Simulating Quantum Turbulence with Matrix Product States
Felipe Gómez-Lozada, Nicolas Perico-García, Nikita Gourianov, Hayder Salman, Juan José Mendoza-Arenas · 2025
Read Simulating Quantum Turbulence with Matrix Product States by Felipe Gómez-Lozada, Nicolas Perico-García, Nikita Gourianov, Hayder Salman, Juan José Mendoza-Arenas (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-12191.txt.
Introduction
Felipe Gómez-Lozada and four co-authors present a cheaper way to simulate quantum turbulence in this 2025 paper. Quantum turbulence spans scales from the whole system down to a tiny healing length. Direct simulation of the governing equation across that range costs enormous computing power.
Their solver compresses the wave pattern by cutting weak links between scales, using a chain-like data form. Memory use fell by factors from ten to over ten thousand against direct simulation. Tests on dark solitons and spinning vortices matched known behavior, including wave motion along vortex lines. For full turbulent states, savings grew with the density of the features. Standard checks on energy patterns matched established results with little memory cost.
Catalog record
| CREATOR | Felipe Gómez-Lozada, Nicolas Perico-García, Nikita Gourianov, Hayder Salman, Juan José Mendoza-Arenas |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 101 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-12191
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Simulating Quantum Turbulence with Matrix Product States Felipe Gómez-Lozada, Nicolas Perico-García, Nikita Gourianov, Hayder Salman, Juan José Mendoza-Arenas Abstract Quantum turbulence spans length scales from the system size L L to the healing length ξ \xi , making direct numerical simulations (DNS) of the Gross-Pitaevskii (GP) equation computationally expensive when L ≫ ξ L\gg\xi . We present a matrix product state (MPS) solver for the GP equation that efficiently compresses the wavefunction by truncating weak interlength-scale correlations. This approach reduces memory use by factors ranging from 10x to over 10,000x compared to DNS. We benchmark our approach on nonlinear excitations, namely dark solitons (1D) and quantized vortices (2D, 3D), capturing key dynamics like Kelvin wave propagation and vortex ring emission in the case of vortex line reconnection. For turbulent states composed of multiple nonlinear excitations, we find that the memory compression of the MPS representation is directly proportional to the soliton or vortex densities. We also accurately reproduce established results from two-point correlation functions and energy spectra, where we recover the incompressible kinetic energy spectrum with little memory overhead. These results demonstrate the representative capabilities of the MPS ansatz for quantum turbulence and pave the way for studying this nonequilibrium state using previously-prohibited system sizes to uncover possible new physics. I Introduction The advent of the Bose-Einstein condensates (BECs) in cold atom experiments in 1995 Anderson et al. 1995 ; Davis et al. 1995 is a milestone of many-body quantum phenomena. The striking properties of this state of matter, such as superfluidity, nonlinear dynamics, quantization of vortex circulation, and emergence of second sound Pethick and Smith 2008 ; Paoletti and Lathrop 2011 ; Madeira et al. 2020 ; Proukakis 2025 , have motivated a great deal of research. Moreover, the high tunability of cold-atom experiments Dalfovo et al. 1999 ; Bloch et al. 2008 ; Bloch et al. 2012 means that these systems hold promise in a wide range of applications, including quantum metrology Estève et al. 2008 ; Ngo et al. 2021 ; Mao et al. 2023 , interferometry Wales et al. 2020 ; Grimshaw et al. 2022 , and quantum simulators Bloch et al. 2012 . Nowadays, BECs have been realized in diverse unconventional quantum fluid systems Bennemann and Ketterson 2013 ; Bennemann and Ketterson 2014 ; Proukakis 2025 such as exciton-polariton condensates, magnon condensates, and dipolar molecules Demokritov et al. 2006 ; Klaers et al. 2010 ; Altfeder et al. 2017 ; Bigagli et al. 2024 . As a result, such systems are ideal for studying fundamental nonequilibrium physics under controlled experimental conditions. This includes nonequilibrium thermal Lamporesi et al. 2013 and quantum phase transitions Yi et al. 2020 , dynamics of Bose-gas mixtures such as spinor condensates Stamper-Kurn and Ueda 2013 , and quantum turbulence in BECs White et al. 2014 ; Navon et al. 2016 ; Fischer and Bradley 2025 . Moreover, the ability to realize