FDTDX
What it is
FDTDX is an open-source Python package for simulating and designing three-dimensional photonic nanostructures with the finite-difference time-domain method. Built on JAX, it provides automatic differentiation for gradient-based optimization and describes GPU-accelerated simulations with multi-GPU scaling. Its documentation says it can handle simulations with billions of grid cells, and the package uses time-reversibility in Maxwell's equations to compute gradients efficiently. The API supports positioning and sizing objects with absolute or relative coordinates, as well as importing GDS polygons and working with GDS layer stacks and ports. FDTDX is available under the MIT license for research, development, and commercial use. Installation is documented with `pip install fdtdx`; the project recommends JAX GPU acceleration packages for NVIDIA GPUs, AMD GPUs, and TPUs. Documentation includes interactive Jupyter quickstarts, physics tutorials, component examples, best-practice material, and an API reference. Community help is available through GitHub issues and discussions. The FAQ describes FDTDX as relatively new and says bugs may occur despite testing. Its listed platforms are Linux and self-hosted.
For and against
For
- MIT-licensed for research, development, and commercial use.
- Automatic differentiation supports gradient-based optimization.
- GPU acceleration and multi-GPU scaling are described.
- API includes GDS polygon and layer-stack functions.
- Documentation includes tutorials and an API reference.
Against
- FAQ says bugs may occur because the project is relatively new.
- Listed platforms are Linux and self-hosted.
- GPU acceleration requires JAX packages for the relevant hardware.
The verdict
FDTDX offers simulation and optimization tools for photonic nanostructures, with GDS functions and GPU options. Its own FAQ flags the possibility of bugs in this relatively new project.
Who it suits
FDTDX suits researchers and developers working on photonic nanostructure simulation or design who use Python and JAX. It also targets commercial use, with GPU acceleration options for NVIDIA, AMD, and TPU hardware.
Plans and prices
FDTDX
Open source · freely available for research, development, and commercial use
Read 4 Oct 2026 · fdtdx.readthedocs.io
On the record
- Accelerators
- The project recommends JAX GPU acceleration packages for NVIDIA GPUs, AMD GPUs, and TPUs.github.com · read 3 Oct 2026
- API
- Objects can be positioned and sized using absolute or relative coordinates.fdtdx.readthedocs.io · read 4 Oct 2026
- Assessment tools
- Nofdtdx.readthedocs.io · read 23 Sep 2026
- Audience
- The project presents FDTDX for research, development, and commercial use in photonic nanostructure simulation and design.fdtdx.readthedocs.io · read 3 Oct 2026
- Authoring tools
- Yesfdtdx.readthedocs.io · read 23 Sep 2026
- Data export
- Yesfdtdx.readthedocs.io · read 23 Sep 2026
- Delivery mode
- desktopfdtdx.readthedocs.io · read 23 Sep 2026
- Documentation
- The project provides an API reference, tutorial guides, and best-practice documentation.github.com · read 3 Oct 2026
- GPU integrations
- The project README lists JAX installation options for NVIDIA GPUs, AMD GPUs, and TPUs.github.com · read 4 Oct 2026
- GPU support
- The project describes GPU-accelerated simulations with multi-GPU scaling capabilities.fdtdx.readthedocs.io · read 3 Oct 2026
- Guides
- Documentation includes interactive Jupyter quickstarts, physics tutorials, component examples, and an API reference.fdtdx.readthedocs.io · read 4 Oct 2026
- Installation
- The maker provides a pip installation command and an editable installation path for development.fdtdx.readthedocs.io · read 4 Oct 2026
- Integrations
- The API includes functions for importing GDS polygons and working with GDS layer stacks and ports.fdtdx.readthedocs.io · read 3 Oct 2026
- Intended users
- The project describes FDTDX as freely usable for research, development, and commercial work involving photonic nanostructures.fdtdx.readthedocs.io · read 4 Oct 2026
- JAX
- FDTDX is built on JAX and provides automatic differentiation for gradient-based optimization.fdtdx.readthedocs.io · read 3 Oct 2026
- JAX foundation
- FDTDX is built on JAX and provides automatic differentiation and native GPU support.fdtdx.readthedocs.io · read 4 Oct 2026
- Lesson library
- Nofdtdx.readthedocs.io · read 23 Sep 2026
- License and cost
- FDTDX is freely available for research, development, and commercial use under the MIT license.github.com · read 3 Oct 2026
