FLORAF is the FLORIS wake models and API with support for NVIDIA and Apple GPUs and automatic differentiation. This unlocks running large wind farm studies and gradient-based layout and yaw optimization in a fraction of the time.
FLORIS runtime grows quickly with turbine count. FLORAF keeps large problems interactive.
Illustrative data. Lower is better. Both axes are logarithmic.
Automatic differentiation gives exact gradients in one pass, so optimizers reach good layouts and yaw strategies sooner.
Illustrative data. AEP gain over the initial layout vs. wall-clock time.
Illustrative data. AEP gain from wake steering vs. wall-clock time.
Same wake modeling you trust, built for scale.
| Capability | FLORIS | FLORAF |
|---|---|---|
| Compute | CPU | NVIDIA and Apple GPUs |
| Gradients | Finite differences | Automatic differentiation |
| Layout and yaw optimization | Gradient-free or finite-difference | Gradient-based |
| Large farms and many wind conditions | Runtime grows quickly | Scales on the GPU |
| Workflow | Python API | Familiar FLORIS-style Python API |
| Availability | Open source | Licensed, with a free trial |
Start with a free trial. Upgrade when you're ready.
Academic and Commercial plans are paid at checkout and reviewed before the full license is issued. Your trial key works in the meantime.
Tell us a little about yourself and we'll send your trial key right away.
Get FLORAF and other updates or reach out to discuss your project.
Occasional emails about releases, benchmarks, and new tools. Probably one per quarter. No spam.