Summary
Can RCWA differentiation and parallel execution improve together?
Replaces eigendecomposition with matrix-square-root and matrix-exponential evaluations.
Targets a numerical bottleneck in optical-response optimization and parameter retrieval.
Related modelling approaches
Relation labels come from the method guide and are editorial. Unlabelled links are conceptual connections or comparison methods; the paper does not necessarily implement them.
Commentary
Matrix-function convergence, branch choices and precision remain numerical concerns; no universal speed factor is assumed here.
Changes the forward-solver machinery while retaining the layered electromagnetic problem.
Interpretation and connections in this note, separate from the authors’ reported findings.