Kombyne®  is Intelligent Light’s third-generation simulation co-processing software. Simulation co-processing reduces the amount of data that the simulation must write to disk by performing “post-processing operations” while the solver is running.

The co-processing occurs in either the same computational compute nodes as the simulation or by sending the simulation data to separate compute nodes.  The Kombyne® API is open-source and the product is delivered as source code, with a license that provides uncounted usage.

By co-processing the simulation data in this manner, Kombyne® reduces the total amount of simulation data written to disk by creating data extracts. These extracts consist of point, line or surface samples, coordinate slice planes and iso-surfaces or rendered images.

It is easy to save samples at high temporal rates, so that any subsequent frequency-based analysis will be accurate and is free from aliasing artifacts.  High-fidelity extracts are essential when building surrogates or training AI models.