Matrix Holography on an Optical Lattice

arXiv:2609.05613 · hep-th, cond-mat.quant-gas, physics.atom-ph, quant-ph · Submitted 2026-09-04 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: High Energy Physics: Theory. The day in high-energy theory: gravity, strings, fields and the questions left open.

Kai: I'm Kai.

Mira: And I'm Mira, here with Lev.

Kai: We're looking at the paper: "Matrix Holography on an Optical Lattice".

Mira: Here's the gist of the paper. We propose an analog quantum-simulation protocol based on a lattice of atomic ensembles and time-averaged Floquet dynamics to construct four-body potentials typically found in bosonic BFSS-like matrix quantum mechanics.

Kai: This method avoids the deep circuits required by gate-based digital implementations, instead generating the target model through a fixed set of control stages per Floquet cycle, maintaining a constant number of consecutive unitaries with respect to the matrix size N.

Mira: The primary scaling cost is determined by the frequency range needed to control the growing lattice, which grows at most as O(N 3), offering a polynomial scaling route favorable for experimental realization.

Kai: Furthermore, we demonstrate that the coupling parameter can be tuned beyond O(1), enabling simulations of strong-coupling physics relevant to holographic phases.

Title and authors: Kai: So let's look at this Matrix Holography on an Optical Lattice paper and who wrote it. It's Robin Loewenberg and Julian Sonner from the University of Geneva.

Mira: They are focusing on mapping the physics of matrix models onto a lattice setup, which is a pretty cool way to bridge theoretical string theory concepts with experimental quantum simulation.

Lev: From an engineering standpoint, it's interesting that they are tackling four body potentials directly rather than just two body terms first.

Kai: They claim their approach generates the target model through a fixed set of control stages per Floquet cycle instead of needing deep circuits for gate based implementations.

Mira: The implication there is that you can achieve this model construction with a constant number of consecutive unitaries no matter how big N gets, which simplifies the hardware requirements.

Lev: That's promising if the overhead doesn't balloon too fast when you try to implement it on real hardware, because we know those deep circuits are where most of the error comes from.

Kai: But they also note that their principal scaling cost is O of n cubed due to the frequency range needed for lattice control.

Mira: So while the circuit depth might be manageable, you still need good control over a growing number of modes at certain frequencies.

The paper's summary: Kai: To summarize this Matrix Holography on an Optical Lattice paper, they are constructing four body potentials typically found in bosonic BFSS like matrix quantum mechanics using an analog quantum simulation protocol.

Mira: They use a lattice of atomic ensembles and time averaged Floquet dynamics to build these four body interactions.

Lev: The core idea is that by using time averaging and the Magnus expansion, you generate the target model from a fixed sequence of control stages in each Floquet cycle.

Kai: This method is valid for general four body couplings of that specific form, demonstrated first with a single monomial like lambda m cubed X one through X four.

Mira: They then show how more complex coupling structures, like those in equation one point one, can be assembled using this time averaging approach.

Lev: To get there they truncate the bosonic degrees of freedom using Holstein Primakoff to map them onto large spin operators that are in a ferromagnetic regime where m is close to j.

Kai: That results in a potential form V of J x equals lambda m to the power of j over two J x one J x two J x three J x four plus other terms.

Mira: The main result is that the time averaged Hamiltonian accurately reconstructs the four body potential, showing high unitary fidelity between the Floquet model and the target model even when lambda is greater than one.

Lev: So what this means for us is that strong coupling physics in matrix models might actually be accessible through these experimental setups.

The paper's improvements: Kai: They propose a blueprint for Matrix Potential Engineering in section four, detailing how to build the O of n four couplings on the lattice.

Mira: It requires two separate N by N lattices, all to all coupling between them, and independent control over chi yy and chi zz.

Lev: And they use time averaging techniques from the Magnus expansion to generate those higher order interactions from simpler two body terms.

Kai: They also show how the driving frequency omega acts as an effective stabilizing parameter that protects convergence to the target model.

Mira: This suggests you can probe strong coupling physics by tuning the coupling parameter beyond one using a high driving frequency omega, which we discussed earlier.

Lev: And they point out that achieving all pairwise couplings corresponding to non vanishing structure constants needs a frequency count scaling as omega cubed, which is a caveat for the experimental realization.

Conclusion: Kai: So to wrap up this Matrix Holography on an Optical Lattice paper, they've shown they can construct a single non local four body interaction using a Floquet Hamiltonian driven by three two body interactions.

Mira: This protocol is scalable to large N without being directly hampered by decoherence time, aiming to map the intricate structure of matrix models into Hamiltonians closer to experimental quantum optics.

