Micro dB Convergence for Quantum Device Applications

For a 3D simulation of this inverse taper of length 80um, a convergence of 0.1udB was reached with a simulation runtime of just 95s.

These simulations were performed on a Standard Desktop PC without requiring the use of any cloud computing thanks to Photon Design's advanced EME algorithm.

A metric for the validity of an output is whether it converges to a value as the resolution of the simulation increases. This is known as convergence, and determines the accuracy of a simulation's output. It is governed by a set of parameters within Photon Design's FIMMPROP which can be fine tuned to minimise fluctuations in the data. 

Here the convergence parameter from the EME solver is the number of modes. Increasing the number of modes in the basis set of an EME simulation enables a more accurate depiction of the propagating field. As the number of modes in the simulation increases, results converge with decreasing error.

For a given set of convergence parameters, the discretisation of the simulation becomes an automated process the FIMMPROP's EME solver offers (reach out to our team for more details). 

The error is the percentage difference between the transmission at the current number of modes with respect to the final convereged value at 20 modes. 

Importantly, even with a small number of nodes and short runtime of less than a minute, the software reaches errors of the order 10udB.

Reach out to us for more details on advanced taper optimisations including FIMMPROP's automated simulation discretisation.