[Paper Review] NuMagSANS: a GPU-accelerated open-source software package for the generic computation of nuclear and magnetic small-angle neutron scattering observables of complex systems

Michael P. Adams, Andreas Michels|arXiv (Cornell University)|Jan 26, 2026

Advanced NMR Techniques and Applications2 citations

TL;DR

NuMagSANS is a GPU-accelerated, open-source tool that computes nuclear and magnetic SANS observables from real-space data, with extensive support for polarized, unpolarized, and interference terms and a large library of response functions.

ABSTRACT

We present NuMagSANS, a GPU-accelerated software package for calculating nuclear and magnetic small-angle neutron scattering (SANS) cross sections and correlation functions. The program allows users to import position-dependent nuclear density and magnetization data, providing a large flexibility for analyzing the scattering signatures of complex systems, particularly magnetic materials. Full rotational control of the sample is supported, allowing a comprehensive exploration of angular-dependent scattering features. NuMagSANS includes a versatile library of approximately 100 response functions that encompass two-dimensional SANS cross sections, correlation functions, and azimuthally averaged quantities. These capabilities allow users to gain detailed insight into the structural and magnetic characteristics of their samples. GPU acceleration ensures rapid computations, even for large data sets, making NuMagSANS a powerful and efficient tool for advanced SANS analysis.

Motivation & Objective

• Address computational challenges in evaluating nuclear and magnetic SANS observables for structurally and magnetically complex systems.
• Provide a flexible framework to compute 2D SANS cross sections, correlation functions, and azimuthally averaged quantities from user-supplied real-space data.
• Enable seamless integration with micromagnetic simulations and separate form-factor from interparticle contributions for advanced analysis.
• Offer a modular, open-source solution under the MIT license with straightforward installation and benchmarking datasets.

Proposed method

• GPU acceleration using NVIDIA CUDA to rapidly compute SANS observables from real-space nuclear and magnetization data.
• C++/CUDA implementation with inputs including nuclear scattering-length density, magnetization fields, and structural data.
• Supports nuclear, magnetic, and nuclear-magnetic interference terms, plus polarized and unpolarized cross sections and various measurement geometries.
• Provides a library of ~100 response functions for two-dimensional SANS cross sections, correlation functions, and azimuthally averaged quantities.
• Includes azimuthal averaging and a Fourier/modal decomposition to quantify anisotropy in 2D scattering patterns.
• Enables data handling from atomistic or micromagnetic (MuMax3) inputs and outputs to facilitate integration with micromagnetic pipelines.

Figure 1: Sketch of the neutron scattering geometry that is implemented in NuMagSANS . The neutron optical elements (polarizer, spin flipper, analyzer) that are required to measure the spin-resolved SANS cross sections are not drawn. The external magnetic field $\mathbf{B}_{0}\parallel\mathbf{e}_{z}

Experimental results

Research questions

• RQ1 How can nuclear and magnetic SANS observables be efficiently computed from complex real-space data using GPU acceleration?
• RQ2 How can nuclear form factors be separated from interparticle interference to analyze structure factors in SANS?
• RQ3 How well can NuMagSANS reproduce and extend existing SANS calculations across polarized, unpolarized, and chiral cross sections?
• RQ4 What is the utility of integrating SANS computations with micromagnetic simulations for magnetic nanostructures and skyrmion-related systems?

Key findings

• NuMagSANS provides GPU-accelerated computation of two-dimensional SANS cross sections and correlation functions from user-defined real-space data.
• The package supports nuclear, magnetic, and nuclear-magnetic interference terms, along with polarized and unpolarized cross sections and chiral contributions.
• It includes a versatile library of about 100 response functions for various SANS observables and supports azimuthal averaging and modal decomposition to quantify anisotropy.
• NuMagSANS integrates with micromagnetic tools (e.g., MuMax3), enabling direct computation of magnetic SANS observables from simulated spin configurations.
• The software is open-source under the MIT license and comes with benchmark datasets (Zenodo) to validate SANS simulations and facilitate reproducibility.

Figure 2: Schematic of the basic workflow of NuMagSANS (see the main text for explanations).