SPINS Photonics
HyperWave
Inverse Design for Next-Generation Photonics.
Industrial-grade 3D FDTD built for efficient, manufacturable device optimization.
Inverse Design in Action
Watch gradient-based optimization converge to high-efficiency photonic devices in real time.
import hyperwave_community as hwc
import gdsfactory as gf
# Load photonic component and discretize
device = gf.components.mmi2x2_with_sbend()
theta, info = hwc.component_to_theta(device, resolution=0.02)
# Build 3D structure (SOI stack)
structure = hwc.create_structure(layers=[
hwc.Layer(density_clad, eps_clad, clad_cells),
hwc.Layer(density_core, (eps_clad, eps_core), wg_cells),
hwc.Layer(density_clad, eps_clad, clad_cells),
])
# Solve waveguide mode and run on cloud GPU
source, offset, mode = hwc.create_mode_source(structure, ...)
results = hwc.simulate(
structure_recipe=structure.extract_recipe(),
source_field=source,
simulation_steps=20000,
)
# Analyze port transmission
hwc.analyze_transmission(results, input_monitor="Input_o1")Developer-Friendly. Enterprise-Ready.
A clean, Pythonic API built for photonics engineers. Define your simulation, run optimization, and analyze results, directly from a notebook or production environment.
GPU-Accelerated Backend
Optimized 3D FDTD powered by NVIDIA B200 GPUs for high-throughput simulation and optimization.
Integrated Adjoint Differentiation
Compute gradients directly within the solver for efficient inverse design.
Jupyter-Native Workflow
Inline visualization with seamless integration into Jupyter and Colab notebooks.
GDSII Import and Export
Build from GDSFactory components and export fabrication-ready designs.
Simple, Transparent Pricing
Usage-based pricing for individuals. Custom infrastructure options for enterprise teams.
Why HyperWave
Built for photonics engineers who demand speed, control, and practical workflows.
Accelerate Design Cycles
GPU-accelerated 3D FDTD delivers simulation results in minutes, enabling rapid iteration compared to traditional CPU-based workflows.
Purpose-Built for Inverse Design
Automatic differentiation enables efficient gradient-based topology optimization that converges to fabrication-ready designs.
Python-First Developer Experience
A clean Python SDK with Jupyter and Colab support, letting you offload compute to cloud GPUs, integrate with GDSII, and manage your full simulation workflow programmatically.
Ready to accelerate your photonics design?
Get started and run your first GPU-accelerated simulation today.