pith. sign in

arxiv: 2606.31379 · v1 · pith:HVPMZRUKnew · submitted 2026-06-30 · ⚛️ physics.comp-ph

P3MaZe: a Mass-Zero constrained-dynamics formulation of particle-mesh electrostatics

Federica Troni , Violette Gontran , Davide Grassano , Sara Bonella This is my paper

Pith reviewed 2026-07-01 03:00 UTC · model grok-4.3

classification ⚛️ physics.comp-ph
keywords particle-mesh electrostaticsconstrained dynamicsP3MMaZemolecular dynamicsPoisson equationmultigrid solverholonomic constraint
0
0 comments X

The pith

P3MaZe replaces the multigrid Poisson solver in particle-mesh electrostatics with a mass-zero constrained dynamics formulation that enforces the discretized Poisson equation as a holonomic constraint.

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

The paper presents P3MaZe as a combination of the standard particle-particle particle-mesh electrostatics decomposition with the mass-zero constrained dynamics framework. The long-range electrostatic potential is represented as a zero-inertia auxiliary field on a mesh, and the discretized Poisson equation is imposed as a holonomic constraint during the dynamics. This substitution preserves the usual accuracy controls from the real-space cutoff, Ewald splitting parameter, mesh spacing, and charge assignment scheme. Tests on molten NaCl and flexible point-charge water show that structural properties, translational and rotational dynamics, and collective observables match those obtained with real-space P3M and Ewald summation. The constrained formulation requires fewer multigrid iterations than the corresponding unconstrained solver while preserving linear scaling with system size.

Core claim

P3MaZe combines the short-range/long-range decomposition of P3M electrostatics with the MaZe framework by representing the smooth long-range electrostatic potential on a mesh as a zero-inertia auxiliary field and enforcing the discretized Poisson equation as a holonomic constraint. This replaces the conventional multigrid Poisson solver by a constrained correction problem. The formulation retains the systematic accuracy controls associated with the real-space cutoff, the Ewald splitting, the mesh spacing, and the charge-assignment procedure, produces observables in quantitative agreement with established methods, and requires fewer multigrid iterations while retaining linear scaling with sys

What carries the argument

Mass-zero constrained dynamics (MaZe) applied to the long-range electrostatic potential, with the discretized Poisson equation enforced as a holonomic constraint.

If this is right

  • Structural, translational, collective, and rotational dynamical observables remain in quantitative agreement with those from real-space P3M and Ewald summation.
  • The method requires fewer multigrid iterations than the corresponding real-space P3M solver.
  • Linear scaling with system size is retained.
  • All standard accuracy controls from the P3M decomposition remain available for systematic convergence checks.

Where Pith is reading between the lines

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

  • The constrained formulation could be combined with other auxiliary-field constraints already present in a simulation without changing the integrator structure.
  • Because the Poisson equation becomes part of the constraint manifold, the method may allow direct control over electrostatic energy contributions during sampling.
  • The reduction in multigrid iterations suggests the approach could be particularly advantageous when electrostatics dominate the cost in very large periodic systems.

Load-bearing premise

Enforcing the discretized Poisson equation as a holonomic constraint in the MaZe framework produces dynamics equivalent to the unconstrained P3M solution without introducing numerical artifacts or altering physical observables.

What would settle it

A direct side-by-side simulation of molten NaCl showing a measurable difference in radial distribution functions or diffusion coefficients between P3MaZe and standard real-space P3M at identical cutoff, splitting, and mesh parameters would falsify the claimed equivalence.

Figures

Figures reproduced from arXiv: 2606.31379 by Davide Grassano, Federica Troni, Sara Bonella, Violette Gontran.

