The hawk is on the wing
A native optimization plug-in for Grasshopper: single- and multi-objective solvers behind one component, built for the few thousand evaluations that real design models allow. The first public release is being prepared for food4Rhino.
Fewer evaluations, better designs
Evaluations are only half the budget. On a real model the wall clock decides what you can run, so SilverHawk lets you cap both: a number of evaluations, a time limit, or whichever comes first.
What it is for
One component, many solvers
Connect sliders or gene pools and an objective once, then switch between evolution strategies, differential evolution, swarms and multi-objective methods without rewiring.
Built for small budgets
Stop a run after a fixed number of evaluations or a fixed time, repeat it in batches, and compare solvers under the same conditions on your own model.
Runs you can retrace
Every run writes a record with its settings and results, which can be reinstated on the canvas to check a design or a published result.
Flight log
Where the project stands before release.
- DoneSolvers20 entries: 11 single-objective and 9 multi-objective, including hybrids designed for small budgets and an automatic selector.
- DoneResearchTwo preprints on arXiv document the solvers and how optimizer implementations differ across design tools.
- In flightRelease buildFinal tests of the plug-in on Rhino 8 (Windows).
- In flightTutorialsShort videos with example files, from a first run to multi-objective studies.
- NextRelease on food4RhinoFree download, versioned and archived for reproducible research.
Independent research. SilverHawk is developed and released without external funding. About the project →
Rhino 8
Grasshopper
C# / .NET
Rhino and Grasshopper are trademarks of Robert McNeel & Associates; C# and .NET are trademarks of Microsoft. SilverHawk is an independent plug-in and is not endorsed by either.