MagTense version 2.2.5

Version 2.2.5 of MagTense includes the following:

MEX-files with support for Matlab R2025a and R2025b.

The MEX-files for Windows and Linux distributed directly in the archive along with examples files on how to use these.

For Python, installation of MagTense is recommended via pip (requires >=Python 3.12) using pip install magtense. Examples files for python are included in this release as well.

MagTense version 2.2.4

Version 2.2.4 of MagTense includes the following:

• Fix memory leak using CVODE.
• Updated example of unstructured cartesian (prism) meshes in python.
• Version bump to v2.2.4 to ensure consistency with PyPI.

The MEX-files for Windows and Linux distributed directly in the archive along with examples files on how to use these.

For Python, installation of MagTense is recommended via pip (requires >=Python 3.12) using pip install magtense. Examples files for python are included in this release as well.

MagTense version 2.1.0

Version 2.1.0 of MagTense includes the following:

• Unstructured cartesian (prism) meshes in python. An example file for an unstructured mesh is provided for standard problem 4 in python (identical to the Matlab example already provided).
• Cubic and general anisotropies in both Matlab and python. In Matlab an example of a CoFe nanopiller with cubic anisotropy is provided.
• Magnetostatic scalar potentials for a prism and a cylinder (full and slice). The magnetostatic scalar potentials for a prism and a cylinder are provided in Matlab, for the time being.

The MEX-files for Windows and Linux distributed directly in the archive along with examples files on how to use these.

For Python, installation of MagTense is recommended via pip (requires >=Python 3.12) using pip install magtense. Examples files for python are included in this release as well.

MagTense version 2.0

Version 2.0 of MagTense includes Matlab MEX-files for Windows and Linux with CUDA support and statically linked MKL and CVODE libraries.

The MEX-files for Windows and Linux distributed directly in the archive along with examples files on how to use these.

For Python, installation of MagTense is recommended via pip (requires >=Python 3.12) using pip install magtense. Examples files for python are included in this release as well.

MagTense version 1.2

Version 1.2 of MagTense includes an update to singularities in the tetrahedron demag tensor, as well as support for CUDA 12.1. It also disables support for OpenMP for constructing the demagnetization tensor for micromagnetic simulations, due to bugs.

The MEX-files for Matlab for Windows are distributed directly in the archive.

Instructions on how to build and use the Python interface are listed in the python folder of the repository. Installation is recommended via conda package manager (requires >= Python 3.9). Installation with CUDA 12.x: conda install magtense -c cmt-dtu-energy/label/cuda-12 -c nvidia/label/cuda-12.2.2 . Installation without CUDA support: conda install magtense -c cmt-dtu-energy/label/cpu

MagTense version 1.1

Version 1.1 of MagTense includes a python interface to the micromagnetism model, as well as updates MEX-files for Matlab on Windows.

The MEX-files for Matlab for Windows are distributed directly in the archive.

Instructions on how to build and use the Python interface are listed in the python folder in the archive. Installation is recommended via conda package manager (requires Python 3.9 or 3.10). For CUDA support {from 11.4.4 , 11.6.2 , 11.7.0 } use: conda install magtense -c cmt-dtu-energy/label/cuda-11.7.0 -c nvidia/label/cuda-11.7.0 . Installation without CUDA support: conda install magtense -c cmt-dtu-energy/label/cpu

MagTense version 1.0.1

This is the first release of MagTense in which the MEX-files for Windows are directly distributed and the Python version is available on PyPi.

The MEX-files for Matlab integration, along with examples for both Matlab and python, are available as an archive below.

The python version can be installed via PyPi:

python3 -m pip install magtense