Figure 1
Figure 1. Figure 1: Partial radial distribution functions of molten NaCl obtained with Poisson MaZe and P3MaZe. [PITH_FULL_IMAGE:figures/full_fig_p008_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Comparison of single-particle and collective dynamical observables for molten NaCl obtained with [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Comparison of oxygen–oxygen, oxygen–hydrogen and hydrogen–hydrogen radial distribution func [PITH_FULL_IMAGE:figures/full_fig_p011_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Comparison of single-particle and collective dynamical observables for SPC/Fw water obtained [PITH_FULL_IMAGE:figures/full_fig_p011_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Comparison of dipole rotational relaxation in SPC/Fw water obtained with P3M and P3MaZe. [PITH_FULL_IMAGE:figures/full_fig_p012_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Residual convergence of the long-range electrostatic solver for [PITH_FULL_IMAGE:figures/full_fig_p014_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Average number of multigrid iterations per MD step for [PITH_FULL_IMAGE:figures/full_fig_p014_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Electrostatic CPU time per MD step as a function of system size shown on logarithmic axes. Linear [PITH_FULL_IMAGE:figures/full_fig_p015_8.png] view at source ↗
read the original abstract

We introduce P3MaZe, a real-space particle-mesh electrostatic method that combines the standard short-range/long-range decomposition of Particle-Particle Particle-Mesh (P3M) electrostatics with the Mass-Zero constrained dynamics (MaZe) framework. In this formulation, the smooth long-range electrostatic potential is represented on a mesh as a zero-inertia auxiliary field, while the discretized Poisson equation is enforced as a holonomic constraint during molecular dynamics. By retaining the standard P3M decomposition, P3MaZe preserves the systematic accuracy controls associated with the real-space cutoff, the Ewald splitting, the mesh spacing, and the charge-assignment procedure, while replacing the conventional multigrid Poisson solver by a constrained correction problem. The method is validated for molten NaCl and simple point-charge flexible water (SPC/Fw). Structural, translational, collective, and rotational dynamical observables are in quantitative agreement with those obtained with established electrostatic methods, including real-space P3M, and Ewald summation. The constrained formulation consistently requires fewer multigrid iterations than the corresponding real-space P3M solver while retaining the expected linear scaling with system size. These results establish P3MaZe as a promising new direction for scalable real-space electrostatics in large-scale molecular simulations.

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 / 2 minor

Summary. The manuscript introduces P3MaZe, which augments the standard short-range/long-range decomposition of real-space P3M electrostatics with the Mass-Zero (MaZe) constrained-dynamics framework. The long-range potential is represented as a zero-inertia auxiliary field on the mesh, and the discretized Poisson equation is imposed as a holonomic constraint. The approach retains the usual P3M accuracy controls (real-space cutoff, Ewald parameter, mesh spacing, charge assignment) while replacing the conventional multigrid Poisson solve with a constrained correction step. Validation on molten NaCl and SPC/Fw reports quantitative agreement with reference real-space P3M and Ewald results for structural, translational, collective, and rotational observables, together with a reduction in multigrid iterations and retention of linear scaling with system size.

Significance. If the dynamical equivalence to unconstrained P3M is confirmed without introducing artifacts, the formulation offers a conceptually distinct route to real-space electrostatics that may reduce solver cost while preserving established accuracy parameters. The explicit retention of P3M controls and the reported linear scaling are positive features; the iteration reduction, if reproducible, would be of practical interest for large-scale MD. The work is grounded in an existing constrained-dynamics framework rather than introducing ad-hoc parameters.

major comments (2)
  1. [Abstract] Abstract: the central performance claim that the constrained formulation 'consistently requires fewer multigrid iterations' is stated without supporting data (iteration counts, tolerances, or scaling plots). Because this is the primary advertised advantage over standard real-space P3M, the absence of quantitative comparison undermines assessment of the claim.
  2. [Validation] Validation (abstract and results): quantitative agreement is asserted for structural, translational, collective, and rotational observables, yet no error bars, statistical uncertainties, RMSD values, or convergence diagnostics are supplied. This information is required to evaluate whether the holonomic constraint introduces measurable deviations from the unconstrained P3M reference.
minor comments (2)
  1. [Introduction] The phrase 'zero-inertia auxiliary field' is used without an immediate definition or reference to its MaZe origin; a one-sentence clarification in the introduction would improve accessibility.
  2. [Method] Notation for the charge-assignment procedure and the Ewald splitting parameter should be aligned explicitly with standard P3M literature to facilitate direct comparison.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the positive assessment and recommendation for minor revision. We address the two major comments below and will revise the manuscript accordingly to strengthen the presentation of quantitative evidence.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the central performance claim that the constrained formulation 'consistently requires fewer multigrid iterations' is stated without supporting data (iteration counts, tolerances, or scaling plots). Because this is the primary advertised advantage over standard real-space P3M, the absence of quantitative comparison undermines assessment of the claim.

