Khronos is a GPU-accelerated Maxwell FDTD solver written entirely in Julia.
- GPU acceleration compatible with NVIDIA CUDA, AMD ROCm, Intel OneAPI, and Apple Metal (via KernelAbstractions.jl)
- 100% Julia code (works with Windows, Mac, and Linux)
- Specifiable precision (e.g.
Float64,Float32) - 2D (TE/TM) and 3D simulation support
- Parameterizable geometry via GeometryPrimitives.jl with subpixel smoothing
- GDS-II layout import for photonic device geometries
- Diagonally anisotropic permittivity (
$\varepsilon$ ) and permeability ($\mu$ ), and electric/magnetic conductivity terms ($\sigma_D$ /$\sigma_B$ ) for either the permittivity or permeability, respectively. - Dispersive materials via Drude and Lorentzian susceptibility poles (auxiliary differential equation method)
-
$\chi^{(3)}$ Kerr nonlinearity - Predefined current sources (including planewaves, Gaussian beams, and total-field/scattered-field sources) and arbitrary current sources.
- Equivalent sources from predefined electric and magnetic fields (e.g. to inject modes computed from a mode solver).
- Continuous-wave and Gaussian-pulse time profiles
- Arbitrary 1D, 2D, or 3D (rectilinear) DFT monitors
- Flux, mode, and diffraction monitors
- Near-to-far field transformation with GPU-accelerated Green's function evaluation
- Layered-medium far-field projection via transfer matrix method (e.g. project from inside GaN to air without simulating the air region)
- Perfectly matched layer (PML) and absorber boundaries
- Periodic and Bloch boundary conditions
- Multi-GPU support via MPI/NCCL-based domain decomposition with halo exchange
- Concurrent GPU batch execution via multi-stream parallelism
- Adjoint-method inverse design with automatic differentiation (Zygote/ChainRules integration)
- Composable kernel framework with fused PML kernels and per-component specialization
- Meep-compatible Python API (
import khronos.meep as mp) - Predefined simulation runtime functions, including run for an arbitrary time and run until the DFT fields have converged.
- Simple plotting of visualization cross sections and overlaid field response, DFT monitors, and source frequency responses.
- Benchmark tooling to track performance regressions/enhancements on arbitrary hardware
- All simulations require PML or absorber boundaries.
- Uniform gridding.
Khronos is licensed under the MIT license.