Lev: And they flag that the lowest order four body interaction doesn't directly resemble bosonic physics and the Magnus expansion generates terms that suppress the four body interaction, which is a limitation we need to consider.

Kai: They suggest using the Hoppe model as a minimal matrix model for testing this protocol first.

Mira: The ultimate objective they are aiming for is to develop a quantum simulator capable of providing insights into the full supersymmetric BFSS model at large coupling.

Robin Löwenberg, Julian Sonner

University of Geneva · Geneva Quantum Center

hep-th, cond-mat.quant-gas, physics.atom-ph, quant-ph

Submitted: 2026-09-04

Updated: 2026-11-02

Comments: 33 + 24 pages, 13 figures, 1 table; v2: references added and minor clarifications

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 83/100

The gist: We propose an analog quantum-simulation protocol based on a lattice of atomic ensembles and time-averaged Floquet dynamics to construct four-body potentials typically found in bosonic BFSS-like…

Key concepts

Floquet Dynamics
This involves applying a periodic driving force (like an external laser pulse) to a quantum system. By time-averaging the resulting evolution over many cycles, the complex time-dependent interactions are effectively mapped onto a simpler, time-independent Hamiltonian that reproduces the desired four-body potential.
Matrix Quantum Mechanics (MQM)
This is a specific type of quantum mechanics where the degrees of freedom are represented by large matrices. The paper focuses on constructing potentials for these models, specifically those involving four-body interactions like $V(\mathbf{X}) = \lambda \text{Tr}\mathbf{X}_4}$, which are central to studying string theory and holographic duals.
Holstein-Primakoff Transformation
This transformation maps the bosonic degrees of freedom (which describe particles) onto large spin operators. This is done by relating the particle number/creation operators to spin operators, allowing the complex matrix model interactions to be expressed in terms of simpler, approximately ferromagnetic spin interactions.
Time-Averaging and Magnus Expansion
These mathematical tools are used to derive higher-order, non-local four-body interactions from a set of simpler two-body driving terms. The time average effectively captures the resulting complex dynamics, allowing the construction of the target four-body potential through a fixed sequence of control operations.

Terminology

Summary

We propose an analog quantum-simulation protocol based on a lattice of atomic ensembles and time-averaged Floquet dynamics to construct four-body potentials typically found in bosonic BFSS-like matrix quantum mechanics. p. 0

Where the paper says it

sonner@unige. ch Abstract: We propose an analog quantum-simulation protocol based on a lattice of atomic ensembles and time-averaged Floquet dynamics to construct the four-body potentials that typically appear in bosonic BFSS-like matrix quantum mechanics.

This method avoids the deep circuits required by gate-based digital implementations, instead generating the target model through a fixed set of control stages per Floquet cycle, maintaining a constant number of consecutive unitaries with respect to the matrix size N. p. 0
Where the paper says it

In contrast to gate-based digital implementations requiring deep circuits, our approach generates the target model through a fixed set of control stages per Floquet cycle, keeping the number of consecutive unitaries constant with respect to the matrix size N. The principal scaling cost in our construction is given by the frequency range required to control the growing lattice.

The primary scaling cost is determined by the frequency range needed to control the growing lattice, which grows at most as O(N 3), offering a polynomial scaling route favorable for experimental realization. p. 0
Where the paper says it

The principal scaling cost in our construction is given by the frequency range required to control the growing lattice. We show that this cost grows at most as O(N3 ), giving a polynomial scaling route to large−N, attractive to experimental implementation.

Furthermore, we demonstrate that the coupling parameter can be tuned beyond O(1), enabling simulations of strong-coupling physics relevant to holographic phases. p. 0
Where the paper says it

We show that this cost grows at most as O(N3 ), giving a polynomial scaling route to large−N, attractive to experimental implementation. Additionally, we demonstrate that the coupling parameter can be tuned beyond O(1), enabling the simulation of strong-coupling physics necessary for holographic phases.

The study focuses on bosonic matrix models featuring potentials of the form V(X) = lambda Tr X 4 = lambda X abX bcX cdX da, which expand into a sum of four-body interactions governed by various couplings. p. 3

Where the paper says it

In this study, we are primarily interested in bosonic matrix models containing potentials of the form V (X) = λTrX4 = λXabXbcXcdXda , (1. 1) which expand into a sum of four-body interaction among the matrix degrees of freedom with a set of couplings governing the strength of each monomial.

