Releases: loganbvh/superscreen
v0.10.5
Contributors
Assets 2
v0.10.4
v0.10.3
v0.10.2
v0.10.1
v0.10.0
v0.9.2
v0.9.1
v0.9.0
SuperScreen v0.9.0 is a significant update intended to improve the CPU/memory efficiency and ergonomics of the package. The major upgrades are contained in the following Pull Requests: #96, #99.
Major changes
-
Rather than creating a single mesh containing all films in a device, each film now gets its own mesh and finite element operators.
- This approach avoids meshing large vacuum regions between films, which is costly.
- This approach also means that films with transport terminals can exist in the same
superscreen.Deviceas films without terminals. As a result, thesuperscreen.TransportDeviceclass is no longer needed and has been removed. - It is no longer necessary to explicitly define a bounding box around the film(s) being modeled. A suitable bounding box is automatically generated for each film within
superscreen.Device.make_mesh().
-
Reduced memory footprint and improved performance using
numbaJIT compiled functions.- Several costly numerical operations have been migrated from
numpy/scipyto customnumbajust-in-time (JIT) compiled functions. Thenumbafunctions are automatically performed in parallel on multiple CPU cores and avoid the allocation of large intermediate arrays, which can cause significant memory usage innumpy/scipy. - For devices with multiple films, the inductive coupling between films is now calculated using the supercurrent density and a
numbaimplementation of the Biot-Savart law, rather than the stream function and Ampere's law. The new approach is both more robust for closely-stacked films and avoids storage of large temporary arrays. - The default for
superscreen.Device.solve_dtypehas been changed fromfloat64tofloat32.
- Several costly numerical operations have been migrated from
-
The linear system describing the magnetic response of each film is now LU factored only once per call to
superscreen.solve().- This dramatically speeds up self-consistent simulations involving multiple films and makes the solver code more modular.
- The portions of the model that are independent of the applied field can be pre-factorized using
superscreen.factorize_model(), which returns asuperscreen.FactorizedModelobject that can be saved for future use. Asuperscreen.FactorizedModelinstance can be passed directly tosuperscreen.solve().
-
As a result of the above optimizations, GPU support using
jaxand parallel processing with shared memory usingrayno longer seem to add much value to the package, so they have been removed.- The
gpuargument tosuperscreen.solve()has been removed, along with the (optional) dependency onjax. superscreen.solve_many()has been removed, along with the dependency onray.
- The
-
All IO operations, including writing
superscreen.Deviceandsuperscreen.Solutionobjects to disk, are now performed using the HDF5 file format viah5py.- All objects within
superscreenthat can be serialized to disk now have.to_hdf5()and.from_hdf5()methods.
- All objects within
-
SuperScreenhas dropped support for Python 3.7, which will reach end-of-life in June 2023.- Added support for Python 3.11, which was being blocked by the dependency on
ray.
- Added support for Python 3.11, which was being blocked by the dependency on