    Authors: We agree that the abstract would benefit from explicit reference to the supporting data. The results section already contains direct comparisons of multigrid iteration counts (with fixed tolerances) between P3MaZe and standard real-space P3M for both NaCl and SPC/Fw systems, together with a demonstration of linear scaling. In the revised manuscript we will add a short clause in the abstract pointing to these quantitative results and include a compact table of average iteration counts in the main text or as a new figure panel. revision: yes

  2. Referee: [Validation] Validation (abstract and results): quantitative agreement is asserted for structural, translational, collective, and rotational observables, yet no error bars, statistical uncertainties, RMSD values, or convergence diagnostics are supplied. This information is required to evaluate whether the holonomic constraint introduces measurable deviations from the unconstrained P3M reference.

    Authors: We acknowledge that the current validation would be strengthened by explicit uncertainty estimates. The manuscript already reports that observables match reference P3M and Ewald results to within visual agreement on the plotted scales, but we will add statistical error bars (obtained via block averaging over independent trajectories) to all key figures and report RMSD or mean absolute deviations between P3MaZe and reference P3M runs in a new table. This will allow readers to confirm that any differences lie within statistical uncertainty. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation self-contained

full rationale

The paper introduces P3MaZe by combining the standard P3M short/long-range decomposition with the MaZe constrained-dynamics framework, enforcing the discretized Poisson equation as a holonomic constraint. All accuracy controls (cutoff, Ewald splitting, mesh spacing, charge assignment) are retained from the established P3M method. Validation against independent external references (real-space P3M and Ewald summation) shows quantitative agreement on structural, dynamical, and collective observables for molten NaCl and SPC/Fw, with the reported reduction in multigrid iterations presented as an empirical outcome of the simulations rather than a quantity forced by definition or by a self-citation chain. No load-bearing step reduces to a fitted input renamed as prediction or to an ansatz smuggled via prior self-work; the central claims remain falsifiable against the cited benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 1 invented entities

Only the abstract is available; the ledger is therefore minimal and provisional. The method inherits standard P3M parameters and introduces one new representational device.

axioms (1)
  • domain assumption The discretized Poisson equation can be enforced as a holonomic constraint without changing the underlying electrostatic physics.
    Invoked when the long-range potential is recast as a zero-inertia auxiliary field under constraint.
invented entities (1)
  • zero-inertia auxiliary field no independent evidence
    purpose: Representation of the smooth long-range electrostatic potential on the mesh
    Introduced to enable the constrained-dynamics formulation

pith-pipeline@v0.9.1-grok · 5767 in / 1237 out tokens · 58535 ms · 2026-07-01T03:00:03.418043+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

40 extracted references · 4 canonical work pages

  1. [1]

    Ryckaert and A

    J-P. Ryckaert and A. Bellemans and G. Ciccotti and , title =. Molecular Physics , volume =. 1981 , publisher =

    1981

  2. [2]

    Performance Analysis of Parallel FFT on Large Multi-GPU Systems , year=

    Ayala, Alan and Tomov, Stan and Stoyanov, Miroslav and Haidar, Azzam and Dongarra, Jack , booktitle=. Performance Analysis of Parallel FFT on Large Multi-GPU Systems , year=

  3. [3]

    Accelerating MPI All-to-All Communication with Online Compression on Modern GPU Clusters

    Zhou, Qinghua and Kousha, Pouya and Anthony, Quentin and Shafie Khorassani, Kawthar and Shafi, Aamir and Subramoni, Hari and Panda, Dhabaleswar K. Accelerating MPI All-to-All Communication with Online Compression on Modern GPU Clusters. High Performance Computing. 2022

    2022

  4. [4]

    The Journal of Chemical Physics , volume =

    Sagui, Celeste and Darden, Thomas , title =. The Journal of Chemical Physics , volume =. 2001 , month =

    2001

  5. [5]

    and Maggs, A

    Rossetto, V. and Maggs, A. C. , title =. Physical Review Letters , volume =

  6. [6]

