1. User Guide

This guide is meant to help Tidy3D users understand and get started with the code.

Defining Simulation

All of the information needed to run a Tidy3D simulation are contained in the tidy3d.Simulation object. This object can be thought of as a large, nested data structure containing various components and parameters. As such, this package gives a python interface for defining this object as well as tools for running the simulation and analyzing the results.

Tidy3d Components

The Simulation object contains a set of general parameters, such as the size and resolution, which define the global context of the simulation and physics. Additionally, the Simulation contains a series of "components" that further define the specific physics. For example, a few of the main components are:

• tidy3d.Structure objects, which define the geometry of objects in your simulation, as well as their material properties.
• tidy3d.Source objects, which define the electric current distributions that give rise to the electromagnetic fields in the simulation.
• tidy3d.Monitor objects, which define how data from the simulation is measured and stored for analyzing later.

Plugins

While these components can be hard-coded or defined programmatically, at times they require more exploration or experimentation to define. For this purpose, we also provide various "plugins" to help users with more complicated aspects of defining these components. Here are a few examples of plugins used to help define the Simulation.

• A library of commonly used dispersive materials that can be imported and used in the simulation.
• A tool for fitting optical (n, k) data to create customized materials.
• A mode solver that allows the user to define custom modal sources and monitors.

Many more plugins are in development and will be released in the coming months, with tools for optimization, structure analysis, nonuniform meshing, and others.

Running Simulation

With the Simulation object defined, we can then run it on Flexcompute's servers using the web api also defined in this repository. This section will describe how to do this at a high level.

Requirements

While anyone can define a Tidy3D simulation through the python API provided here, to run a simulation on our servers requires a Tidy3D account and "credits". The credits required for running a simulation depend on the amount of "work" required to perform the computation. For more details on signing up and purchasing credits, see this link.

Web API

Because all simulations must be run on our remote servers, we provide an API for uploading, managing, and downloading your simulations. A set of tidy3d.web provides a set of functions for granular control of tasks, but we also provide tidy3d.web.Job and tidy3d.web.Batch interfaces for more convenient handling of single or multiple jobs, respectively.

At a high level, the steps required for running a simulation are:

1. Upload Simulation object to server with some metadata, which is then referred to as a 'task'.
2. Tell the server to start running the task.
3. Monitor task progress.
4. When finished, download data, including the solver log and data for each Monitor.
5. Load the data locally into a tidy3d.SimulationData object, which provides an interface for analyzing, and visualizing the results.

Browser Interface

Alternatively, users may use a browser-based interface for managing all previous and current tasks, which will be updated continuously in the coming months with additional for task editing and data visualization, among others.

Finally

If you have a specific question, see the documentation linked in the footer of this wiki or leave a question in our Issues or Discussion page of this repository.