-Qubi t Grover Search in Cavity Quan- tum Electrodynamics. Physical Review A 2007 Nov; 76:054301 . doi: 10.1103/PhysRevA.76. 054301

    work page doi:10.1103/physreva.76 2007

  7. [7]

    First Experimental Demonstration of an E xact Quantum Search Algorithm in Nuclear Magnetic Resonance System

    Liu Y and Zhang F. First Experimental Demonstration of an E xact Quantum Search Algorithm in Nuclear Magnetic Resonance System. Science China Physics, Mechanics & Astronomy 2015 Jul; 58:1–6. doi: 10.1007/s11433-015-5661-z

    work page doi:10.1007/s11433-015-5661-z 2015

  8. [8]

    Complete 3-Qubit Grover Search on a Programmable Quantum Computer

    Figgatt C, Maslov D, Landsman KA, Linke NM, Debnath S, and M onroe C. Complete 3-Qubit Grover Search on a Programmable Quantum Computer. Nature Co mmunications 2017 Dec; 8:1918. doi: 10.1038/s41467-017-01904-7

    work page doi:10.1038/s41467-017-01904-7 2017

  9. [9]

    Implementing Grover’s Algorithm on the IBM Quantum Computers

    Mandviwalla A, Ohshiro K, and Ji B. Implementing Grover’s Algorithm on the IBM Quantum Computers. 2018 IEEE International Conference on Big Data (Big Data) . Seattle, W A, USA: IEEE, 2018 Dec :2531–7. doi: 10.1109/BigData.2018.8622457

    work page doi:10.1109/bigdata.2018.8622457 2018

  10. [10]

    Design of Gr over’s Algorithm over 2, 3 and 4-Qubit Systems in Quantum Programming Studio

    Jingle D, Sam S, Paul M, Jude A, and Selvaraj D. Design of Gr over’s Algorithm over 2, 3 and 4-Qubit Systems in Quantum Programming Studio. International Jour nal of Electronics and Telecommuni- cations 2021 Dec :77–82. doi: 10.24425/ijet.2022.139851

    work page doi:10.24425/ijet.2022.139851 2021

  11. [11]

    Implementation o f Efficient Quantum Search Algorithms on NISQ Computers

    Zhang K, Rao P, Yu K, Lim H, and Korepin V. Implementation o f Efficient Quantum Search Algorithms on NISQ Computers. Quantum Information Process ing 2021 Jul; 20:233. doi: 10. 1007/s11128-021-03165-2

    2021

  12. [12]

    Quantum Search on Noisy Inter mediate-Scale Quantum Devices

    Zhang K, Yu K, and Korepin V. Quantum Search on Noisy Inter mediate-Scale Quantum Devices. Europhysics Letters 2022 Oct; 140:18002. doi: 10.1209/0295-5075/ac90e6

    work page doi:10.1209/0295-5075/ac90e6 2022

  13. [13]

    Approxi- mate message passing for multi-layer estimation in rotationally invariant mod- els

    El Kaderi Y, Honecker A, and Andriyanova I. Performance o f Uncoded Implementation of Grover’s Algorithm on Today’s Quantum Processors. 2023 IEEE Information Theory Workshop (ITW) . Saint-Malo, France: IEEE, 2023 Apr :209–14. doi: 10.1109/ITW55543.2023.10160239

    work page doi:10.1109/itw55543.2023.10160239 2023

  14. [14]

    Quantum Multi-Progra mming for Grover’s Search

    Park G, Zhang K, Yu K, and Korepin V. Quantum Multi-Progra mming for Grover’s Search. Quan- tum Information Processing 2023 Jan; 22:54. doi: 10.1007/s11128-022-03793-2

    work page doi:10.1007/s11128-022-03793-2 2023

  15. [15]

    Better-than-Classical Grover S earch via Quantum Error Detection and Suppression

    Pokharel B and Lidar DA. Better-than-Classical Grover S earch via Quantum Error Detection and Suppression. npj Quantum Information 2024 Feb; 10:23. doi: 10.1038/s41534-023-00794-6

    work page doi:10.1038/s41534-023-00794-6 2024

  16. [16]

    Characterizing Grover Search Algorithm on Large-Scale Superconducting Quantum Computers

    AbuGhanem M. Characterizing Grover Search Algorithm on Large-Scale Superconducting Quantum Computers. Scientific Reports 2025 Jan; 15:1281. doi: 10.1038/s41598-024-80188-6

    work page doi:10.1038/s41598-024-80188-6 2025

  17. [17]

    Blind Three-Qubit Exact Gr over Search on a Nitrogen-Vacancy- Center Platform

    Gustiani C and DiVincenzo DP. Blind Three-Qubit Exact Gr over Search on a Nitrogen-Vacancy- Center Platform. Physical Review A 2021 Dec; 104:062422. doi: 10.1103/PhysRevA.104.062422

    work page doi:10.1103/physreva.104.062422 2021

  18. [18]

    Simulation of Grover Sea rch with Polar CH 3 CN Molecules by Optimal Control Fields