The technique is valid for general four-body couplings of this form, demonstrated first using a single monomial V(X) = lambda m cubed X 1X 2X 3X 4, and subsequently showing how more complex coupling structures, such as those in equation (1. p. 3
Where the paper says it

The technique we develop in this paper is valid for general four-body couplings of the form above. We demonstrate the scheme using single monomial of the form V (X) = λm3X1X2X3X4 , (1.

1), can be assembled. p. 3
Where the paper says it

2) and demonstrate how more elaborate couplings structures such as in eqn. 1) can be assembled.

The protocol employs a method to truncate bosonic degrees of freedom using techniques from [35, 36], but instead of mapping them to qubits, it utilizes the Holstein-Primakoff transformation to map the model onto large spin operators that are approximately in the ferromagnetic regime (m about j, where j is the spin and m is the magnetic quantum number). p. 4

Where the paper says it

[35, 36], but instead of mapping them to qubits, we use the Holstein-Primakoff transformation to map the model to large spin operators that are approximately in the ferromagnetic regime (i. , m ∼ j, where j is the spin and m is the magnetic quantum number).

This results in a potential of the form V(J x) = lambda m j/2 J x1 J x2 J x3 J x4 +. p. 1
Where the paper says it

6 Convergence of the Magnus Expansion 19 3. 7 Numerical Results 19 4 A Blueprint for Matrix Potential Engineering 23 4.

To construct the O(N 4) couplings in Matrix Quantum Mechanics (MQM), we propose translating the complex, non-local interactions of matrix models into a lattice geometry. p. 2

Where the paper says it

We are particularly interested in a special class of models known as matrix quantum mechanics (MQM), where some representatives, like the supersymmetric BFSS [22] or BMN [3] models, have long been known to possess holographic duals. The bosonic and fermionic degrees of freedom of these models are organized in Hermitian matrices that naturally produce highly non-local theories.

This experimental setup requires eight properties: two separate N times N lattices, all-to-all coupling between them, and independent control over two types of couplings (chi yy and chi zz). p. 4
Where the paper says it

We will show that the terms in the trace can be constructed individually, assuming that a Hamiltonian Hin = X a,b,c,d χ yy ab;cdI y abJ y cd + χ zz ab;cdJ z abJ z cd + HZ is realized in the experiment, where χ yy ab;cd and χ zz ab;cd are independently controllable couplings and HZ is a Zeeman term adding another degree of control to the system. The notation should be understood as follows: there are two individual lattices representing the matrices XI and XJ .

The resulting Hamiltonian is built using time-averaging techniques derived from the Magnus expansion to generate higher-order interactions from simpler two-body terms. p. 4
Where the paper says it

3) from a two-body interaction, thereby achieving high controllability also for higher orders. Our proposal is to add an additional step to the Floquet scheme that allows, by constructing oscillating two-body couplings, to achieve higher orders via time averaging using James’ effective Hamiltonian method [50] and the Magnus expansion.

Numerical results confirm that the time-averaged Hamiltonian accurately reconstructs the four-body potential of the target Hamiltonian, exhibiting high unitary fidelity between Floquet and target models, even in non-perturbative regimes where lambda > 1. p. 0

Where the paper says it

ch Abstract: We propose an analog quantum-simulation protocol based on a lattice of atomic ensembles and time-averaged Floquet dynamics to construct the four-body potentials that typically appear in bosonic BFSS-like matrix quantum mechanics. In contrast to gate-based digital implementations requiring deep circuits, our approach generates the target model through a fixed set of control stages per Floquet cycle, keeping the number of consecutive unitaries constant with respect to the matrix size N.

The driving frequency omega acts as an effective stabilizing parameter protecting convergence to the target model. p. 4
Where the paper says it

4) reproduces the potential (1. 3), where the frequency ω acts as a control parameter that determines the convergence of the Magnus expansion and, consequently, the coupling regime of the final model.

This suggests that probing strong-coupling physics is feasible with experimentally reasonable frequencies (O(MHz) - O(GHz)). p. 0
Where the paper says it

We show that this cost grows at most as O(N3 ), giving a polynomial scaling route to large−N, attractive to experimental implementation. Additionally, we demonstrate that the coupling parameter can be tuned beyond O(1), enabling the simulation of strong-coupling physics necessary for holographic phases.

The study concludes by outlining experimental hurdles, such as achieving highly controllable non-local interactions and managing heating, and suggests using the Hoppe model as a minimal matrix model for testing the protocol. p. 2

Where the paper says it

The bosonic and fermionic degrees of freedom of these models are organized in Hermitian matrices that naturally produce highly non-local theories. Even simpler (non-supersymmetric) realizations, such as the Hoppe model [23], can exhibit phenomena typical of holographic theories, like fast 1We acknowledge that the analog gravity community traditionally refers to this as the “analog” approach.