    Physical Review Letters , volume =

    Local Molecular Dynamics with Coulombic Interactions , author =. Physical Review Letters , volume =. 2004 , month =

    2004

  7. [7]

    and Hardy, David J

    Phillips, James C. and Hardy, David J. and Maia, Julio D. C. and Stone, John E. and Ribeiro, João V. and Bernardi, Rafael C. and Buch, Ronak and Fiorin, Giacomo and Hénin, Jérôme and Jiang, Wei and McGreevy, Ryan and Melo, Marcelo C. R. and Radak, Brian K. and Skeel, Robert D. and Singharoy, Abhishek and Wang, Yi and Roux, Benoît and Aksimentiev, Aleksei ...

  8. [8]

    SoftwareX1-2, 19–25 (2015)

    Abraham, Mark J. and Murtola, Teemu and Schulz, Roland and Páll, Szilárd and Smith, Jeremy C. and Hess, Berk and Lindahl, Erik , title =. SoftwareX , year =. doi:10.1016/j.softx.2015.06.001 , issn =

    work page doi:10.1016/j.softx.2015.06.001 2015

  9. [9]

    and Lundborg, Magnus and Gray, Alan and Hess, Berk and Lindahl, Erik , title =

    Páll, Szilárd and Zhmurov, Artem and Bauer, Paul and Abraham, Mark J. and Lundborg, Magnus and Gray, Alan and Hess, Berk and Lindahl, Erik , title =. Journal of Chemical Physics , year =

  10. [10]

    The Journal of Physical Chemistry

    Time Step Rescaling Recovers Continuous-Time Dynamical Properties for Discrete-Time Langevin Integration of Nonequilibrium Systems , author=. The Journal of Physical Chemistry. B , year=

  11. [11]

    Mouhat, S

    F. Mouhat, S. Bonella and C. Pierleoni , title =. Molecular Physics , volume =. 2013 , publisher =

    2013

  12. [12]

    and Costa Cabral, B

    Galamba, N. and Costa Cabral, B. J. , title = ". The Journal of Chemical Physics , volume =. 2007 , month =

    2007

  13. [13]

    1977 , issn =

    Numerical integration of the cartesian equations of motion of a system with constraints: molecular dynamics of n-alkanes , journal =. 1977 , issn =

    1977

  14. [14]

    The Journal of Chemical Physics , volume =

    Coretti, Alessandro and Bonella, Sara and Ciccotti, Giovanni , title = ". The Journal of Chemical Physics , volume =. 2018 , month =

    2018

  15. [15]

    1989 , author =

    Computer Simulation of Liquids , publisher =. 1989 , author =

    1989

  16. [16]

    and Gezelter, J

    Fennell, Christopher J. and Gezelter, J. Daniel , title =. The Journal of Chemical Physics , volume =. 2006 , month =

    2006

  17. [17]

    2004 , publisher=

    Fast Multipole Methods for the Helmholtz Equation in Three Dimensions , author=. 2004 , publisher=

    2004

  18. [18]

    Journal of Computational Physics , volume=

    A hierarchical O(N) force calculation algorithm , author=. Journal of Computational Physics , volume=. 2002 , publisher=

    2002

  19. [19]

    and Darden, Tom and Lee, Hsing and Pedersen, Lee G

    Essmann, Ulrich and Perera, Lalith and Berkowitz, Max L. and Darden, Tom and Lee, Hsing and Pedersen, Lee G. , title =. The Journal of Chemical Physics , volume =. 1995 , month =

    1995

  20. [20]

    and Tezcan, Ismail and Hardy, David J

    Skeel, Robert D. and Tezcan, Ismail and Hardy, David J. , title =. Journal of Computational Chemistry , volume =

  21. [21]

    and Wu, Zhe and Phillips, James C

    Hardy, David J. and Wu, Zhe and Phillips, James C. and Stone, John E. and Skeel, Robert D. and Schulten, Klaus , title =. Journal of Chemical Theory and Computation , volume =

  22. [22]

    J. V. L. Beckers and C. P. Lowe and S. W. De Leeuw , title =. Molecular Simulation , volume =. 1998 , publisher =

    1998

  23. [23]