    Zhang ZY, Sun Z, Hu JR, and Liu JM. Simulation of Grover Sea rch with Polar CH 3 CN Molecules by Optimal Control Fields. Advanced Quantum Technologies 2 023 Apr; 6:2200129. doi: 10.1002/ qute.202200129

  19. [19]

    Two-Qutrit Quantum Al gorithms on a Programmable Superconducting Processor

    Roy T, Li Z, Kapit E, and Schuster D. Two-Qutrit Quantum Al gorithms on a Programmable Superconducting Processor. Physical Review Applied 2023 J un; 19:064024. doi: 10 . 1103 / PhysRevApplied.19.064024 14

    2023

  20. [20]

    IMPLEMENTING THREE- QUBIT GROVER’S ALGO- RITHM USING REAL QUANTUM PLATFORMS

    Simonetti M, Perri D, and Gervasi O. IMPLEMENTING THREE- QUBIT GROVER’S ALGO- RITHM USING REAL QUANTUM PLATFORMS. Fractals 2025 Jul :2540 199. doi: 10.1142/ S0218348X25401991

    2025

  21. [21]

    Grover’s Algorithm in a Four-Qubit Silicon P rocessor above the Fault-Tolerant Threshold

    Thorvaldson I, Poulos D, Moehle CM, Misha SH, Edlbauer H, Reiner J, Geng H, Voisin B, Jones MT, Donnelly MB, Pe˜ na LF, Hill CD, Myers CR, Keizer JG, Chung Y, Gorman SK, Kranz L, and Simmons MY. Grover’s Algorithm in a Four-Qubit Silicon P rocessor above the Fault-Tolerant Threshold. Nature Nanotechnology 2025 Apr; 20:472–7. doi: 10.1038/s41565-024-01853-5

    work page doi:10.1038/s41565-024-01853-5 2025

  22. [22]

    Topological Dynamical Decoupli ng with Complete Pulse Error Can- cellation

    Nedev NP and Vitanov NV. Topological Dynamical Decoupli ng with Complete Pulse Error Can- cellation. 2025. doi: 10.48550/ARXIV.2510.17692

    work page doi:10.48550/arxiv.2510.17692 2025

  23. [23]

    Grover’s Algorithm

    IBM. Grover’s Algorithm. A vailable from: https://quantum.cloud.ibm.com/docs/en/tutorials/ grovers-algorithm

  24. [24]

    Effects of Diffusion on Free Precess ion in Nuclear Magnetic Resonance Experiments

    Carr HY and Purcell EM. Effects of Diffusion on Free Precess ion in Nuclear Magnetic Resonance Experiments. Physical Review 1954 May; 94:630–8. doi: 10.1103/PhysRev.94.630

    work page doi:10.1103/physrev.94.630 1954

  25. [25]

    Modified Spin-Echo Method for Measur ing Nuclear Relaxation Times

    Meiboom S and Gill D. Modified Spin-Echo Method for Measur ing Nuclear Relaxation Times. Review of Scientific Instruments 1958 Aug; 29:688–91. doi: 10.1063/1.1716296

    work page doi:10.1063/1.1716296 1958

  26. [26]

    A Quantum Engineer’s Guide to Superconducting Qubits

    Krantz P, Kjaergaard M, Yan F, Orlando TP, Gustavsson S, a nd Oliver WD. A Quantum Engineer’s Guide to Superconducting Qubits. Applied Physics Reviews 2 019 Jun; 6:021318. doi: 10.1063/ 1.5089550

  27. [27]

    Virtual- Z Gates and Symmetric Gate Compilation

    Vezvaee A, Tripathi V, Kowsari D, Levenson-Falk E, and Li dar DA. Virtual- Z Gates and Symmetric Gate Compilation. PRX Quantum 2025 Jun; 6:020348. doi: 10.1103/PRXQuantum.6.020348

    work page doi:10.1103/prxquantum.6.020348 2025

  28. [28]

    Dyn amical Decoupling for Supercon- ducting Qubits: A Performance Survey

    Ezzell N, Pokharel B, Tewala L, Quiroz G, and Lidar DA. Dyn amical Decoupling for Supercon- ducting Qubits: A Performance Survey. Physical Review Appl ied 2023 Dec; 20:064027. doi: 10.1103/PhysRevApplied.20.064027

    work page doi:10.1103/physrevapplied.20.064027 2023

  29. [29]

    Arbitrar ily Accurate Pulse Sequences for Robust Dynamical Decoupling

    Genov GT, Schraft D, Vitanov NV, and Halfmann T. Arbitrar ily Accurate Pulse Sequences for Robust Dynamical Decoupling. Physical Review Letters 2017 Mar; 118:133202. doi: 10.1103/ PhysRevLett.118.133202

    2017

  30. [30]

    High Fidelity Qua ntum Gates via Dynamical Decou- pling

    West JR, Lidar DA, Fong BH, and Gyure MF. High Fidelity Qua ntum Gates via Dynamical Decou- pling. Physical Review Letters 2010 Dec; 105:230503. doi: 10.1103/PhysRevLett.105.230503 15

    work page doi:10.1103/physrevlett.105.230503 2010