Six numerical engines for VCSELs and semiconductor devices.
Transport, optical and thermal solvers, all ready and validated, wired together in Flow. Run it on your own machine or on our EU cloud.
Three physics, six numerical engines.
Each one is production ready on its own and designed to couple with the others.
Transport
Drift-diffusion carrier transport for lasers, detectors, CMOS and power devices.
- Currents and electrical characteristics
- Carrier densities and potential
- Heterostructures, doping and contacts
Optical
Four solvers behind one interface, from fast VCSEL modes to full-wave propagation.
- CMT: full-vector 3D modal analysis for VCSELs
- FEM: modes of arbitrary cross sections
- FDTD: full-wave time domain
- Ray tracing: geometric-optics propagation
Thermal
FEM heat transfer coupled to optics and transport.
- Internal temperature distribution
- Self-heating effects
- Thermal lensing
- Temperature-induced index changes
Available and validated today.
| Module | What it does | Status |
|---|---|---|
| Transport | Drift-diffusion carrier transport. Validated on photodetectors, CMOS and diodes. | Available and validated |
| Optical: CMT | Full-vector 3D modal analysis from our VCSEL research. | Available and validated |
| Optical: FEM | Finite-element modal solver for arbitrary cross sections. | Available and validated |
| Optical: FDTD | Full-wave time-domain solver. | Available and validated |
| Optical: ray tracing | Geometric-optics propagation. | Available and validated |
| Thermal | FEM heat transfer, coupled to optics and transport. | Available and validated |
The VCSEL optical engine, our foundation.
The coupled mode theory solver comes straight from more than twenty years of VCSEL research.
- Transverse modes, emission wavelengths and threshold gain
- Longitudinal confinement factors and outcoupling efficiency
- Near-field and far-field distributions, polarization, Poynting vector
- Arbitrary transverse geometries, surface patterning and complex apertures
- Seconds to minutes per structure, no brute-force PDE solving
Answers, not just solvers.
Companies do not license a solver, they need answers before a design goes to fabrication.
Will it survive at operating current?
Couple optics and thermal to predict rollover current and reliability margin, so a marginal design shows up in simulation instead of in qualification.
Which aperture wins?
Sweep the oxide aperture and compare beam quality and thermal impedance for every variant, so the trade-off is chosen on purpose.
Which of fifty variants to fabricate?
Batch-run every geometry and drive current overnight and review a ranked shortlist in the morning.
One canvas for all the physics.
Connect solvers with single arrows for one-way passes and double arrows for self-consistent loops. Let the local AI agent suggest the setup, write the input files and run the simulations.
GUI or Python. Same solvers.
GUI
Import a structure, sweep parameters with sliders and watch mode profiles, beam patterns and temperature maps update live.
Import structure
Read standard epi-structure files straight from your existing tools, in the GUI or in a script, with no manual re-entry of layer stacks.
Scripting
Full Python access to every computed quantity. Loop over geometries and drive currents and feed results into your own pipeline.
One aperture sweep, from geometry to reliability margin, in a few lines (illustrative example):
import polite as plt
for aperture_um in [4, 6, 8, 10]:
optical = plt.OpticalSolver(structure="wafer_A.epi", aperture_um=aperture_um)
modes = optical.solve()
thermal = plt.ThermalSolver.from_optical(optical, drive_current_mA=8)
rollover_mA = thermal.find_rollover_current()
print(aperture_um, modes.threshold_gain, rollover_mA)Purpose-built, not adapted.
Device physics first
Dedicated solvers born from VCSEL research, extended to the semiconductor devices that share the same physics.
You choose the coupling
One-way where it is enough, self-consistent where the physics demands it. You decide where computation is spent.
IP stays with you
Local by default, with the AI agent included. No HPC cluster required for the VCSEL workflows. How local and cloud work.
Every device program has its own constraints.
Our engineers work directly with yours to fit Po.Li.Te into the program.
Custom templates
Simulation templates matched to your epitaxial stack, wavelength or fabrication process, so every project starts from a validated baseline.
Custom features
A new figure of merit, a solver extension or an integration with your design tools, built to your specification.
Tool integration
We connect the external tools you already use to Flow, so Po.Li.Te sits inside your toolchain.
Licensed per seat. Deployed your way.
Annual or multi-year
Single-user licenses on annual or multi-year terms, so the commitment matches your project timeline.
Discounts for teams
Multi-user licenses come with volume discounts that scale with your team size.
Local or EU cloud
Install on your own hardware, offline if needed, or use our cloud solution hosted in the European Union. Learn more.
Frequently asked questions
What does Po.Li.Te simulate?
Po.Li.Te is a multiphysics simulation platform for semiconductor devices. It couples drift-diffusion transport, optical solvers (coupled mode theory, FEM, FDTD and ray tracing) and thermal solvers. It was born from VCSEL modeling and now covers most semiconductor device families, including photodetectors, CMOS, diodes and power devices.
What is Flow?
Flow is the graphical workflow environment of Po.Li.Te. You connect solvers and external tools on a canvas with single arrows for one-way passes and double arrows for self-consistent loops, and an AI agent can write the flow and the input files for you.
Where does my data live? Is the AI agent cloud-based?
Po.Li.Te runs locally by default, including the AI agent, which uses a compact language model on a standard consumer graphics card. Your designs stay on your hardware. If you prefer, we offer a cloud solution hosted in the European Union.
How is Po.Li.Te different from general-purpose TCAD tools?
General-purpose TCAD suites are adapted to many device types. Po.Li.Te started from dedicated, device-specific solvers validated against the team’s own peer-reviewed VCSEL research, and extends them with transport, thermal and further optical solvers in one workflow where you choose the coupling.
Can Po.Li.Te be automated or scripted?
Yes. You can work in the GUI or in Python to sweep parameters, batch-run design variants and feed results into your own analysis pipeline.
Is Po.Li.Te validated against measurements or published research?
Yes. The VCSEL optical engine is built on the team’s peer-reviewed research in Optics Express, Optics Letters and IEEE Photonics Journal. The transport module is validated on photodetectors and power semiconductor devices such as CMOS and diodes.