For simulating propagation loss on Ge waveguide for Mid-infrared on Si substrate such as a strip structure or other structures like ridge, I understand both Mode or FDTD solver can do the job. What are the advantages and disadvantages for each solver in this case?
If the structure is uniform in the direction of propagation, then the FDE solver in MODE Solutions will be much faster. FDTD will be required if there variation in the device in the direction of propagation.
It is also worth noting a potential issue with using FDTD to measure waveguide loss: If the loss is small, it is necessary to simulate a very long propagation distance before the amount of loss is measurable, which might require impractically large simulations. In some situations, it is easier to measure waveguide loss experimentally, rather than trying to simulate it.