BECs in low-dimensional systems Kinoshita et al. 2006 ; Langen et al. 2015 allows the study of 1D physics where integrable quantum field theories become relevant (e.g., in generalized hydrodynamics Ruggiero et al. 2020 ; Bonnemain et al. 2022 ; Doyon et al. 2025 ), as well as 2D quantum turbulence Neely et al. 2013 ; Johnstone et al. 2019 ; Gauthier et al. 2019 ; Gałka et al. 2022 ; Reeves et al. 2022 ; Karailiev et al. 2024 and Berezinskii–Kosterlitz–Thouless physics Hadzibabic et al. 2006 ; Sunami et al. 2023 . In modeling quantum fluids, much of the physics referred to in the above examples is dictated by a BEC well-represented by a macroscopic occupation of a single-particle state Pitaevskii and Stringari 2016 ; Pethick and Smith 2008 . A natural starting point in modeling this physics is through a mean-field or classical-field approximation, given by the Gross-Pitaevskii (GP) equation Pethick and Smith 2008 . This model captures the nonequilibrium spatio-temporal dynamics of superfluids and BECs, including nonlinear excitations such as quantized vortices. In the context of quantum turbulence for BECs, the GP model accounts for compressibility in the system, enabling the coexistence and energy exchange between phonons and vortices, mediating processes such as vortex dipole annihilation in 2D, vortex reconnections in 3D, dissipative vortex dynamics, and phonon scattering of vortices Pismen 1999 . These mechanisms generate multiscale turbulent regimes Barenghi et al. 2023a , from wave turbulence dominated by compressible modes Kolmakov et al. 2014 to Vinen or Kolmogorov turbulence where vortices dominate the dynamics Barenghi et al. 2014 . As these studies provide key insights into fundamental processes at prescribed scales, extending GP-type simulations to span a broader spectrum of scales is needed for a better understanding of these diverse fields mentioned above. Nevertheless, direct numerical simulations (DNS) for this endeavor are prohibitively costly when the system size L is much larger than the healing length \xi ( \sim size of a vortex) due to an exploding number of computational variables. A similar problem occurs for quantum many-body systems. For the latter, a challenge in their modeling corresponds to accurately representing the quantum state described by the many-body wavefunction that resides in an exponentially large-dimensional Hilbert space. However, since the many-body Schrödinger equation governs the wavefunction, it contains a significant amount of structure that a direct numerical solution does not exploit. Tensor networks target the underlying structure by recognizing that, for many systems of interest, the governing Hamiltonian leads to limited entanglement within the many-body wavefunction Schollwöck 2011 . By decomposing the extensive quantum states into a series of low-rank tensors connected by virtual bonds, tensor networks can lead to a drastic reduction in the number of parameters that are required to describe the state of the system accurately Orús 2014 . The quantics representation of continuous variable functions leverages the analogy between many-body quantum systems and the solution of partial differential equations (PDEs) that resolve a large number of length scales. It does so by encoding real-space functions into quantum states of multiple qubits, where each two-level component represents a particular length scale in real space. Then, a tensor network decomposition is used to compress the given state, depending on the correlation distribution among length scales Oseledets 2010 ; Khoromskij 2011 . In particular, the hallmark matrix product state (MPS) Orús 2014 (also called tensor train Oseledets 2011 ), has been widely used for these purposes. This insight has opened the door for using tensor network methods in the solution of large-scale PDEs. By exploiting the interlength-scale correlation structure, it has been possible to achieve low-rank representations of classical turbulent flows Gourianov et al. 2022 ; Gourianov 2022 ; Kornev et al. 2024 ; Kiffner and Jaksch 2023 ; Peddinti et al. 2024 ; Hölscher et al. 2025 ; Gourianov et al. 2025 ; Adak et al. 2025 ; Pisoni et al. 2025 ; Arenstein et al. 2025 , plasma states Ye and Loureiro 2022 ; Ye and Loureiro 2024 , wave phenomena Fraschini et al. 2024 ; Lively et al. 2025 and different properties of continuous quantum systems Shinaoka et al. 2023 ; Ritter et al. 2024 ; Rohshap et al. 2025 . In this work, we develop a tensor network algorithm to simulate the nonlinear GP equation. We perform a systematic study