Search papers, labs, and topics across Lattice.
This paper introduces an enhanced Fourier Neural Operator (FNO) specifically designed for modeling two-dimensional Rayleigh-B茅nard convection by predicting time increments rather than full solutions. The improved model boasts a compact architecture with only 314k parameters and achieves rapid inference times of 7 ms, while maintaining comparable accuracy to traditional FNOs. Notably, the study reveals that while FNOs can generalize to finer meshes, their accuracy is ultimately constrained by the resolution of the training data used.
Predicting time increments with a compact Fourier Neural Operator leads to faster and equally accurate modeling of Rayleigh-B茅nard convection.
We propose an improved Fourier Neural Operator (FNO) for modeling two-dimensional Rayleigh-B\'enard convection by predicting time increments instead of full solutions, achieving higher accuracy than a standard FNO baseline. The resulting model is compact (314k parameters, 1.26 MB) and fast (7 ms inference), while maintaining similar accuracy as demonstrated in previous benchmarks. We show that although FNOs generalize to finer meshes, accuracy remains limited by the resolution of the training data.