    Journal of Parallel and Distributed Computing , volume=

    FFT, FMM, and multigrid on the road to exascale: Performance challenges and opportunities , author=. Journal of Parallel and Distributed Computing , volume=. 2020 , publisher=

    2020

  24. [24]

    ACS Omega , volume =

    George, Anu and Mondal, Sandip and Purnaprajna, Madhura and Athri, Prashanth , title =. ACS Omega , volume =. 2022 , doi =

    2022

  25. [25]

    Adiabatic motion and statistical mechanics via mass-zero constrained dynamics

    Bonella, Sara and Coretti, Alessandro and Vuilleumier, Rodolphe and Ciccotti, Giovanni. Adiabatic motion and statistical mechanics via mass-zero constrained dynamics. Physical Chemistry Chemical Physics. 2020

    2020

  26. [26]

    and Scalfi,L

    Coretti,A. and Scalfi,L. and Bacon,C. and Rotenberg,B. and Vuilleumier,R. and Ciccotti,G. and Salanne,M. and Bonella,S. , title =. The Journal of Chemical Physics , volume =

  27. [27]

    Mass-Zero constrained dynamics and statistics for the shell model in magnetic field , volume =

    Girardier, David and Coretti, Alessandro and Ciccotti, Giovanni and Bonella, Sara , year =. Mass-Zero constrained dynamics and statistics for the shell model in magnetic field , volume =

  28. [28]

    The Journal of Chemical Physics

    Coretti, Alessandro and Baird, Taylor and Vuilleumier, Rodolphe and Bonella, Sara , year =. The Journal of Chemical Physics

  29. [29]

    The Journal of Chemical Physics , author =

    Mass-zero constrained molecular dynamics for electrostatic interactions , volume =. The Journal of Chemical Physics , author =. 2025 , pages =. doi:10.1063/5.0283356 , number =

    work page doi:10.1063/5.0283356 2025

  30. [30]

    Annalen der physik , volume=

    Die Berechnung optischer und elektrostatischer Gitterpotentiale , author=. Annalen der physik , volume=. 1921 , publisher=

    1921

  31. [31]

    The Journal of Chemical Physics , volume=

    Flexible simple point-charge water model with improved liquid-state properties , author=. The Journal of Chemical Physics , volume=. 2006 , publisher=

    2006

  32. [32]

    2023 , publisher=

    Understanding molecular simulation: from algorithms to applications , author=. 2023 , publisher=

    2023

  33. [33]

    Computer Physics Communications , volume=

    Comments on P3M, FMM, and the Ewald method for large periodic Coulombic systems , author=. Computer Physics Communications , volume=. 1996 , publisher=

    1996

  34. [34]

    Proceedings of the National Academy of Sciences , volume=

    Electrically driven first-order phase transition of a 2D ionic crystal at the electrode/electrolyte interface , author=. Proceedings of the National Academy of Sciences , volume=. 2025 , publisher=

    2025

  35. [35]

    Computer Physics Communications , author =

    A particle–particle particle-multigrid method for long-range interactions in molecular simulations , volume =. Computer Physics Communications , author =. 2005 , pages =. doi:10.1016/j.cpc.2005.03.077 , language =

    work page doi:10.1016/j.cpc.2005.03.077 2005

  36. [36]

    Physical Review E , author =

    Comparison of scalable fast methods for long-range interactions , volume =. Physical Review E , author =. 2013 , pages =. doi:10.1103/PhysRevE.88.063308 , language =

    work page doi:10.1103/physreve.88.063308 2013

  37. [37]

    Computer Simulation Using Particles

  38. [38]

    The Journal of Chemical Physics , volume =

    Darden, Tom and York, Darrin and Pedersen, Lee , title = ". The Journal of Chemical Physics , volume =. 1993 , month =

    1993

  39. [39]

    The Journal of Chemical Physics , volume=

    A compression strategy for particle mesh Ewald theory , author=. The Journal of Chemical Physics , volume=. 2021 , publisher=

    2021

  40. [40]

    Journal of Chemical Theory and Computation , volume=

    Periodic coulomb tree method: an alternative to parallel particle Mesh Ewald , author=. Journal of Chemical Theory and Computation , volume=. 2019 , publisher=

    2019