Home / Research / Quantum Bootstrap

Quantum bootstrap, without the quantum part.

A phase-space twist on Nakayama's classical bootstrap that finally tames the unphysical region of the double-well potential, and lays the moment-based scaffolding for hybrid classical-quantum simulation.

At a glance

VenueIEEE ICPCT2025
DOI10.1109/ICPCT64145
.2025.10939270
Systems validated4HO · DW · Coulomb · Toda
Method typePhase-spacemoment-constrained
Headline result↓ unphysicalE<0 region in DW potential
Author positionSole1/1

01The problem with the classical bootstrap

The bootstrap method is one of those quietly powerful ideas that keeps resurfacing in physics. The premise is almost embarrassingly simple: instead of solving the Schrödinger equation, you write down everything you can't get away with, positivity of expectation values, Hamiltonian recursion relations, commutator algebra, and let those constraints squeeze the answer out. It works. It has worked since the 1960s for large-N systems, conformal field theory, lattice models and matrix models.

Nakayama showed in earlier work that the same trick translates to classical microcanonical ensembles in the ℏ → 0 limit. But there was a wart: in the double-well potential, the bootstrap leaves an unphysical region at E < 0 that refuses to converge no matter how much you grow the moment grid. That region is real and structural, not a numerical artefact you can compute your way out of.

"The bootstrap is constraint satisfaction in its purest form. The interesting question isn't can we constrain, it's which constraints buy us convergence."

02What this paper changes

The contribution is small in surface area and large in consequence: instead of bootstrapping in position space alone, I add phase-space information, momentum moments, mixed moments, and the cross-constraints they imply via the canonical Poisson bracket. Geometrically, you're enforcing positivity on a richer object. Numerically, that means more equations, more rejected candidate energies, and a sharper convergence cone.

The headline finding: the unphysical E < 0 region in the double-well potential shrinks substantially. It does not vanish, that would be too neat, but it stops being the obstacle it was for downstream applications.

03The four test cases

  • Harmonic oscillator, the sanity check. Closed-form energies, the bootstrap should hit them. It does.
  • Double-well potential, the load-bearing case. Pre-existing unphysical region; phase-space constraints visibly compress it.
  • Coulomb potential, singular, divergent expectation values; the bootstrap has to behave gracefully under those. It does, with the caveat that you need careful regularisation of the lowest moments.
  • Non-relativistic Toda model, the integrable system; tests whether the method respects conservation laws automatically. It does.

04Why this matters for quantum computing

Phase-space bootstrap is a moment-positivity problem. Moment-positivity problems map onto semidefinite programs. SDPs admit quantum speedups under specific structural conditions, and quantum simulators can in principle estimate the relevant moment expectation values directly. The contribution here isn't a quantum algorithm; it's a cleaner formulation that a quantum algorithm could plug into.

That's the throughline I keep coming back to in this work and the IEEE Access 2026 quantum-kernels paper: the engineering question for the next decade isn't where do we run quantum algorithms, it's which classical formulations are quantum-friendly enough to be worth the round-trip.

FAQWhat people ask me about this paper

Q1Why "quantum bootstrap" if you're working classically?
The technique is named after its quantum-mechanical origin, self-consistency over operator expectation values. The classical correspondence (this paper) is what you get in the ℏ → 0 limit; the constraints survive, the operators become functions on phase space.
Q2Does this fully solve the double-well unphysical-region problem?
No. It significantly reduces the region's footprint. There's still residual non-convergence at the boundary that I believe is intrinsic to the moment hierarchy at finite truncation, not a defect of the phase-space approach.
Q3How does this fit alongside the IEEE Access 2026 quantum-kernels work?
Different problem, same intellectual stance: take a classical method, identify the structural slot where quantum primitives plug in cleanly, and validate the slot, not vapourware end-to-end pipelines. The bootstrap moment problem is one such slot; QSVM kernels are another.
Q4Is the code public?
Reference implementations of the four test cases are available on request; reach me at s.ali.badami@gmail.com. A full repository drop is planned alongside follow-up work currently in preparation.
Q5Where does this go next?
Two threads. (1) Pushing the phase-space bootstrap onto matrix-model and large-N problems where the constraint structure is even richer. (2) A hybrid pipeline where quantum hardware estimates the moments and classical SDP solvers close the loop.

CITEHow to cite this paper

@inproceedings{badami2025bootstrap,
  author    = {Shujaatali Badami},
  title     = {Quantum Bootstrap in Microcanonical Ensembles:
               Computational Insights and Applications},
  booktitle = {2025 IEEE International Conference on
               Pervasive Computing and Technology (ICPCT)},
  year      = {2025},
  pages     = {1--6},
  publisher = {IEEE},
  doi       = {10.1109/ICPCT64145.2025.10939270}
}
S. Badami, "Quantum Bootstrap in Microcanonical Ensembles: Computational Insights and Applications," in 2025 IEEE International Conference on Pervasive Computing and Technology (ICPCT), 2025, pp. 1-6, doi: 10.1109/ICPCT64145.2025.10939270.
Badami, S. (2025). Quantum bootstrap in microcanonical ensembles: Computational insights and applications. In 2025 IEEE International Conference on Pervasive Computing and Technology (ICPCT) (pp. 1-6). IEEE. https://doi.org/10.1109/ICPCT64145.2025.10939270
TY  - CONF
AU  - Badami, Shujaatali
TI  - Quantum Bootstrap in Microcanonical Ensembles: Computational Insights and Applications
T2  - 2025 IEEE International Conference on Pervasive Computing and Technology (ICPCT)
PB  - IEEE
PY  - 2025
DO  - 10.1109/ICPCT64145.2025.10939270
ER  -

SEE ALSORelated work in this portfolio