FAQ

You will receive a 20-digit code via email. The first time you launch the app, a license dialog will appear where you can enter this code and register.

sioRAY is a Python-based ray tracing framework for optical simulations. It is designed for engineers and researchers who need to simulate how light propagates through optical systems — including lenses, diffuse surfaces, scattering layers, detectors, and complex multi-surface geometries.

At its core, sioRAY delivers massive simulation throughput through native GPU ray tracing algorithms running directly on NVIDIA RTX hardware — the same acceleration architecture used in real-time rendering, applied here to rigorous physical optics.

sioRAY integrates seamlessly into modern Python-based workflows and is built for deep AI interoperability. From a single line of code to a full-scale multi-surface simulation, AI assistants can leverage the software natively — enabling a new generation of AI-assisted optical design without leaving the coding environment. Crucially, everything the AI produces remains plain, readable Python code: no black boxes, no hidden logic — every simulation step is fully transparent and auditable by the engineer.

sioRAY is built for engineers and researchers who refuse to let artificial limitations of existing tools dictate what is possible. The software is designed around a single philosophy: give the user maximum freedom to operate at the maximum performance of modern hardware.

– **Physics-based simulation** — full Fresnel splitting at every interface, spectrally resolved wavelength grids, and physically calibrated light quantities (radiant flux, irradiance, luminous output) throughout the entire pipeline.

– **Fully scalable scenes** — from a single surface to hundreds of millions of geometric structures, the scene model scales without architectural changes to the workflow.

– **Mesh-based geometry, free from CAD overhead** — geometry is defined directly as triangulated meshes. No CAD kernel, no format conversion pipeline — import what you have or generate geometry programmatically. STEP files remain supported via freely available open-source packages.

– **Deep integration of custom optical properties** — R/T tables and arbitrary BSDF models plug directly into patterned and structured surfaces, giving full control over scatter, diffraction, and thin-film behavior at the triangle level.

– **The data is yours** — all simulation results are returned as plain NumPy arrays. No proprietary result formats, no locked-in post-processing — the output is yours to analyse, export, or feed into any downstream tool.

– **Optimized for batching, tolerancing, and optimization** — the batching module is designed to make full use of available CPU and GPU resources, including multi-GPU configurations, to run parameter sweeps and optimization loops efficiently.

– **AI-readable by design** — the entire API, documentation, and example library are structured to be fully readable by AI assistants. This dramatically shortens onboarding time: instead of weeks spent navigating unfamiliar tooling, a new user can reach productive simulations in hours, guided step by step by an AI that understands the software natively.

Yes. sioRAY includes a structured documentation set:

– **Getting Started guides** that walk you through installation and your first simulation step by step
– **Concept chapters** covering scene setup, materials, ray sources, detectors, visualization, and result analysis
– **Worked example notebooks** grouped by complexity — from minimal examples to advanced multi-surface systems
– **API reference** generated from the public interface, with parameter descriptions and usage notes
– **Cookbook notebooks** for common tasks such as detector comparison, spatial analysis, and BSDF validation

The documentation is designed to let you go from installation to a running simulation in a single session.

Yes. sioRAY ships with a collection of Jupyter notebooks organized into four levels:

– **Getting Started** — minimal working examples, first materials, first detector
– **Intermediate** — volume absorption, patterned surfaces, spatial detector modes, thin-lens camera usage
– **Advanced Mechanics** — BSDF workflows, batching, optimization, freeform lens design
– **Cookbooks & Reference** — focused recipes for specific analysis tasks

All notebooks are self-contained and runnable after a standard installation.

An NVIDIA RTX GPU is required. The minimum is an RTX 3050. Any current consumer or workstation RTX card works well.

The “RTX” designation is important — it indicates dedicated RT cores that sioRAY relies on for ray-triangle intersection. Older GTX cards and data-center GPUs without RT cores (such as A100 or H100) are not supported.

Everything else runs on standard Python hardware. No special CPU or memory requirements beyond what a typical engineering workstation provides.

sioRAY is distributed as a Python wheel and supports **Python 3.13** on **Windows**. A standard conda or pip environment is sufficient.

sioRAY uses an online activation system. After installation, you run a short activation step from Python that opens a license dialog. Once activated, the license is stored locally. An internet connection is not required for regular use; however, the software performs periodic online checks every 5 days to validate the license.

If you are interested in evaluating sioRAY or would like to request a trial license, please feel free to get in touch with us.

1. Receive the installation wheel and follow the installation guide.
2. Run the built-in preflight check to confirm everything is working.
3. Open the **Getting Started** notebooks — they cover the full workflow from scene setup to your first result.
4. Use `trace.help()` inside Python to explore the API interactively at any time.

The documentation and example notebooks are included in the installation and do not require an internet connection after setup.

Yes. Simulations can be run headlessly in scripts, batch jobs, CI pipelines, or optimization loops. The batching module is specifically designed for parameter sweeps and multi-configuration studies.

sioRAY ships with a comprehensive set of resources that cover the most common questions and workflows:

– The **Getting Started notebooks** and **concept chapters** address the majority of typical setup and modelling questions.
– The built-in `trace.help()` command provides instant access to the full public API from within any Python session.
– The **FAQ and troubleshooting sections** in the documentation cover the most frequently encountered issues with geometry, materials, detectors, and simulation setup.

If you run into something that is not covered there, please reach out directly. 

For licensing questions, or technical inquiries, please contact us through the website.