The ultimate objective is to develop a quantum simulator capable of providing insights into the full, supersymmetric BFSS model at large coupling. p. 5
Where the paper says it

We begin with the BFSS model as the most thoroughly studied example, which offers deep insights into holography, M-theory, fast scrambling, and various other phenomena. Although investigating the dynamics of its strong-coupling regime would be highly interesting, the complex coupling structure and large number of degrees of freedom make the BFSS model difficult to implement in a quantum simulator.

A detailed analysis on time-averaged spin-exchange interactions reveals that spin-conserving approaches face two issues: first, the lowest-order four-body interaction does not directly resemble bosonic physics; and second, the Magnus expansion generates numerous terms that interfere with the four-body interaction and suppress its associated physics, although the mathematical structure of these terms is non-trivial.

The analysis also indicates that realizing all pairwise couplings corresponding to non-vanishing structure constants requires a frequency count scaling as about O(N 4), contrasting with the O(N 2) scaling suggested by a two-dimensional architecture. In summary, we have demonstrated the construction of a single, non-local four-body interaction using a Floquet Hamiltonian driven by three two-body interactions governed by angular momentum operators. p. 4

Where the paper says it

3) from a two-body interaction, thereby achieving high controllability also for higher orders. Our proposal is to add an additional step to the Floquet scheme that allows, by constructing oscillating two-body couplings, to achieve higher orders via time averaging using James’ effective Hamiltonian method [50] and the Magnus expansion.

This protocol is scalable to large N without being directly hampered by decoherence time, aiming to map the intricate structure of matrix models into a class of Hamiltonians closer to experimental quantum optics. p. 9
Where the paper says it

In Chapters 3 and 4, we outline these primary challenges, notably the simultaneous realization of interacting bosonic and fermionic degrees of freedom, alongside the requisite precise control over large matrix dimensions and strong couplings. Although the protocol presented might be able to realize full BFSS on the long-term, it therefore is reasonable from an experimental point of view to first concentrate on simpler matrix models.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Direct simulation of four-body potentials in matrix quantum mechanics can be achieved by constructing a Floquet Hamiltonian from two-body interactions using time averaging and the Magnus expansion, which generates the target model through a fixed set of control stages per Floquet cycle.

  2. The system can generate programmable four-body couplings by constructing a drive Hamiltonian Hf(t) that reproduces the four-body interaction as well as in other ways, such as Hf(t) = 3s8w2λm j2J x1J y2cos(ωt) + J x3J y4cos(2ωt) − J z2Jz4cos(3ωt).

  3. The analog quantum simulation protocol can scale polynomially with the matrix size N, as the principal scaling cost is given by the frequency range required to control the growing lattice, which grows at most as O(N 3).

  4. The system can be engineered to simulate strong-coupling physics necessary for holographic phases by tuning the coupling parameter beyond O(1) using a high driving frequency omega.

  5. Experimental platforms such as photon-mediated spin-exchange architectures are well-suited to simulating Hamiltonians of the form H(t) = χ12J x1J y2cos(ω1t) + χ34J x3J y4cos(2ωt) − χ24J z2Jz4cos(3ωt).

  6. The system can be designed to realize complex matrix model potentials by partitioning the system into two subsystems and controlling interactions with specific frequency regimes that form a Golomb ruler.

  7. The AI system can perform state preparation tasks, such as preparing an initial thermofield double state, which is necessary for studying the dynamics of matrix models.

  8. The AI can utilize Jordan-Wigner transformations to encode fermionic degrees of freedom onto a lattice of spin-1/2 atoms using two-dimensional generalizations to simulate supersymmetric matrix models.

Abstract

We propose an analog quantum-simulation protocol based on a lattice of atomic ensembles and time-averaged Floquet dynamics to construct the four-body potentials that typically appear in bosonic BFSS-like matrix quantum mechanics. In contrast to gate-based digital implementations requiring deep circuits, our approach generates the target model through a fixed set of control stages per Floquet cycle, keeping the number of consecutive unitaries constant with respect to the matrix size N. The principal scaling cost in our construction is given by the frequency range required to control the growing lattice. We show that this cost grows at most as O(N 3), giving a polynomial scaling route to large-N, attractive to experimental implementation. Additionally, we demonstrate that the coupling parameter can be tuned beyond O(1), enabling the simulation of strong-coupling physics necessary for holographic phases.

Sources

Related papers